Assignment2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "8ecfa60c",
   "metadata": {},
   "source": [
    "# Applied Machine Learning 2nd Assignment\n",
    "\n",
    "Τhis assignment contains a replication of a study of belief in supernatural evil and attitudes towards guns in the United States. The study is:\n",
    "\n",
    "Christopher G. Ellison, Benjamin Dowd-Arrow, Amy M. Burdette, Pablo E. Gonzalez, Margaret S. Kelley, Paul Froese, \"Peace through superior firepower: Belief in supernatural evil and attitudes toward gun policy in the United States\", Social Science Research, Volume 99, 2021, https://doi.org/10.1016/j.ssresearch.2021.102595.\n",
    "\n",
    "---\n",
    "\n",
    "> Georgios Liargkovas, Undergraduate Student <br />\n",
    "> Department of Management Science and Technology <br />\n",
    "> Athens University of Economics and Business <br />\n",
    "> gliargovas@aueb.gr"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4484e2ee",
   "metadata": {},
   "source": [
    "## Pre-work\n",
    "\n",
    "### Import libraries to use\n",
    "\n",
    "We procceed by importing the libraries used in this assignment. In the "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "cb5afee1",
   "metadata": {},
   "outputs": [],
   "source": [
    "import math\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ef442fd8",
   "metadata": {},
   "source": [
    "### Downloading the dataset\n",
    "\n",
    "* Before attempting to tackle the assignment's questions, the correct dataset utilized in the paper above should be downloaded.\n",
    "\n",
    "\n",
    "* In sec. 1.4. of the aforementioned paper, it is stated that the data analyzed come from the **Wave Four of the Baylor Religion Survey** (BRS), conducted in January 2014.\n",
    "\n",
    "\n",
    "* Additionally, the dataset contains some missing values, that are addressed via multiple imputation. The imputation will be conducted after coding the variables.\n",
    "\n",
    "\n",
    "* By searching for the study on Google, we come accross the survey's website where we are redirected to the ARDA website, to download the dataset. The EXCEL (.xlsx) format is downloaded. The download link can be found [here](https://osf.io/hf9u3/download). \n",
    "\n",
    "\n",
    "* An initial attempt for reading the file resulted in a parse error. It turned out that the file contained text formatting that prohibited pandas from reading it correctly. \n",
    "\n",
    "We proceed to read the dataset from the .sav file:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "ac2d26cc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>MOTHERLODE_ID</th>\n",
       "      <th>PROJECT_ID</th>\n",
       "      <th>RESPONDENT_ID</th>\n",
       "      <th>METHOD_ID</th>\n",
       "      <th>METHOD_TYPE</th>\n",
       "      <th>PROJECT_NUMBER</th>\n",
       "      <th>CREATED_ON</th>\n",
       "      <th>PRACTICE</th>\n",
       "      <th>RESPONDENT_DATE</th>\n",
       "      <th>RESPONDENT_LANGUAGE</th>\n",
       "      <th>...</th>\n",
       "      <th>STATE</th>\n",
       "      <th>PROJWT</th>\n",
       "      <th>WEIGHT</th>\n",
       "      <th>Q2</th>\n",
       "      <th>RELTRAD</th>\n",
       "      <th>I_GENDER</th>\n",
       "      <th>I_EDUC</th>\n",
       "      <th>I_MARITAL</th>\n",
       "      <th>I_RELIGION</th>\n",
       "      <th>I_ATTEND</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>128170892</td>\n",
       "      <td>33370</td>\n",
       "      <td>01_000001_00000001</td>\n",
       "      <td>162048140_01</td>\n",
       "      <td>SCAN</td>\n",
       "      <td>162048140</td>\n",
       "      <td>3/26/2014</td>\n",
       "      <td>OTHER</td>\n",
       "      <td>1/24/2014</td>\n",
       "      <td>en-US</td>\n",
       "      <td>...</td>\n",
       "      <td>TX</td>\n",
       "      <td>20086.696606</td>\n",
       "      <td>0.270000</td>\n",
       "      <td>Church of Christ</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>128170893</td>\n",
       "      <td>33370</td>\n",
       "      <td>01_000002_00000001</td>\n",
       "      <td>162048140_01</td>\n",
       "      <td>SCAN</td>\n",
       "      <td>162048140</td>\n",
       "      <td>3/26/2014</td>\n",
       "      <td>OTHER</td>\n",
       "      <td>1/24/2014</td>\n",
       "      <td>en-US</td>\n",
       "      <td>...</td>\n",
       "      <td>TX</td>\n",
       "      <td>26218.573175</td>\n",
       "      <td>0.352423</td>\n",
       "      <td>Church of God In Christ</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>128170894</td>\n",
       "      <td>33370</td>\n",
       "      <td>01_000003_00000001</td>\n",
       "      <td>162048140_01</td>\n",
       "      <td>SCAN</td>\n",
       "      <td>162048140</td>\n",
       "      <td>3/26/2014</td>\n",
       "      <td>OTHER</td>\n",
       "      <td>1/24/2014</td>\n",
       "      <td>en-US</td>\n",
       "      <td>...</td>\n",
       "      <td>ND</td>\n",
       "      <td>41408.889891</td>\n",
       "      <td>0.556608</td>\n",
       "      <td>ELCA</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>128170895</td>\n",
       "      <td>33370</td>\n",
       "      <td>01_000004_00000001</td>\n",
       "      <td>162048140_01</td>\n",
       "      <td>SCAN</td>\n",
       "      <td>162048140</td>\n",
       "      <td>3/26/2014</td>\n",
       "      <td>OTHER</td>\n",
       "      <td>1/24/2014</td>\n",
       "      <td>en-US</td>\n",
       "      <td>...</td>\n",
       "      <td>TX</td>\n",
       "      <td>50023.027356</td>\n",
       "      <td>0.672397</td>\n",
       "      <td>Independent Methodist</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>128170896</td>\n",
       "      <td>33370</td>\n",
       "      <td>01_000005_00000001</td>\n",
       "      <td>162048140_01</td>\n",
       "      <td>SCAN</td>\n",
       "      <td>162048140</td>\n",
       "      <td>3/26/2014</td>\n",
       "      <td>OTHER</td>\n",
       "      <td>1/24/2014</td>\n",
       "      <td>en-US</td>\n",
       "      <td>...</td>\n",
       "      <td>IA</td>\n",
       "      <td>230734.428388</td>\n",
       "      <td>3.101474</td>\n",
       "      <td>Don't have specific name</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1567</th>\n",
       "      <td>128172920</td>\n",
       "      <td>33371</td>\n",
       "      <td>02_000301_00000013</td>\n",
       "      <td>162048140_02</td>\n",
       "      <td>SCAN</td>\n",
       "      <td>162048140</td>\n",
       "      <td>3/26/2014</td>\n",
       "      <td>OTHER</td>\n",
       "      <td>3/10/2014</td>\n",
       "      <td>es-AR</td>\n",
       "      <td>...</td>\n",
       "      <td>CA</td>\n",
       "      <td>17054.912661</td>\n",
       "      <td>0.229248</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1568</th>\n",
       "      <td>128172921</td>\n",
       "      <td>33371</td>\n",
       "      <td>02_000326_00000014</td>\n",
       "      <td>162048140_02</td>\n",
       "      <td>SCAN</td>\n",
       "      <td>162048140</td>\n",
       "      <td>3/26/2014</td>\n",
       "      <td>OTHER</td>\n",
       "      <td>3/12/2014</td>\n",
       "      <td>es-AR</td>\n",
       "      <td>...</td>\n",
       "      <td>CA</td>\n",
       "      <td>104454.336796</td>\n",
       "      <td>1.404049</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1569</th>\n",
       "      <td>128172922</td>\n",
       "      <td>33371</td>\n",
       "      <td>02_000351_00000015</td>\n",
       "      <td>162048140_02</td>\n",
       "      <td>SCAN</td>\n",
       "      <td>162048140</td>\n",
       "      <td>3/26/2014</td>\n",
       "      <td>OTHER</td>\n",
       "      <td>3/19/2014</td>\n",
       "      <td>es-AR</td>\n",
       "      <td>...</td>\n",
       "      <td>CA</td>\n",
       "      <td>153453.642034</td>\n",
       "      <td>2.062685</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1570</th>\n",
       "      <td>128172335</td>\n",
       "      <td>33371</td>\n",
       "      <td>02_000352_00000015</td>\n",
       "      <td>162048140_02</td>\n",
       "      <td>SCAN</td>\n",
       "      <td>162048140</td>\n",
       "      <td>3/26/2014</td>\n",
       "      <td>OTHER</td>\n",
       "      <td>3/19/2014</td>\n",
       "      <td>es-AR</td>\n",
       "      <td>...</td>\n",
       "      <td>CT</td>\n",
       "      <td>33324.200008</td>\n",
       "      <td>0.447935</td>\n",
       "      <td>Catolica Romana - Latinos</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1571</th>\n",
       "      <td>128172923</td>\n",
       "      <td>33371</td>\n",
       "      <td>02_000376_00000016</td>\n",
       "      <td>162048140_02</td>\n",
       "      <td>SCAN</td>\n",
       "      <td>162048140</td>\n",
       "      <td>3/26/2014</td>\n",
       "      <td>OTHER</td>\n",
       "      <td>3/21/2014</td>\n",
       "      <td>es-AR</td>\n",
       "      <td>...</td>\n",
       "      <td>GA</td>\n",
       "      <td>117172.713311</td>\n",
       "      <td>1.575006</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1572 rows × 282 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      MOTHERLODE_ID  PROJECT_ID       RESPONDENT_ID     METHOD_ID METHOD_TYPE  \\\n",
       "0         128170892       33370  01_000001_00000001  162048140_01        SCAN   \n",
       "1         128170893       33370  01_000002_00000001  162048140_01        SCAN   \n",
       "2         128170894       33370  01_000003_00000001  162048140_01        SCAN   \n",
       "3         128170895       33370  01_000004_00000001  162048140_01        SCAN   \n",
       "4         128170896       33370  01_000005_00000001  162048140_01        SCAN   \n",
       "...             ...         ...                 ...           ...         ...   \n",
       "1567      128172920       33371  02_000301_00000013  162048140_02        SCAN   \n",
       "1568      128172921       33371  02_000326_00000014  162048140_02        SCAN   \n",
       "1569      128172922       33371  02_000351_00000015  162048140_02        SCAN   \n",
       "1570      128172335       33371  02_000352_00000015  162048140_02        SCAN   \n",
       "1571      128172923       33371  02_000376_00000016  162048140_02        SCAN   \n",
       "\n",
       "      PROJECT_NUMBER CREATED_ON PRACTICE RESPONDENT_DATE RESPONDENT_LANGUAGE  \\\n",
       "0          162048140  3/26/2014    OTHER       1/24/2014               en-US   \n",
       "1          162048140  3/26/2014    OTHER       1/24/2014               en-US   \n",
       "2          162048140  3/26/2014    OTHER       1/24/2014               en-US   \n",
       "3          162048140  3/26/2014    OTHER       1/24/2014               en-US   \n",
       "4          162048140  3/26/2014    OTHER       1/24/2014               en-US   \n",
       "...              ...        ...      ...             ...                 ...   \n",
       "1567       162048140  3/26/2014    OTHER       3/10/2014               es-AR   \n",
       "1568       162048140  3/26/2014    OTHER       3/12/2014               es-AR   \n",
       "1569       162048140  3/26/2014    OTHER       3/19/2014               es-AR   \n",
       "1570       162048140  3/26/2014    OTHER       3/19/2014               es-AR   \n",
       "1571       162048140  3/26/2014    OTHER       3/21/2014               es-AR   \n",
       "\n",
       "      ...  STATE         PROJWT    WEIGHT                         Q2  RELTRAD  \\\n",
       "0     ...     TX   20086.696606  0.270000           Church of Christ      1.0   \n",
       "1     ...     TX   26218.573175  0.352423    Church of God In Christ      3.0   \n",
       "2     ...     ND   41408.889891  0.556608                       ELCA      2.0   \n",
       "3     ...     TX   50023.027356  0.672397      Independent Methodist      2.0   \n",
       "4     ...     IA  230734.428388  3.101474   Don't have specific name      1.0   \n",
       "...   ...    ...            ...       ...                        ...      ...   \n",
       "1567  ...     CA   17054.912661  0.229248                        NaN      4.0   \n",
       "1568  ...     CA  104454.336796  1.404049                        NaN      4.0   \n",
       "1569  ...     CA  153453.642034  2.062685                        NaN      6.0   \n",
       "1570  ...     CT   33324.200008  0.447935  Catolica Romana - Latinos      4.0   \n",
       "1571  ...     GA  117172.713311  1.575006                        NaN      4.0   \n",
       "\n",
       "      I_GENDER I_EDUC  I_MARITAL  I_RELIGION  I_ATTEND  \n",
       "0          2.0    5.0        3.0         1.0       6.0  \n",
       "1          2.0    3.0        3.0         3.0       2.0  \n",
       "2          2.0    3.0        1.0         2.0       6.0  \n",
       "3          2.0    4.0        3.0         2.0       5.0  \n",
       "4          1.0    2.0        2.0         1.0       6.0  \n",
       "...        ...    ...        ...         ...       ...  \n",
       "1567       2.0    5.0        2.0         4.0       2.0  \n",
       "1568       2.0    2.0        2.0         4.0       1.0  \n",
       "1569       2.0    3.0        1.0         5.0       6.0  \n",
       "1570       1.0    5.0        2.0         4.0       3.0  \n",
       "1571       2.0    1.0        NaN         4.0       4.0  \n",
       "\n",
       "[1572 rows x 282 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_excel(\"data/Baylor Religion Survey, Wave IV (2014).xlsx\")\n",
    "data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7c6225ee",
   "metadata": {},
   "source": [
    "### Q1: Belief in Supernatural Evil Metric\n",
    "\n",
    "* For this question, we will be examining three variables. To select the correct varuables to use, it was necessary to examine the dataset's labels. Question group 23 contains questions about belief in supernatural forces. From there we can spot the questions that concern us:\n",
    "\n",
    " * **Q23A**: whether the respondent believes in the devil.\n",
    " * **Q23C**: whether the respondent believes in hell.\n",
    " * **Q23G**: whether the respondent believes in demons.\n",
    " \n",
    " \n",
    "* Additionally, in the paper, it is described that the use of BRS sample weights help align the responses of the survey to the characteristics of the adult population of the continental United States. \n",
    "\n",
    "* However, every analysis performed does not take weights into account, so we won't be using them in our analysis either.\n",
    "\n",
    "We create a new dataframe containing the answers of the above questions with more descriptive labels:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f684c35a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Satan</th>\n",
       "      <th>Hell</th>\n",
       "      <th>Demons</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1567</th>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1568</th>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1569</th>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1570</th>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1571</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1572 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Satan  Hell  Demons\n",
       "0       4.0   4.0     4.0\n",
       "1       4.0   4.0     4.0\n",
       "2       3.0   2.0     2.0\n",
       "3       4.0   4.0     4.0\n",
       "4       4.0   4.0     4.0\n",
       "...     ...   ...     ...\n",
       "1567    4.0   2.0     4.0\n",
       "1568    3.0   3.0     3.0\n",
       "1569    4.0   4.0     3.0\n",
       "1570    3.0   3.0     3.0\n",
       "1571    NaN   NaN     NaN\n",
       "\n",
       "[1572 rows x 3 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "q1 = data[[\"Q23A\", \"Q23C\", \"Q23G\"]]\n",
    "q1.columns = [\"Satan\", \"Hell\", \"Demons\"]\n",
    "q1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d72ef1a1",
   "metadata": {},
   "source": [
    "* We want to combine the three responses into one single metric.\n",
    "\n",
    "\n",
    "* We can think of this as a problem of describing three different dimentions (belief in satan, hell and demons) with just one dimention (belief in supernatural evil). \n",
    "\n",
    "\n",
    "* This can be done by conducting a **Principal Component Analysis (PCA)**.\n",
    "\n",
    "\n",
    "* In the above dataframe, the rows represent our samples and the columns represent the different features (belief in satan, hell and demons). So, for each sample, we have three features.\n",
    "\n",
    "\n",
    "* We are now going to conduct PCA with a signle dimention to the above three questions. PCA requires our data to be free of missing values. We haven't conducted any form of imputation yet, so we are going to conduct PCA by ignoring the missing values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "5f9359c1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PCA(n_components=1)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.decomposition import PCA\n",
    "\n",
    "pca = PCA(n_components=1)\n",
    "pca.fit(q1.dropna())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a6ff4d99",
   "metadata": {},
   "source": [
    "Let's have a look to the explained variance ratio of the PCA:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "634764aa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.90171495])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca.explained_variance_ratio_"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1eccab42",
   "metadata": {},
   "source": [
    "* Principal Component Analysis explains a very large percentage (just above 90%) of the total variance of the three metrics. \n",
    "* This means that combining them into a single metric can be done without significant information loss.\n",
    "\n",
    "\n",
    "* We have seen that the components can be combined together, but how do they combine? \n",
    "* For this, we will have to take a look at the PCA components:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "33a39dbb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.58568735, -0.57818501, -0.56804263]])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca.components_"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9c1745b2",
   "metadata": {},
   "source": [
    "* So from the above, we can conclude that belief in satan, hell and demons can be combined into a single metric: **Supernatural Evil** (SE). The coefficients of the different beliefs are the following:\n",
    "\n",
    "$$Supernatural Evil = 0,586 * Satan + 0,578 * Hell + 0,568 * Demons$$\n",
    "\n",
    "* The coefficients are very close to ane another, which indicates that the authors' approach to combine them into a single metric by calculating their average (the average assumes all variables have the same weight) can adequately describe the relationship between the three metrics and be used to combine them into a single one. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "df9d799e",
   "metadata": {},
   "source": [
    "### Q2: Variables Selection\n",
    "\n",
    "* Fow this section of the assignment, we need to select and rename the different columns of the dataset in order to make it easier to process.\n",
    "\n",
    "* We identify the related to the question variables by reading the dataset description.\n",
    "\n",
    "Below are our variables with their column and their newer encoding (**ans** represents the specific option/s in a question if the question has to be coded into dummy variables):\n",
    "\n",
    "* Dependent Variables\n",
    "\n",
    "  * **Q73A**: Ban on Semi-Auto Guns\n",
    "  * **Q73E**: Ban on High-Capacity Ammo Clips \n",
    "  * **Q73G**: Banning Civilian Handguns\n",
    "  * **Q73H**: Support for Concealed Carry Laws\n",
    "  * **Q73C**: More Armed Security at Schools\n",
    "  * **Q73F**: More Teachers/Faculty having Guns\n",
    "  * **Q73B**: More Gun Safety Programs\n",
    "  * **Q73D**: Expanded Mental Health Screening\n",
    "  \n",
    "* Independept Variables\n",
    " \n",
    "   * Religious Variables\n",
    "     * **Q4**: Attendance\n",
    "     * **Q17 (ans: 3,4,8)**: Bible (Human Error, History and Fables) \n",
    "     * **Q17 (ans: 2)**: Biblical Inerrancy\n",
    "     * **Q17 (ans: 1)**: Biblical Literalism\n",
    "     \n",
    "  * Religious Affiliation\n",
    "    * **I_RELIGION (ans: 1)**: Conservative Protestant (Evangelical)\n",
    "    * **I_RELIGION (ans: 2)**: Mainline Protestant \n",
    "    * **I_RELIGION (ans: 3)**: Black Protestant (African Methodist)\n",
    "    * **I_RELIGION (ans: 4)**: Catholic\n",
    "    * **I_RELIGION (ans: 5)**: Other (Judaism, Islam, other non-Christian world religion, minority faiths)\n",
    "    * **I_RELIGION (ans: 6)**: No Affiliation\n",
    "  * **Q31**: Political Ideology\n",
    "  * **AGE**: Age\n",
    "  * Sex\n",
    "    * **Q77 (ans: 2)**: Female\n",
    "    * **Q77 (ans: 1)**: Male\n",
    "  * Race\n",
    "    * **Q88A** : White\n",
    "    * **Q89 (ans: 2,3,4)**: Hispanic\n",
    "    * **Q88B** : African American/Black \n",
    "    * **Q88C,Q88D,Q88E,Q88F**: Other\n",
    "  * Education\n",
    "    * **Q90 (ans: 1,2)**: Less Than High School \n",
    "    * **Q90 (ans: 3)**: High School or Equivalent \n",
    "    * **Q90 (ans: 4, 5)**: Some College\n",
    "    * **Q90 (ans: 6)**: College Degree (contains technical degree)\n",
    "    * **Q90 (ans: 7)**: Post-graduate Degree \n",
    "  * **Q95**: Household Income\n",
    "  * Marital Status\n",
    "    * **I_MARITAL (ans: 1,3,4)**: Not Partnered/Single \n",
    "    * **I_MARITAL (ans: 2)**: Married/Cohabitating\n",
    "  * Children\n",
    "    * **Q92_NONE**: No kids under 18 in home \n",
    "    * **Q92**: Kids under 18 in home\n",
    "  * Area Demographics\n",
    "    * **Q80 (ans: 3,4)**: Small Town/Rural\n",
    "    * **Q80 (ans: 1,2)**: Urban Area\n",
    "  * Region\n",
    "    * **STATE (ans: MD,DE,VA,WV,KY,TN,NC,SC,FL,GA,AL,MS,LA,AK,TX,OK)**: South\n",
    "    * **STATE (ans: not belonging in the above)**: Other Region"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0f3c76a",
   "metadata": {},
   "source": [
    "Now, let's code the variables and create dummy variables where needed.\n",
    "\n",
    "For this purpose, we create a new DataFrame:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "93967445",
   "metadata": {},
   "outputs": [],
   "source": [
    "q2 = pd.DataFrame()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "074cfee2",
   "metadata": {},
   "source": [
    "* During the whole variable coding process, we must be very careful with NaNs. Every coding approach utilized taked into consieration the missing values and does not alter them, unless described otherwise.\n",
    "\n",
    "\n",
    "* Firstly, we will code the dependent variables. This can be done very easily with the approach below, which also keeps the missing values intact. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "ea2b91b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "q2[\"BAN_SEMI_AUTO\"] = 2 - data.Q73A\n",
    "q2[\"BAN_AMMO_CLIPS\"] = 2 - data.Q73E\n",
    "q2[\"BAN_HANDGUNS\"] = 2 - data.Q73G\n",
    "q2[\"SUPPORT_CONCEALED_CARRY\"] = 2 - data.Q73H\n",
    "q2[\"ARMED_SECURITY\"] = 2 - data.Q73C\n",
    "q2[\"ARMED_TEACHERS\"] = 2 - data.Q73F\n",
    "q2[\"MORE_GUN_SAFETY\"] = 2 - data.Q73B\n",
    "q2[\"EXPANDED_SCREENING\"] = 2 - data.Q73D"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b5d9a855",
   "metadata": {},
   "source": [
    "* Then, we will proceed with some easy-to-code variables, such as the belief in diffferent supernatural evil powers, the religious attendance, the political ideology and the income of the respondents.\n",
    "\n",
    "* Note that there are two versions of the religious attendance metric in the variables. The one we chose is the one that ranges from 1-8, which aligns with the one used in the **Appendix A table** (range 1-8)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "7b72ca54",
   "metadata": {},
   "outputs": [],
   "source": [
    "q2[\"DEVIL\"] = data.Q23A\n",
    "q2[\"HELL\"] = data.Q23C\n",
    "q2[\"DEMONS\"] = data.Q23G\n",
    "\n",
    "q2[\"ATTENDANCE\"] = data.Q4\n",
    "\n",
    "q2[\"POLITICAL_IDEOLOGY\"] = data.Q31\n",
    "\n",
    "q2[\"INCOME\"] = data.Q95"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "55e1f412",
   "metadata": {},
   "source": [
    "* We then proceed with the respondents' beliefs about the bible.\n",
    "* In this case, \"I don't know\" answers were mapped as NaN. We will assume that all NaNs are classified as \"I don't know\", which means that we know the category they belong to and therefore will be mappped as 0, instead of NaN. \n",
    "* This approach leads to an identical value in the descriptive statistics section and lead to more identical regression results in q3.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "50b9be11",
   "metadata": {},
   "outputs": [],
   "source": [
    "q2[\"BIBLICAL_LITERALISM\"] = np.where(data.Q17 == 1, 1, np.where(data.Q17.isna(), np.NaN, 0))\n",
    "q2[\"BIBLICAL_INERRACY\"] = np.where(data.Q17 == 2, 1, np.where(data.Q17.isna(), np.NaN, 0))\n",
    "q2[\"BIBLE_HUMAN_ERROR\"] = np.where((data.Q17 == 3) | \n",
    "                                   (data.Q17 == 4) | \n",
    "                                   (data.Q17 == 8), 1, np.where(data.Q17.isna(), np.NaN, 0))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6455f57c",
   "metadata": {},
   "source": [
    "* In the respondents' religious affiliation, missing values were left as is. I_RELIGION variable was utilised as it is already coded in the way we want (one category per variable value). \n",
    "\n",
    "* For leaving missing values intact the following numpy.where() format was used:\n",
    "\n",
    "\n",
    "    np.where(question == value, 1, np.where(question.isna(), np.NaN, 0))\n",
    "\n",
    "\n",
    "* This way, zeros will remain zero and NaNs will remain NaN."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "448e07f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "q2[\"CONSERVATIVE_PROTESTANT\"] = np.where((data.I_RELIGION == 1), 1, np.where(data.I_RELIGION.isna(), np.NaN, 0))\n",
    "q2[\"MAINLINE_PROTESTANT\"] = np.where((data.I_RELIGION == 2), 1, np.where(data.I_RELIGION.isna(), np.NaN, 0))\n",
    "q2[\"BLACK_PROTESTANT\"] = np.where((data.I_RELIGION == 3), 1, np.where(data.I_RELIGION.isna(), np.NaN, 0))\n",
    "q2[\"CATHOLIC\"] = np.where((data.I_RELIGION == 4), 1, np.where(data.I_RELIGION.isna(), np.NaN, 0))\n",
    "q2[\"OTHER_RELIGION\"] = np.where(data.I_RELIGION == 5, 1, np.where(data.I_RELIGION.isna(), np.NaN, 0))\n",
    "q2[\"NO_AFFILIATION\"] = np.where((data.I_RELIGION == 6), 1, np.where(data.I_RELIGION.isna(), np.NaN, 0))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a0dff930",
   "metadata": {},
   "source": [
    "* A similar approach was used for education level variables, NaNs are left intact:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "6404ad3d",
   "metadata": {},
   "outputs": [],
   "source": [
    "q2[\"LESS_THAN_HS\"] = np.where((data.Q90 == 1) | (data.Q90 == 2), 1, np.where(data.Q90.isna(), np.NaN, 0))\n",
    "q2[\"HIGH_SCHOOL_OR_EQUIVALENT\"] = np.where(data.Q90 == 3, 1, np.where(data.Q90.isna(), np.NaN, 0))\n",
    "q2[\"SOME_COLLEGE\"] = np.where((data.Q90 == 4) | (data.Q90 == 5), 1, np.where(data.Q90.isna(), np.NaN, 0))\n",
    "q2[\"COLLEGE_GRADUATE\"] = np.where((data.Q90 == 6), 1, np.where(data.Q90.isna(), np.NaN, 0))\n",
    "q2[\"POSTGRADUATE\"] = np.where((data.Q90 == 7), 1, np.where(data.Q90.isna(), np.NaN, 0))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58ee7d8a",
   "metadata": {},
   "source": [
    "* Regarding age, it was noticed that 104 responses had a value of 0:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "3186e39b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0     104\n",
       "67     40\n",
       "59     40\n",
       "58     39\n",
       "61     37\n",
       "     ... \n",
       "95      2\n",
       "99      2\n",
       "89      2\n",
       "94      2\n",
       "96      1\n",
       "Name: AGE, Length: 80, dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.AGE.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39f008a4",
   "metadata": {},
   "source": [
    "* It is obvious that these are missing values so they were replaced by NaN."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "e3d77511",
   "metadata": {},
   "outputs": [],
   "source": [
    "q2[\"AGE\"] = np.where(data.AGE > 0, data.AGE, np.NaN)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8e8867be",
   "metadata": {},
   "source": [
    "* Genders were easily mapped to separate variables as seen below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "304cb99d",
   "metadata": {},
   "outputs": [],
   "source": [
    "q2[\"FEMALE\"] = data.Q77 - 1\n",
    "q2[\"MALE\"] = 2 - data.Q77"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0124e695",
   "metadata": {},
   "source": [
    "* For the marital status, we follow a similar approach as the other dummy variables.\n",
    "\n",
    "* It was noticed that the downloaded excel file (and the SPSS file as well) **has different values than the one in the [codebook](https://www.thearda.com/Archive/Files/Codebooks/BRS2014_CB.asp) of the dataset**. This might be due to the weights taken into consideration. Our analysis is\n",
    "* This might be a problem on other questions as well but it is not something we can control."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "7c8b5ede",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.0    859\n",
       "1.0    309\n",
       "3.0    254\n",
       "4.0    132\n",
       "Name: I_MARITAL, dtype: int64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "q2[\"MARRIED\"] = np.where(data.I_MARITAL == 2, 1, np.where(data.I_MARITAL.isna(), np.NaN, 0))\n",
    "q2[\"NOT_PARTNERED\"] = np.where((data.I_MARITAL == 1) | \n",
    "                               (data.I_MARITAL == 3) | \n",
    "                               (data.I_MARITAL == 4), 1, np.where(data.I_MARITAL.isna(), np.NaN, 0))\n",
    "data.I_MARITAL.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "59e87dcc",
   "metadata": {},
   "source": [
    "* Having children under 18 in home was a bit tricky to map as two variables were used (Q93 & Q93_NONE).\n",
    "\n",
    "* There were cases where only one of the two answers was NaN and cases were both answers were missing.\n",
    "\n",
    "* The approach was to only take into account answers where at least one of the answers was not missing.\n",
    " * If children were more than 0 in Q93, we were sure that the respondent had a child in home. In any other case Q93 can either be 0 or NaN. If either Q93 or Q93_NONE are not NaN we are sure that the respondent does not have children.\n",
    " * If Q93_NONE has a value, it means that the respondent does not have a child.\n",
    " * If both Q93 and Q93_NONE are NONE, we are sure that the value of both mapped variables will be NaN."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "0b85b0b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "q2[\"Both_child_na\"] = np.where((data.Q93.isna()) & (data.Q93_NONE.isna()), 1, 0)\n",
    "q2[\"CHILDREN\"] = np.where(data.Q93 > 0, 1, np.where(q2.Both_child_na == 1, np.NaN, 0))\n",
    "q2[\"NO_CHILDREN\"] = np.where((data.Q93_NONE == 1) | \n",
    "                             (data.Q93 == 0), 1, \n",
    "                             np.where(q2.Both_child_na == 1, np.NaN, 0))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "149bdcc7",
   "metadata": {},
   "source": [
    "* Mapping of urban and rural areas was straightforward:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "07a8aa7d",
   "metadata": {},
   "outputs": [],
   "source": [
    "q2[\"URBAN\"] = np.where((data.Q80 == 1)|(data.Q80 == 2),\n",
    "                1, np.where(data.Q80.isna(), \n",
    "                    np.NaN, 0))\n",
    "q2[\"RURAL\"] = np.where((data.Q80 == 3)|(data.Q80 == 4),\n",
    "                1, np.where(data.Q80.isna(), \n",
    "                    np.NaN, 0))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "babe9d95",
   "metadata": {},
   "source": [
    "* To determine whether a respondent resides in a southern region, the **STATE** variable should be analyzed. The south states were selected according to this [Wikipedia article](https://en.wikipedia.org/wiki/Southern_United_States) about the **Southern United States**.\n",
    "* The states belonging to the american south are the ones below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "299989c8",
   "metadata": {},
   "outputs": [],
   "source": [
    "south_states = {\"MD\",\"DE\",\"VA\",\"WV\",\"KY\",\n",
    "                \"TN\",\"NC\",\"SC\",\"FL\",\"GA\",\n",
    "                \"AL\",\"MS\",\"LA\",\"AK\",\"TX\",\"OK\"}"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "82871b72",
   "metadata": {},
   "source": [
    "* We then procceed with the mapping of southern regions:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "44d2769d",
   "metadata": {},
   "outputs": [],
   "source": [
    "q2[\"SOUTH\"] = np.where(data.STATE.isin(south_states), 1, \n",
    "                np.where(data.STATE.isna(), \n",
    "                   np.NaN, 0))\n",
    "q2[\"OTHER_REGION\"] = np.where(~data.STATE.isin(south_states), 1, \n",
    "                        np.where(data.STATE.isna(), \n",
    "                           np.NaN, 0))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "013b6b5a",
   "metadata": {},
   "source": [
    "* Finally, the race of the respondents must be coded to different dummy variables.\n",
    "\n",
    "* For the coding, Q88A-F & Q89 were examined.\n",
    "\n",
    "* Mapping was tricky as a separate variable was used for respondents of hispanic origin.\n",
    "\n",
    "* In addition, some answers to Q88 had only one race set to 1 and the other ones as NaN, while others, had the other races set to 2 (negative).\n",
    "\n",
    "* There were also cases where no race was not NaN but set to 2 (negative) in some occations (this specific case is considered to be \"Other Race\", more detailed description below)\n",
    "\n",
    "* We must first fill in some NaNs that should not be missing. If at least one race is selected in columns Q88A to Q88F, then all other answers can be set to 0 if not already set. \n",
    "\n",
    "\n",
    "* The issue described above can be seen in row 1 below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "d8bdf169",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Q88A</th>\n",
       "      <th>Q88B</th>\n",
       "      <th>Q88C</th>\n",
       "      <th>Q88D</th>\n",
       "      <th>Q88E</th>\n",
       "      <th>Q88F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Q88A  Q88B  Q88C  Q88D  Q88E  Q88F\n",
       "0   NaN   NaN   NaN   NaN   NaN   NaN\n",
       "1   2.0   1.0   2.0   2.0   2.0   NaN\n",
       "2   1.0   2.0   2.0   2.0   2.0   2.0\n",
       "3   1.0   2.0   2.0   2.0   2.0   2.0\n",
       "4   1.0   2.0   2.0   2.0   2.0   2.0"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "race_qs = [\"Q88A\",'Q88B', 'Q88C', 'Q88D', 'Q88E', 'Q88F']\n",
    "data[race_qs].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39fd45cc",
   "metadata": {},
   "source": [
    "* This issue is fixed with the code below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "436d48fa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Q88A</th>\n",
       "      <th>Q88B</th>\n",
       "      <th>Q88C</th>\n",
       "      <th>Q88D</th>\n",
       "      <th>Q88E</th>\n",
       "      <th>Q88F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Q88A  Q88B  Q88C  Q88D  Q88E  Q88F\n",
       "0   NaN   NaN   NaN   NaN   NaN   NaN\n",
       "1   2.0   1.0   2.0   2.0   2.0   2.0\n",
       "2   1.0   2.0   2.0   2.0   2.0   2.0\n",
       "3   1.0   2.0   2.0   2.0   2.0   2.0\n",
       "4   1.0   2.0   2.0   2.0   2.0   2.0"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_races_na = (12 - data.fillna(2)[race_qs].sum(axis=1)) > 0\n",
    "\n",
    "race = pd.DataFrame(data[race_qs].copy().values)\n",
    "race.columns = race_qs\n",
    "\n",
    "for row in range(len(race.values)):\n",
    "    if all_races_na.loc[row]:\n",
    "        for column in range(6):\n",
    "            if math.isnan(race.loc[row][column]):\n",
    "                race.loc[row][column] = 2           \n",
    "for row in range(len(race.values)):\n",
    "    if all_races_na.loc[row]:\n",
    "        for column in range(6):\n",
    "            if math.isnan(race.loc[row][column]):\n",
    "                race.loc[row][column] = 2\n",
    "race.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4e811744",
   "metadata": {},
   "source": [
    "* The issue in row 1 has been fixed.\n",
    "\n",
    "\n",
    "* We then take into account Q89 as well. We create a series to spot the cases where neither of the questions described above were answered.\n",
    "\n",
    "* We then code each answer to the related dummy variable.\n",
    "* In case of NaN in any of the qustions, the if both Q88A-F and Q89 are NaN, then NaN is assigned to all the bummy vaiables, otherwise 0 (negative) is assigned."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "6bf3b113",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>WHITE</th>\n",
       "      <th>BLACK_AFRICAN_AMERICAN</th>\n",
       "      <th>HISPANIC</th>\n",
       "      <th>OTHER_RACE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   WHITE  BLACK_AFRICAN_AMERICAN  HISPANIC  OTHER_RACE\n",
       "0    0.0                     0.0       0.0         0.0\n",
       "1    0.0                     1.0       0.0         0.0\n",
       "2    1.0                     0.0       0.0         0.0\n",
       "3    1.0                     0.0       0.0         0.0\n",
       "4    1.0                     0.0       0.0         0.0"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Neither hispanic and other races are all NaN\n",
    "RaceNaN = race[race_qs].isna().all(axis=1) & data.Q89.isna()\n",
    "\n",
    "q2[\"OTHER_RACE\"] = np.where((race.Q88C == 1) |(race.Q88D == 1) |(race.Q88E == 1) |(race.Q88F == 1), 1, \n",
    "                       np.where(RaceNaN, np.NaN, 0))\n",
    "q2[\"BLACK_AFRICAN_AMERICAN\"] = np.where((race.Q88B == 1), 1, np.where(RaceNaN, np.NaN, 0))\n",
    "q2[\"WHITE\"] = np.where((race.Q88A == 1), 1, np.where(RaceNaN, np.NaN, 0))\n",
    "q2[\"HISPANIC\"] = np.where((data.Q89 > 1), 1, np.where(RaceNaN, np.NaN, 0))\n",
    "\n",
    "race_columns = [\"WHITE\", \"BLACK_AFRICAN_AMERICAN\", \"HISPANIC\", \"OTHER_RACE\"]\n",
    "q2[race_columns].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9cdb1389",
   "metadata": {},
   "source": [
    "* The coding process of the respondents' race is not finished yet. In some cases, we can notice that a respondent has selected more than one race (row 32 below):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "59559c18",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>WHITE</th>\n",
       "      <th>BLACK_AFRICAN_AMERICAN</th>\n",
       "      <th>HISPANIC</th>\n",
       "      <th>OTHER_RACE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    WHITE  BLACK_AFRICAN_AMERICAN  HISPANIC  OTHER_RACE\n",
       "31    1.0                     0.0       0.0         0.0\n",
       "32    1.0                     0.0       1.0         1.0\n",
       "33    1.0                     0.0       0.0         0.0"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "q2[race_columns].iloc[31:34]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c3d76365",
   "metadata": {},
   "source": [
    "* From the paper's descriptive statisctics we can cocnclude that the paper is based on the assumption that each respondent has only one race.\n",
    "\n",
    "* So if multiple races are positive, only one should be selected.\n",
    "\n",
    "* The order of precedence used is the following:\n",
    "\n",
    "\n",
    "    HISPANIC > OTHER_RACE > BLACK_AFRICAN_AMERICAN > WHITE\n",
    "\n",
    "\n",
    "* This means that if a respondent selected all the races, they will be classified as HISPANIC.\n",
    "\n",
    "\n",
    "* This is what we do below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "625bbaad",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>WHITE</th>\n",
       "      <th>BLACK_AFRICAN_AMERICAN</th>\n",
       "      <th>HISPANIC</th>\n",
       "      <th>OTHER_RACE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    WHITE  BLACK_AFRICAN_AMERICAN  HISPANIC  OTHER_RACE\n",
       "31    1.0                     0.0       0.0         0.0\n",
       "32    0.0                     0.0       1.0         0.0\n",
       "33    1.0                     0.0       0.0         0.0"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "q2[\"WHITE\"] = np.where(q2.WHITE != 1, \n",
    "                      q2.WHITE, \n",
    "                      q2.WHITE - np.where(q2[[\"BLACK_AFRICAN_AMERICAN\",\"OTHER_RACE\",\"HISPANIC\"]].sum(axis=1) > 0, 1, 0))\n",
    "q2[\"BLACK_AFRICAN_AMERICAN\"] = np.where(q2.BLACK_AFRICAN_AMERICAN != 1, \n",
    "                      q2.BLACK_AFRICAN_AMERICAN, \n",
    "                      q2.BLACK_AFRICAN_AMERICAN - np.where(q2[[\"OTHER_RACE\",\"HISPANIC\"]].sum(axis=1) > 0, 1, 0))\n",
    "q2[\"OTHER_RACE\"] = np.where(q2.OTHER_RACE != 1, \n",
    "                         q2.OTHER_RACE,\n",
    "                         q2.OTHER_RACE - q2.HISPANIC)\n",
    "q2[race_columns].iloc[31:34]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d1930946",
   "metadata": {},
   "source": [
    "* We are now left with the where someone answered as negative in every race. We assume that there cases will be classified as other race, as they have answered at least one race related question."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "b5d94dcc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>WHITE</th>\n",
       "      <th>BLACK_AFRICAN_AMERICAN</th>\n",
       "      <th>HISPANIC</th>\n",
       "      <th>OTHER_RACE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1567</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1568</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1569</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1570</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1571</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1572 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      WHITE  BLACK_AFRICAN_AMERICAN  HISPANIC  OTHER_RACE\n",
       "0       0.0                     0.0       0.0         1.0\n",
       "1       0.0                     1.0       0.0         0.0\n",
       "2       1.0                     0.0       0.0         0.0\n",
       "3       1.0                     0.0       0.0         0.0\n",
       "4       1.0                     0.0       0.0         0.0\n",
       "...     ...                     ...       ...         ...\n",
       "1567    0.0                     0.0       1.0         0.0\n",
       "1568    0.0                     0.0       1.0         0.0\n",
       "1569    0.0                     0.0       1.0         0.0\n",
       "1570    0.0                     0.0       1.0         0.0\n",
       "1571    0.0                     0.0       1.0         0.0\n",
       "\n",
       "[1572 rows x 4 columns]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "empty_race = np.where((q2[race_columns] == 0).all(axis=1), 1, np.where(RaceNaN, np.NaN, 0))\n",
    "\n",
    "q2[\"OTHER_RACE\"] = q2[\"OTHER_RACE\"] + empty_race\n",
    "q2[race_columns]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "033d5fc0",
   "metadata": {},
   "source": [
    "We are now going to specify the columns that contain our variables, in order to display their means:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "94c3f236",
   "metadata": {},
   "source": [
    "* We finally calculate the SUPERNATURAL_EVIL metric"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "4ba03aba",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       4.0\n",
       "1       4.0\n",
       "2       2.0\n",
       "3       4.0\n",
       "4       4.0\n",
       "       ... \n",
       "1567    3.0\n",
       "1568    3.0\n",
       "1569    4.0\n",
       "1570    3.0\n",
       "1571    NaN\n",
       "Name: SUPERNATURAL_EVIL, Length: 1572, dtype: float64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "q2[\"SUPERNATURAL_EVIL\"] = ((q2.DEVIL + q2.HELL + q2.DEMONS) /3).round()\n",
    "q2[\"SUPERNATURAL_EVIL\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "9398cc94",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_cols = [\"BAN_SEMI_AUTO\", \"BAN_AMMO_CLIPS\", \"BAN_HANDGUNS\", \"SUPPORT_CONCEALED_CARRY\", \"ARMED_SECURITY\",\"ARMED_TEACHERS\",\n",
    "    \"MORE_GUN_SAFETY\", \"EXPANDED_SCREENING\", \"SUPERNATURAL_EVIL\", \"ATTENDANCE\", \"BIBLICAL_LITERALISM\", \n",
    "    \"BIBLICAL_INERRACY\", \"BIBLE_HUMAN_ERROR\", \"CONSERVATIVE_PROTESTANT\", \"MAINLINE_PROTESTANT\", \"BLACK_PROTESTANT\", \"CATHOLIC\", \n",
    "    \"OTHER_RELIGION\", \"NO_AFFILIATION\", \"POLITICAL_IDEOLOGY\",\"AGE\",\n",
    "    \"MALE\", \"FEMALE\",\"WHITE\", \"HISPANIC\",\"BLACK_AFRICAN_AMERICAN\",\"OTHER_RACE\",\n",
    "    \"LESS_THAN_HS\",\"HIGH_SCHOOL_OR_EQUIVALENT\", \"SOME_COLLEGE\",\"COLLEGE_GRADUATE\",\n",
    "    \"POSTGRADUATE\",\"INCOME\",\"MARRIED\", \"NOT_PARTNERED\",\"CHILDREN\", \"NO_CHILDREN\" ,\"URBAN\", \"RURAL\",\n",
    "    \"SOUTH\", \"OTHER_REGION\", \"DEVIL\", \"HELL\", \"DEMONS\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cadc0e53",
   "metadata": {},
   "source": [
    "* We then proceed to calculate the means of every column before imputation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "8b419ea7",
   "metadata": {},
   "outputs": [],
   "source": [
    "means = pd.DataFrame(q2[all_cols].mean().transpose())\n",
    "means.columns = [\"Mean Before Imputation\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7cf95666",
   "metadata": {},
   "source": [
    "* We then proceed with the imputation.\n",
    "* For the imputation we will be using scikit learn's IterativeImputer, which is the equivalent of the multiple imputation used in the paper. We initialize and transform our data.\n",
    "\n",
    "\n",
    "* Note that we will be rounding our results as the iterative imputation returns float values.\n",
    "\n",
    "\n",
    "* There is a risk of imputing two or more dummy variables as 1. This in practice is extremely rare and will therefore not be concerning us. The iterative imputer takes into account the relationships between the columns and therefore is is more difficult for such collisions to occure.\n",
    "* In addition, as described by [Johnson & Young 2011](https://www.sciencedirect.com/science/article/pii/S0049089X21000727?via=ihub#bib32) iterative imputation can be used effectively for imputing dummy variables."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "aaaddd6f",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Library/Frameworks/Python.framework/Versions/3.10/lib/python3.10/site-packages/sklearn/impute/_iterative.py:699: ConvergenceWarning: [IterativeImputer] Early stopping criterion not reached.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>BAN_SEMI_AUTO</th>\n",
       "      <th>BAN_AMMO_CLIPS</th>\n",
       "      <th>BAN_HANDGUNS</th>\n",
       "      <th>SUPPORT_CONCEALED_CARRY</th>\n",
       "      <th>ARMED_SECURITY</th>\n",
       "      <th>ARMED_TEACHERS</th>\n",
       "      <th>MORE_GUN_SAFETY</th>\n",
       "      <th>EXPANDED_SCREENING</th>\n",
       "      <th>SUPERNATURAL_EVIL</th>\n",
       "      <th>ATTENDANCE</th>\n",
       "      <th>...</th>\n",
       "      <th>NOT_PARTNERED</th>\n",
       "      <th>CHILDREN</th>\n",
       "      <th>NO_CHILDREN</th>\n",
       "      <th>URBAN</th>\n",
       "      <th>RURAL</th>\n",
       "      <th>SOUTH</th>\n",
       "      <th>OTHER_REGION</th>\n",
       "      <th>DEVIL</th>\n",
       "      <th>HELL</th>\n",
       "      <th>DEMONS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1567</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1568</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1569</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1570</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1571</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1572 rows × 44 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      BAN_SEMI_AUTO  BAN_AMMO_CLIPS  BAN_HANDGUNS  SUPPORT_CONCEALED_CARRY  \\\n",
       "0               0.0             0.0           0.0                      1.0   \n",
       "1               1.0             1.0           0.0                      1.0   \n",
       "2               1.0             1.0           0.0                      0.0   \n",
       "3               1.0             1.0           1.0                      0.0   \n",
       "4               1.0             1.0           1.0                      0.0   \n",
       "...             ...             ...           ...                      ...   \n",
       "1567            1.0             1.0           1.0                      0.0   \n",
       "1568            1.0             0.0           1.0                      0.0   \n",
       "1569            1.0             1.0           1.0                      0.0   \n",
       "1570            1.0             1.0           1.0                      0.0   \n",
       "1571            1.0             1.0           1.0                      0.0   \n",
       "\n",
       "      ARMED_SECURITY  ARMED_TEACHERS  MORE_GUN_SAFETY  EXPANDED_SCREENING  \\\n",
       "0                1.0             1.0              1.0                 1.0   \n",
       "1                1.0             0.0              1.0                 1.0   \n",
       "2                1.0             0.0              1.0                 1.0   \n",
       "3                1.0             0.0              1.0                 1.0   \n",
       "4                1.0             1.0              1.0                 1.0   \n",
       "...              ...             ...              ...                 ...   \n",
       "1567             1.0             0.0              1.0                 1.0   \n",
       "1568             0.0             0.0              1.0                 1.0   \n",
       "1569             0.0             0.0              1.0                 1.0   \n",
       "1570             0.0             0.0              1.0                 1.0   \n",
       "1571             1.0             0.0              1.0                 1.0   \n",
       "\n",
       "      SUPERNATURAL_EVIL  ATTENDANCE  ...  NOT_PARTNERED  CHILDREN  \\\n",
       "0                   4.0         8.0  ...            1.0       0.0   \n",
       "1                   4.0         1.0  ...            1.0       1.0   \n",
       "2                   2.0         7.0  ...            1.0       0.0   \n",
       "3                   4.0         6.0  ...            1.0       0.0   \n",
       "4                   4.0         8.0  ...            0.0       1.0   \n",
       "...                 ...         ...  ...            ...       ...   \n",
       "1567                3.0         1.0  ...            0.0       0.0   \n",
       "1568                3.0         0.0  ...            0.0       1.0   \n",
       "1569                4.0         7.0  ...            1.0       0.0   \n",
       "1570                3.0         3.0  ...            0.0       0.0   \n",
       "1571                3.0         5.0  ...            0.0       0.0   \n",
       "\n",
       "      NO_CHILDREN  URBAN  RURAL  SOUTH  OTHER_REGION  DEVIL  HELL  DEMONS  \n",
       "0             1.0    0.0    1.0    1.0           0.0    4.0   4.0     4.0  \n",
       "1             0.0    1.0    0.0    1.0           0.0    4.0   4.0     4.0  \n",
       "2             1.0    0.0    1.0    0.0           1.0    3.0   2.0     2.0  \n",
       "3             1.0    1.0    0.0    1.0           0.0    4.0   4.0     4.0  \n",
       "4             0.0    1.0    0.0    0.0           1.0    4.0   4.0     4.0  \n",
       "...           ...    ...    ...    ...           ...    ...   ...     ...  \n",
       "1567          1.0    0.0    1.0    0.0           1.0    4.0   2.0     4.0  \n",
       "1568          0.0    1.0    0.0    0.0           1.0    3.0   3.0     3.0  \n",
       "1569          1.0    0.0    1.0    0.0           1.0    4.0   4.0     3.0  \n",
       "1570          1.0    1.0    0.0    0.0           1.0    3.0   3.0     3.0  \n",
       "1571          1.0    1.0    0.0    1.0           0.0    3.0   3.0     3.0  \n",
       "\n",
       "[1572 rows x 44 columns]"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.experimental import enable_iterative_imputer\n",
    "from sklearn.impute import IterativeImputer\n",
    "\n",
    "imp = IterativeImputer(random_state=0)\n",
    "q2[all_cols] = imp.fit_transform(q2[all_cols])\n",
    "q2[all_cols] = q2[all_cols].round()\n",
    "q2[all_cols]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cabdc60b",
   "metadata": {},
   "source": [
    "* We calculate the means after imputation and print them below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "a74ec55a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Mean Before Imputation</th>\n",
       "      <th>Mean After Imputation</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>BAN_SEMI_AUTO</th>\n",
       "      <td>0.629161</td>\n",
       "      <td>0.637405</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BAN_AMMO_CLIPS</th>\n",
       "      <td>0.671105</td>\n",
       "      <td>0.680025</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BAN_HANDGUNS</th>\n",
       "      <td>0.290473</td>\n",
       "      <td>0.288168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SUPPORT_CONCEALED_CARRY</th>\n",
       "      <td>0.557923</td>\n",
       "      <td>0.557888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ARMED_SECURITY</th>\n",
       "      <td>0.621854</td>\n",
       "      <td>0.632316</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ARMED_TEACHERS</th>\n",
       "      <td>0.294548</td>\n",
       "      <td>0.287532</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MORE_GUN_SAFETY</th>\n",
       "      <td>0.927393</td>\n",
       "      <td>0.930025</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EXPANDED_SCREENING</th>\n",
       "      <td>0.908069</td>\n",
       "      <td>0.911578</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SUPERNATURAL_EVIL</th>\n",
       "      <td>3.090909</td>\n",
       "      <td>3.091603</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ATTENDANCE</th>\n",
       "      <td>3.676735</td>\n",
       "      <td>3.674936</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BIBLICAL_LITERALISM</th>\n",
       "      <td>0.229180</td>\n",
       "      <td>0.225827</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BIBLICAL_INERRACY</th>\n",
       "      <td>0.338283</td>\n",
       "      <td>0.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BIBLE_HUMAN_ERROR</th>\n",
       "      <td>0.432537</td>\n",
       "      <td>0.426209</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CONSERVATIVE_PROTESTANT</th>\n",
       "      <td>0.284267</td>\n",
       "      <td>0.267812</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MAINLINE_PROTESTANT</th>\n",
       "      <td>0.168805</td>\n",
       "      <td>0.159033</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BLACK_PROTESTANT</th>\n",
       "      <td>0.071573</td>\n",
       "      <td>0.067430</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CATHOLIC</th>\n",
       "      <td>0.249156</td>\n",
       "      <td>0.234733</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OTHER_RELIGION</th>\n",
       "      <td>0.081702</td>\n",
       "      <td>0.076972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NO_AFFILIATION</th>\n",
       "      <td>0.144497</td>\n",
       "      <td>0.136132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>POLITICAL_IDEOLOGY</th>\n",
       "      <td>3.800789</td>\n",
       "      <td>3.798982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AGE</th>\n",
       "      <td>52.645095</td>\n",
       "      <td>52.697837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MALE</th>\n",
       "      <td>0.421224</td>\n",
       "      <td>0.411578</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FEMALE</th>\n",
       "      <td>0.578776</td>\n",
       "      <td>0.588422</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>WHITE</th>\n",
       "      <td>0.705017</td>\n",
       "      <td>0.719466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HISPANIC</th>\n",
       "      <td>0.128428</td>\n",
       "      <td>0.122137</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BLACK_AFRICAN_AMERICAN</th>\n",
       "      <td>0.113043</td>\n",
       "      <td>0.107506</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OTHER_RACE</th>\n",
       "      <td>0.053512</td>\n",
       "      <td>0.050891</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LESS_THAN_HS</th>\n",
       "      <td>0.059294</td>\n",
       "      <td>0.056616</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HIGH_SCHOOL_OR_EQUIVALENT</th>\n",
       "      <td>0.133911</td>\n",
       "      <td>0.127863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOME_COLLEGE</th>\n",
       "      <td>0.321785</td>\n",
       "      <td>0.307252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>COLLEGE_GRADUATE</th>\n",
       "      <td>0.261825</td>\n",
       "      <td>0.250000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>POSTGRADUATE</th>\n",
       "      <td>0.223185</td>\n",
       "      <td>0.213104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>INCOME</th>\n",
       "      <td>4.151329</td>\n",
       "      <td>4.147583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MARRIED</th>\n",
       "      <td>0.552767</td>\n",
       "      <td>0.557888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NOT_PARTNERED</th>\n",
       "      <td>0.447233</td>\n",
       "      <td>0.442112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CHILDREN</th>\n",
       "      <td>0.343348</td>\n",
       "      <td>0.254453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NO_CHILDREN</th>\n",
       "      <td>0.657511</td>\n",
       "      <td>0.746183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>URBAN</th>\n",
       "      <td>0.520779</td>\n",
       "      <td>0.524173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RURAL</th>\n",
       "      <td>0.466234</td>\n",
       "      <td>0.461196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOUTH</th>\n",
       "      <td>0.377863</td>\n",
       "      <td>0.377863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OTHER_REGION</th>\n",
       "      <td>0.622137</td>\n",
       "      <td>0.622137</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DEVIL</th>\n",
       "      <td>3.149601</td>\n",
       "      <td>3.139949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HELL</th>\n",
       "      <td>3.088726</td>\n",
       "      <td>3.084606</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DEMONS</th>\n",
       "      <td>3.016086</td>\n",
       "      <td>3.022901</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           Mean Before Imputation  Mean After Imputation\n",
       "BAN_SEMI_AUTO                            0.629161               0.637405\n",
       "BAN_AMMO_CLIPS                           0.671105               0.680025\n",
       "BAN_HANDGUNS                             0.290473               0.288168\n",
       "SUPPORT_CONCEALED_CARRY                  0.557923               0.557888\n",
       "ARMED_SECURITY                           0.621854               0.632316\n",
       "ARMED_TEACHERS                           0.294548               0.287532\n",
       "MORE_GUN_SAFETY                          0.927393               0.930025\n",
       "EXPANDED_SCREENING                       0.908069               0.911578\n",
       "SUPERNATURAL_EVIL                        3.090909               3.091603\n",
       "ATTENDANCE                               3.676735               3.674936\n",
       "BIBLICAL_LITERALISM                      0.229180               0.225827\n",
       "BIBLICAL_INERRACY                        0.338283               0.333333\n",
       "BIBLE_HUMAN_ERROR                        0.432537               0.426209\n",
       "CONSERVATIVE_PROTESTANT                  0.284267               0.267812\n",
       "MAINLINE_PROTESTANT                      0.168805               0.159033\n",
       "BLACK_PROTESTANT                         0.071573               0.067430\n",
       "CATHOLIC                                 0.249156               0.234733\n",
       "OTHER_RELIGION                           0.081702               0.076972\n",
       "NO_AFFILIATION                           0.144497               0.136132\n",
       "POLITICAL_IDEOLOGY                       3.800789               3.798982\n",
       "AGE                                     52.645095              52.697837\n",
       "MALE                                     0.421224               0.411578\n",
       "FEMALE                                   0.578776               0.588422\n",
       "WHITE                                    0.705017               0.719466\n",
       "HISPANIC                                 0.128428               0.122137\n",
       "BLACK_AFRICAN_AMERICAN                   0.113043               0.107506\n",
       "OTHER_RACE                               0.053512               0.050891\n",
       "LESS_THAN_HS                             0.059294               0.056616\n",
       "HIGH_SCHOOL_OR_EQUIVALENT                0.133911               0.127863\n",
       "SOME_COLLEGE                             0.321785               0.307252\n",
       "COLLEGE_GRADUATE                         0.261825               0.250000\n",
       "POSTGRADUATE                             0.223185               0.213104\n",
       "INCOME                                   4.151329               4.147583\n",
       "MARRIED                                  0.552767               0.557888\n",
       "NOT_PARTNERED                            0.447233               0.442112\n",
       "CHILDREN                                 0.343348               0.254453\n",
       "NO_CHILDREN                              0.657511               0.746183\n",
       "URBAN                                    0.520779               0.524173\n",
       "RURAL                                    0.466234               0.461196\n",
       "SOUTH                                    0.377863               0.377863\n",
       "OTHER_REGION                             0.622137               0.622137\n",
       "DEVIL                                    3.149601               3.139949\n",
       "HELL                                     3.088726               3.084606\n",
       "DEMONS                                   3.016086               3.022901"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "means[\"Mean After Imputation\"] = pd.DataFrame(q2[all_cols].mean().transpose())\n",
    "means"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5a9d17e3",
   "metadata": {},
   "source": [
    "We can see that all the results are almost identical with the paper's descriptive statistics, except for the children variable. However, the error is less almost 3 percentage units and which is not that large and therefore we can proceed to the next questions.\n",
    "\n",
    "\n",
    "### Q3: Predict Support for Various Gun Policies\n",
    "\n",
    "* Firstly, we create come column groups to aid us create the statsmodels regression formulae."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "07456f71",
   "metadata": {},
   "outputs": [],
   "source": [
    "focal_var_col = [\"SUPERNATURAL_EVIL\"]\n",
    "\n",
    "dependent_variable_cols = [\"BAN_SEMI_AUTO\",\"BAN_AMMO_CLIPS\" ,\"BAN_HANDGUNS\",\n",
    "\"SUPPORT_CONCEALED_CARRY\",\"ARMED_SECURITY\",\"ARMED_TEACHERS\",\n",
    "\"MORE_GUN_SAFETY\",\"EXPANDED_SCREENING\"]\n",
    "\n",
    "mediator_col = [\"POLITICAL_IDEOLOGY\"]\n",
    "\n",
    "religious_variable_cols = [\"ATTENDANCE\", \"BIBLICAL_INERRACY\", \"BIBLICAL_LITERALISM\", \n",
    "                             \"MAINLINE_PROTESTANT\", \"BLACK_PROTESTANT\", \"CONSERVATIVE_PROTESTANT\",  \n",
    "                             \"CATHOLIC\", \"OTHER_RELIGION\"]\n",
    "\n",
    "control_cols = [\"BLACK_AFRICAN_AMERICAN\", \"HISPANIC\", \"OTHER_RACE\", \"FEMALE\",\n",
    "                \"LESS_THAN_HS\", \"SOME_COLLEGE\",\"COLLEGE_GRADUATE\", \"POSTGRADUATE\",\n",
    "                \"AGE\", \"INCOME\", \"MARRIED\", \"CHILDREN\", \"URBAN\", \"SOUTH\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0ee69c67",
   "metadata": {},
   "source": [
    "* We then proceed with the regression.\n",
    "\n",
    "\n",
    "* Two models were run:\n",
    " * One containing only the focal variable and the religious variables (**model_a**)\n",
    " * One containing all the variables (focal + religious + mediator + control) (**model_b**)\n",
    "\n",
    "\n",
    "* Because of the dummy variables, a logistic regression model was run instead of a linear regression model. This is done with the **logit** function.\n",
    "\n",
    "\n",
    "* The logistic regression coefficients are difficult to interpret, therefore in the paper the **Odds Ratio** (**OR**) was calculated. This is done by getting the exponentiating the logistic regression coefficients.\n",
    "\n",
    "* The Odds Ratio describes the single unit increment association between the measured variable and the odds of the dependent variable being 1 or 0. The farther the OR from 1 the larger the magnitude of the association.\n",
    "\n",
    "* For instance for the first two regression models for BAN_SEMI_AUTO, or for supernatural evil is $0,76, p<0.001$ (compared to the paper's $0,73, p<0,001$). This means that an increase of a single unit in one's belief about supernatural evil, reduces the chance of one supporting the ban in semi auto guns by $|1-0,76| = 0,24\\%$. This effect is reduced to $|1-0,86| = 14\\%$ ($OR=0,86$) when the mediator variable of political ideology and the control variables are taken into consideration.\n",
    "\n",
    "\n",
    "* Below 8 tables containing the Odds Ratios and the significance (p-value) for each dependent variable for logistic regression models (a) and (b) can be found."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "0f16d329",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=====================================================================\n",
      "                           Logistic Regression                       \n",
      "=====================================================================\n",
      "\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  BAN_SEMI_AUTO\n",
      "---------------------------------------------------------------------\n",
      "                         OR_a  pval_a sig_a  OR_b  pval_b sig_b\n",
      "Intercept                3.06   0.000   ***  0.02   0.000   ***\n",
      "SUPERNATURAL_EVIL        0.76   0.000   ***  0.86   0.066      \n",
      "ATTENDANCE               1.04   0.089        1.06   0.038     *\n",
      "BIBLICAL_INERRACY        0.92   0.598        0.98   0.897      \n",
      "BIBLICAL_LITERALISM      0.94   0.738        1.11   0.602      \n",
      "MAINLINE_PROTESTANT      1.20   0.366        1.20   0.431      \n",
      "BLACK_PROTESTANT         1.80   0.035     *  0.85   0.686      \n",
      "CONSERVATIVE_PROTESTANT  0.99   0.973        1.43   0.136      \n",
      "CATHOLIC                 1.35   0.110        1.44   0.096      \n",
      "OTHER_RELIGION           1.95   0.010     *  1.64   0.091      \n",
      "POLITICAL_IDEOLOGY        NaN     NaN        1.75   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN    NaN     NaN        1.85   0.049     *\n",
      "HISPANIC                  NaN     NaN        1.59   0.022     *\n",
      "OTHER_RACE                NaN     NaN        0.91   0.707      \n",
      "FEMALE                    NaN     NaN        3.20   0.000   ***\n",
      "LESS_THAN_HS              NaN     NaN        0.62   0.096      \n",
      "SOME_COLLEGE              NaN     NaN        0.82   0.288      \n",
      "COLLEGE_GRADUATE          NaN     NaN        0.67   0.049     *\n",
      "POSTGRADUATE              NaN     NaN        1.23   0.349      \n",
      "AGE                       NaN     NaN        1.03   0.000   ***\n",
      "INCOME                    NaN     NaN        1.12   0.017     *\n",
      "MARRIED                   NaN     NaN        0.74   0.036     *\n",
      "CHILDREN                  NaN     NaN        1.09   0.591      \n",
      "URBAN                     NaN     NaN        1.35   0.015     *\n",
      "SOUTH                     NaN     NaN        0.71   0.007    **\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  BAN_AMMO_CLIPS\n",
      "---------------------------------------------------------------------\n",
      "                         OR_a  pval_a sig_a  OR_b  pval_b sig_b\n",
      "Intercept                3.58   0.000   ***  0.04   0.000   ***\n",
      "SUPERNATURAL_EVIL        0.78   0.001    **  0.86   0.070      \n",
      "ATTENDANCE               1.05   0.072        1.06   0.043     *\n",
      "BIBLICAL_INERRACY        0.93   0.631        0.99   0.951      \n",
      "BIBLICAL_LITERALISM      0.85   0.372        0.89   0.566      \n",
      "MAINLINE_PROTESTANT      1.20   0.366        1.15   0.533      \n",
      "BLACK_PROTESTANT         1.99   0.018     *  1.15   0.727      \n",
      "CONSERVATIVE_PROTESTANT  1.02   0.930        1.39   0.164      \n",
      "CATHOLIC                 1.25   0.254        1.34   0.178      \n",
      "OTHER_RELIGION           1.84   0.023     *  1.57   0.134      \n",
      "POLITICAL_IDEOLOGY        NaN     NaN        1.64   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN    NaN     NaN        1.26   0.452      \n",
      "HISPANIC                  NaN     NaN        1.14   0.518      \n",
      "OTHER_RACE                NaN     NaN        0.98   0.951      \n",
      "FEMALE                    NaN     NaN        2.88   0.000   ***\n",
      "LESS_THAN_HS              NaN     NaN        0.93   0.805      \n",
      "SOME_COLLEGE              NaN     NaN        0.80   0.229      \n",
      "COLLEGE_GRADUATE          NaN     NaN        0.73   0.127      \n",
      "POSTGRADUATE              NaN     NaN        1.21   0.395      \n",
      "AGE                       NaN     NaN        1.03   0.000   ***\n",
      "INCOME                    NaN     NaN        1.00   0.980      \n",
      "MARRIED                   NaN     NaN        0.96   0.750      \n",
      "CHILDREN                  NaN     NaN        0.89   0.450      \n",
      "URBAN                     NaN     NaN        1.40   0.007    **\n",
      "SOUTH                     NaN     NaN        0.97   0.821      \n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  BAN_HANDGUNS\n",
      "---------------------------------------------------------------------\n",
      "                         OR_a  pval_a sig_a  OR_b  pval_b sig_b\n",
      "Intercept                0.94   0.731        0.08   0.000   ***\n",
      "SUPERNATURAL_EVIL        0.65   0.000   ***  0.69   0.000   ***\n",
      "ATTENDANCE               1.04   0.109        1.06   0.039     *\n",
      "BIBLICAL_INERRACY        0.97   0.854        1.01   0.934      \n",
      "BIBLICAL_LITERALISM      1.38   0.107        1.47   0.073      \n",
      "MAINLINE_PROTESTANT      0.87   0.543        0.98   0.931      \n",
      "BLACK_PROTESTANT         3.27   0.000   ***  2.06   0.047     *\n",
      "CONSERVATIVE_PROTESTANT  0.89   0.634        1.07   0.776      \n",
      "CATHOLIC                 1.72   0.007    **  1.50   0.063      \n",
      "OTHER_RELIGION           2.12   0.002    **  1.89   0.013     *\n",
      "POLITICAL_IDEOLOGY        NaN     NaN        1.39   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN    NaN     NaN        1.68   0.056      \n",
      "HISPANIC                  NaN     NaN        2.72   0.000   ***\n",
      "OTHER_RACE                NaN     NaN        2.65   0.000   ***\n",
      "FEMALE                    NaN     NaN        1.36   0.018     *\n",
      "LESS_THAN_HS              NaN     NaN        1.43   0.202      \n",
      "SOME_COLLEGE              NaN     NaN        0.66   0.030     *\n",
      "COLLEGE_GRADUATE          NaN     NaN        0.87   0.496      \n",
      "POSTGRADUATE              NaN     NaN        1.10   0.650      \n",
      "AGE                       NaN     NaN        1.01   0.001    **\n",
      "INCOME                    NaN     NaN        0.93   0.154      \n",
      "MARRIED                   NaN     NaN        1.07   0.648      \n",
      "CHILDREN                  NaN     NaN        1.22   0.209      \n",
      "URBAN                     NaN     NaN        1.35   0.021     *\n",
      "SOUTH                     NaN     NaN        0.76   0.037     *\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  SUPPORT_CONCEALED_CARRY\n",
      "---------------------------------------------------------------------\n",
      "                         OR_a  pval_a sig_a   OR_b  pval_b sig_b\n",
      "Intercept                0.50   0.000   ***  13.00   0.000   ***\n",
      "SUPERNATURAL_EVIL        1.54   0.000   ***   1.41   0.000   ***\n",
      "ATTENDANCE               0.99   0.621         0.97   0.312      \n",
      "BIBLICAL_INERRACY        0.74   0.040     *   0.63   0.004    **\n",
      "BIBLICAL_LITERALISM      0.93   0.675         0.75   0.142      \n",
      "MAINLINE_PROTESTANT      0.98   0.931         0.97   0.878      \n",
      "BLACK_PROTESTANT         0.41   0.001    **   0.49   0.043     *\n",
      "CONSERVATIVE_PROTESTANT  1.00   0.987         0.80   0.327      \n",
      "CATHOLIC                 0.61   0.008    **   0.61   0.014     *\n",
      "OTHER_RELIGION           0.53   0.006    **   0.62   0.062      \n",
      "POLITICAL_IDEOLOGY        NaN     NaN         0.66   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN    NaN     NaN         0.98   0.950      \n",
      "HISPANIC                  NaN     NaN         0.75   0.123      \n",
      "OTHER_RACE                NaN     NaN         0.51   0.009    **\n",
      "FEMALE                    NaN     NaN         0.47   0.000   ***\n",
      "LESS_THAN_HS              NaN     NaN         0.63   0.089      \n",
      "SOME_COLLEGE              NaN     NaN         1.33   0.098      \n",
      "COLLEGE_GRADUATE          NaN     NaN         1.31   0.157      \n",
      "POSTGRADUATE              NaN     NaN         0.78   0.218      \n",
      "AGE                       NaN     NaN         0.99   0.009    **\n",
      "INCOME                    NaN     NaN         0.94   0.129      \n",
      "MARRIED                   NaN     NaN         1.28   0.057      \n",
      "CHILDREN                  NaN     NaN         0.78   0.089      \n",
      "URBAN                     NaN     NaN         0.76   0.018     *\n",
      "SOUTH                     NaN     NaN         1.44   0.003    **\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  ARMED_SECURITY\n",
      "---------------------------------------------------------------------\n",
      "                         OR_a  pval_a sig_a  OR_b  pval_b sig_b\n",
      "Intercept                0.37   0.000   ***  6.21   0.000   ***\n",
      "SUPERNATURAL_EVIL        1.61   0.000   ***  1.37   0.000   ***\n",
      "ATTENDANCE               0.92   0.002    **  0.92   0.004    **\n",
      "BIBLICAL_INERRACY        1.13   0.429        1.04   0.795      \n",
      "BIBLICAL_LITERALISM      1.96   0.001    **  1.49   0.057      \n",
      "MAINLINE_PROTESTANT      1.33   0.154        1.50   0.056      \n",
      "BLACK_PROTESTANT         1.77   0.056        1.68   0.181      \n",
      "CONSERVATIVE_PROTESTANT  1.34   0.174        1.22   0.369      \n",
      "CATHOLIC                 1.23   0.274        1.32   0.172      \n",
      "OTHER_RELIGION           1.26   0.328        1.68   0.038     *\n",
      "POLITICAL_IDEOLOGY        NaN     NaN        0.79   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN    NaN     NaN        1.26   0.413      \n",
      "HISPANIC                  NaN     NaN        1.11   0.585      \n",
      "OTHER_RACE                NaN     NaN        0.78   0.342      \n",
      "FEMALE                    NaN     NaN        0.97   0.816      \n",
      "LESS_THAN_HS              NaN     NaN        0.55   0.044     *\n",
      "SOME_COLLEGE              NaN     NaN        0.64   0.018     *\n",
      "COLLEGE_GRADUATE          NaN     NaN        0.63   0.022     *\n",
      "POSTGRADUATE              NaN     NaN        0.41   0.000   ***\n",
      "AGE                       NaN     NaN        0.99   0.033     *\n",
      "INCOME                    NaN     NaN        0.85   0.000   ***\n",
      "MARRIED                   NaN     NaN        1.31   0.040     *\n",
      "CHILDREN                  NaN     NaN        0.97   0.835      \n",
      "URBAN                     NaN     NaN        0.84   0.134      \n",
      "SOUTH                     NaN     NaN        1.46   0.002    **\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------------------------------------------------------------------\n",
      "Independent Variable:  ARMED_TEACHERS\n",
      "---------------------------------------------------------------------\n",
      "                         OR_a  pval_a sig_a  OR_b  pval_b sig_b\n",
      "Intercept                0.18   0.000   ***  6.09   0.000   ***\n",
      "SUPERNATURAL_EVIL        1.36   0.000   ***  1.24   0.014     *\n",
      "ATTENDANCE               0.95   0.058        0.93   0.012     *\n",
      "BIBLICAL_INERRACY        0.97   0.827        0.91   0.581      \n",
      "BIBLICAL_LITERALISM      0.98   0.901        0.84   0.416      \n",
      "MAINLINE_PROTESTANT      1.00   0.996        0.96   0.866      \n",
      "BLACK_PROTESTANT         0.56   0.060        1.28   0.561      \n",
      "CONSERVATIVE_PROTESTANT  1.49   0.071        1.22   0.395      \n",
      "CATHOLIC                 0.83   0.357        0.76   0.214      \n",
      "OTHER_RELIGION           0.76   0.321        0.88   0.663      \n",
      "POLITICAL_IDEOLOGY        NaN     NaN        0.66   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN    NaN     NaN        0.54   0.059      \n",
      "HISPANIC                  NaN     NaN        0.74   0.144      \n",
      "OTHER_RACE                NaN     NaN        0.94   0.818      \n",
      "FEMALE                    NaN     NaN        0.48   0.000   ***\n",
      "LESS_THAN_HS              NaN     NaN        0.86   0.601      \n",
      "SOME_COLLEGE              NaN     NaN        0.96   0.829      \n",
      "COLLEGE_GRADUATE          NaN     NaN        0.97   0.893      \n",
      "POSTGRADUATE              NaN     NaN        0.67   0.078      \n",
      "AGE                       NaN     NaN        0.99   0.005    **\n",
      "INCOME                    NaN     NaN        0.96   0.447      \n",
      "MARRIED                   NaN     NaN        1.16   0.281      \n",
      "CHILDREN                  NaN     NaN        1.21   0.218      \n",
      "URBAN                     NaN     NaN        0.66   0.001    **\n",
      "SOUTH                     NaN     NaN        0.84   0.176      \n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  MORE_GUN_SAFETY\n",
      "---------------------------------------------------------------------\n",
      "                          OR_a  pval_a sig_a  OR_b  pval_b sig_b\n",
      "Intercept                18.41   0.000   ***  1.20   0.822      \n",
      "SUPERNATURAL_EVIL         0.87   0.318        1.00   0.996      \n",
      "ATTENDANCE                0.95   0.261        0.95   0.261      \n",
      "BIBLICAL_INERRACY         1.14   0.659        1.12   0.694      \n",
      "BIBLICAL_LITERALISM       0.93   0.826        1.11   0.755      \n",
      "MAINLINE_PROTESTANT       1.35   0.428        1.50   0.309      \n",
      "BLACK_PROTESTANT          1.27   0.625        0.73   0.644      \n",
      "CONSERVATIVE_PROTESTANT   1.33   0.465        1.87   0.121      \n",
      "CATHOLIC                  1.66   0.170        1.81   0.128      \n",
      "OTHER_RELIGION            1.51   0.386        1.28   0.626      \n",
      "POLITICAL_IDEOLOGY         NaN     NaN        1.42   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN     NaN     NaN        1.66   0.380      \n",
      "HISPANIC                   NaN     NaN        1.36   0.406      \n",
      "OTHER_RACE                 NaN     NaN        0.60   0.219      \n",
      "FEMALE                     NaN     NaN        1.65   0.019     *\n",
      "LESS_THAN_HS               NaN     NaN        0.34   0.011     *\n",
      "SOME_COLLEGE               NaN     NaN        0.88   0.702      \n",
      "COLLEGE_GRADUATE           NaN     NaN        0.80   0.516      \n",
      "POSTGRADUATE               NaN     NaN        1.17   0.697      \n",
      "AGE                        NaN     NaN        1.01   0.090      \n",
      "INCOME                     NaN     NaN        1.00   0.958      \n",
      "MARRIED                    NaN     NaN        1.18   0.494      \n",
      "CHILDREN                   NaN     NaN        0.68   0.127      \n",
      "URBAN                      NaN     NaN        1.91   0.003    **\n",
      "SOUTH                      NaN     NaN        0.79   0.283      \n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  EXPANDED_SCREENING\n",
      "---------------------------------------------------------------------\n",
      "                          OR_a  pval_a sig_a  OR_b  pval_b sig_b\n",
      "Intercept                19.21   0.000   ***  1.31   0.715      \n",
      "SUPERNATURAL_EVIL         0.77   0.047     *  0.84   0.213      \n",
      "ATTENDANCE                1.01   0.814        1.03   0.519      \n",
      "BIBLICAL_INERRACY         1.12   0.669        1.16   0.593      \n",
      "BIBLICAL_LITERALISM       0.95   0.859        1.02   0.962      \n",
      "MAINLINE_PROTESTANT       1.33   0.414        1.40   0.363      \n",
      "BLACK_PROTESTANT          1.81   0.233        0.98   0.975      \n",
      "CONSERVATIVE_PROTESTANT   0.80   0.521        1.01   0.989      \n",
      "CATHOLIC                  1.65   0.144        1.65   0.165      \n",
      "OTHER_RELIGION            1.58   0.318        1.48   0.420      \n",
      "POLITICAL_IDEOLOGY         NaN     NaN        1.50   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN     NaN     NaN        1.46   0.463      \n",
      "HISPANIC                   NaN     NaN        2.80   0.014     *\n",
      "OTHER_RACE                 NaN     NaN        0.75   0.460      \n",
      "FEMALE                     NaN     NaN        2.04   0.000   ***\n",
      "LESS_THAN_HS               NaN     NaN        0.60   0.280      \n",
      "SOME_COLLEGE               NaN     NaN        0.66   0.175      \n",
      "COLLEGE_GRADUATE           NaN     NaN        0.63   0.165      \n",
      "POSTGRADUATE               NaN     NaN        0.87   0.717      \n",
      "AGE                        NaN     NaN        1.01   0.065      \n",
      "INCOME                     NaN     NaN        0.99   0.863      \n",
      "MARRIED                    NaN     NaN        1.15   0.521      \n",
      "CHILDREN                   NaN     NaN        0.65   0.061      \n",
      "URBAN                      NaN     NaN        1.33   0.141      \n",
      "SOUTH                      NaN     NaN        1.15   0.487      \n",
      "\n"
     ]
    }
   ],
   "source": [
    "from statsmodels.formula.api import logit\n",
    "\n",
    "print(\"=====================================================================\")\n",
    "print(\"                           Logistic Regression                       \")\n",
    "print(\"=====================================================================\", end=\"\\n\\n\\n\")\n",
    "\n",
    "for col in dependent_variable_cols:\n",
    "    \n",
    "    religious_columns_formula = col + \" ~ \" + '+'.join(focal_var_col + religious_variable_cols)\n",
    "    all_columns_formula = col + \" ~ \" + '+'.join(focal_var_col + religious_variable_cols + mediator_col + control_cols)\n",
    "\n",
    "    model_a = logit(formula=religious_columns_formula, data=q2[all_cols])\n",
    "    model_b = logit(formula=all_columns_formula, data=q2[all_cols])\n",
    "    \n",
    "    res_a = model_a.fit(disp=0)\n",
    "    res_b = model_b.fit(disp=0)\n",
    "    \n",
    "    print(\"---------------------------------------------------------------------\")\n",
    "    print(\"Independent Variable: \", col)\n",
    "    print(\"---------------------------------------------------------------------\")\n",
    "    \n",
    "    summary = pd.DataFrame()\n",
    "    summary.index = res_b.params.index\n",
    "    summary[\"OR_a\"] = np.exp(res_a.params).round(2)\n",
    "    summary[\"pval_a\"] = res_a.pvalues.round(3)\n",
    "    summary[\"sig_a\"] = np.where(summary[\"pval_a\"] < 0.001, \"***\", \n",
    "                                     np.where(summary[\"pval_a\"] < 0.01, \"**\", \n",
    "                                             np.where(summary[\"pval_a\"] < 0.05, \"*\", \"\")))\n",
    "    \n",
    "    summary[\"OR_b\"] = np.exp(res_b.params).round(2)\n",
    "    summary[\"pval_b\"] = res_b.pvalues.round(3)\n",
    "    summary[\"sig_b\"] = np.where(summary[\"pval_b\"] < 0.001, \"***\", \n",
    "                                     np.where(summary[\"pval_b\"] < 0.01, \"**\", \n",
    "                                             np.where(summary[\"pval_b\"] < 0.05, \"*\", \"\")))\n",
    "    print(summary, end=\"\\n\\n\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b7ab20f2",
   "metadata": {},
   "source": [
    "The results are very similar to the ones presented in tables 1 and 2 of the paper. Throughout the analyses, the factor of supernatural evil is found to be statistically significant (pvalue < 0,05), which means that it is highly unlikely that if the model is rerun with different sampling the odds ratio will differ.\n",
    "\n",
    "The only cases where the Supernatural Evil variable OR is not significant, is on the \"MORE_GUN_SAFETY\" dependent variable. Also, models containing the mediator and control variables, reduce the significance and the magnitude of the supernatural evil variable, which is an expected behaviour. For measures such as handgun bans ($OR=0.65$), support of concealed weapon carry ($OR=1.54$) and armed school security ($OR=1.61$) the effect of supernarural evil is noticeably larger ($OR>1$ means support and $OR<1$ means opposition of the examined measure).\n",
    "\n",
    "Any minor differences in the Odds Ratios and the significance of the different parameters for every model are expected, as both minor differences in the dataset as well as in the imputation and coding methods utilized do exist.\n",
    "\n",
    "Throughout the analyses, there seems to be a notewothy positive relationship between the belief in supernatural evil and the support of gun use.\n",
    "\n",
    "Another significant factor with a strong relationship between the use of guns is the political ideologies of the respondents. More liberal ideologies (value close to 7) tend to oppose the use of firearms, whereas conservative ones (value close to 1) tend to support them. For instance the odds ratio in model b for political ideology is 1.75, which means that an incremental increase in the political ideology value (range 1-7), increases the chance of one supporting a semi-auto weapon ban by $|1 - 1,75| = 75\\%$.\n",
    "\n",
    "The effect of political ideology is found to be statistically significant in every model run and its magnitude is noticeably larger than the one of belief in supernatural evil.\n",
    "\n",
    "Additionally, the gender plays a very important role regarding the support of firearm use. In the first model run, the gender has an odds ratio of 3.2, which means that is the respondent identifies as a female, the chance of them supporting a ban in semi automatic guns increases by $|1-3,2| = 220\\%$. The gender is found to be statistically significant in all but one of the models run.\n",
    "\n",
    "Moreover, racial characteristics and gun support seem to have an opposite relationship, although not always significant and therefore cannot safely conclude to the above.\n",
    "\n",
    "Finally the age of the respondent seems to be a significant factor that affects the odds of one supporting gun use. The OR in every case is close to 1 because of the many different possible values of the variable (range $19-99$). Despite close to 1 the relationship between age and gun use is always negative, which means the older the respondent, the less likely they are to support guns. This relationship is particularly intense ($OR=1,03$) when the semi auto gun and the high capacity ammo clips bans are concerned (positive relationship between age and ban support).\n",
    "\n",
    "---\n",
    "\n",
    "Regarding the article of the Economist, some possible errors can be found:\n",
    "\n",
    "1. \"A Catholic who believes in supernatural evil is more likely to hold pro-gun views than a Protestant who does not believe that Satan is corrupting souls, and vice versa.\"\n",
    "\n",
    "There are cases where the race is also a strong predictor of the respondents' views regarding gun use. For instance, black protestants have an OR of 1.8 ($|1 - 1.8| = 80\\%$ increased likehood of supporting the ban) when examining the semi-auto weapons ban. In general black protestants and other religious affiliations have a negative attitude towards gun use, that is usually statistically significant as well. Therefore although the effect of supernatural evil might be greater than a person's religion, this is not always clear, and therefore the above argument may sometimes be untrue. \n",
    "\n",
    "In addition, religious affiliation variables and especially Protestants and Catholics do not yield any statistically significant result and therefore it is not safe to be used in the example quoted above.\n",
    "\n",
    "\n",
    "2. \"The effect of belief in supernatural evil on support for the right to carry concealed guns was roughly the same as having conservative politics or not possessing a college degree, but smaller than the effect of gender.\"\n",
    "\n",
    "The effect of political views (and not specifically conservative politics) is much greater than the belief in suernatural evil when opinions about gun use are concerned. In a wider range, political opinion has an OR that is farther away from 1 compared to supernatural evil, which is also statistically significant. In addition, the wider range compared to the belief in supernatural evil, leaves room for a larger effect of the political views variable.\n",
    "\n",
    "The same is true for the gender. Although is is usually found to be well above 2, gender can only be incremented once, compared to the supernatural evil and political views variables. As a result, the cumulative percentage changes from a change in a variable with a wider range, may have a greater effect in the end, compared to the respondents' gender. This does not mean that the gender does not play any role, but certainly is not the only factor with such large efferct.\n",
    "\n",
    "For instance, age has an OR of 1,03 for the semi auto ban, which can have a very significant effect regarding the support of guns of two people of large age difference were examined (3% increase per year increase), compared to the gender (220% if the respondent is female). Gender has a larger OR, but a large age difference (e.g: 60 years) may have a greater effect.\n",
    "\n",
    "\n",
    "Also, in the paper, the **standardized coefficients** that this conclusion was based upon, **are not in fact standardized** (more details in q4). In order to obtain standardized coefficients, the variables if the linear model must also be standardized. Therefore some could argue that the results are not perfectly valid.\n",
    "\n",
    "\n",
    "### Q4: Additional Estimations of the Strength of Predictors\n",
    "\n",
    "* In order to perform standardized regression, we must use standardied coefficients as pointed out in the following [Wikipedia article](https://en.wikipedia.org/wiki/Standardized_coefficient). Specifically it is mentioned that: \"A regression carried out on standardized variables produces standardized coefficients\".\n",
    "\n",
    "* Therefore, we will be using standardized variables.\n",
    "\n",
    "* We can easily standardize our variables with the help of scikit learn's Standard Scanler.\n",
    "\n",
    "* We will put our standardized variables to a new DataFrame: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "02e06268",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>BAN_SEMI_AUTO</th>\n",
       "      <th>BAN_AMMO_CLIPS</th>\n",
       "      <th>BAN_HANDGUNS</th>\n",
       "      <th>SUPPORT_CONCEALED_CARRY</th>\n",
       "      <th>ARMED_SECURITY</th>\n",
       "      <th>ARMED_TEACHERS</th>\n",
       "      <th>MORE_GUN_SAFETY</th>\n",
       "      <th>EXPANDED_SCREENING</th>\n",
       "      <th>SUPERNATURAL_EVIL</th>\n",
       "      <th>ATTENDANCE</th>\n",
       "      <th>...</th>\n",
       "      <th>NOT_PARTNERED</th>\n",
       "      <th>CHILDREN</th>\n",
       "      <th>NO_CHILDREN</th>\n",
       "      <th>URBAN</th>\n",
       "      <th>RURAL</th>\n",
       "      <th>SOUTH</th>\n",
       "      <th>OTHER_REGION</th>\n",
       "      <th>DEVIL</th>\n",
       "      <th>HELL</th>\n",
       "      <th>DEMONS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.325856</td>\n",
       "      <td>-1.457823</td>\n",
       "      <td>-0.636259</td>\n",
       "      <td>0.89021</td>\n",
       "      <td>0.762554</td>\n",
       "      <td>1.574127</td>\n",
       "      <td>0.274298</td>\n",
       "      <td>0.311447</td>\n",
       "      <td>0.856370</td>\n",
       "      <td>1.480778</td>\n",
       "      <td>...</td>\n",
       "      <td>1.12333</td>\n",
       "      <td>-0.584206</td>\n",
       "      <td>0.583227</td>\n",
       "      <td>-1.049573</td>\n",
       "      <td>1.080868</td>\n",
       "      <td>1.283146</td>\n",
       "      <td>-1.283146</td>\n",
       "      <td>0.789139</td>\n",
       "      <td>0.835782</td>\n",
       "      <td>0.897990</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.754230</td>\n",
       "      <td>0.685954</td>\n",
       "      <td>-0.636259</td>\n",
       "      <td>0.89021</td>\n",
       "      <td>0.762554</td>\n",
       "      <td>-0.635273</td>\n",
       "      <td>0.274298</td>\n",
       "      <td>0.311447</td>\n",
       "      <td>0.856370</td>\n",
       "      <td>-0.915822</td>\n",
       "      <td>...</td>\n",
       "      <td>1.12333</td>\n",
       "      <td>1.711724</td>\n",
       "      <td>-1.714599</td>\n",
       "      <td>0.952768</td>\n",
       "      <td>-0.925182</td>\n",
       "      <td>1.283146</td>\n",
       "      <td>-1.283146</td>\n",
       "      <td>0.789139</td>\n",
       "      <td>0.835782</td>\n",
       "      <td>0.897990</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.754230</td>\n",
       "      <td>0.685954</td>\n",
       "      <td>-0.636259</td>\n",
       "      <td>-1.12333</td>\n",
       "      <td>0.762554</td>\n",
       "      <td>-0.635273</td>\n",
       "      <td>0.274298</td>\n",
       "      <td>0.311447</td>\n",
       "      <td>-1.029083</td>\n",
       "      <td>1.138407</td>\n",
       "      <td>...</td>\n",
       "      <td>1.12333</td>\n",
       "      <td>-0.584206</td>\n",
       "      <td>0.583227</td>\n",
       "      <td>-1.049573</td>\n",
       "      <td>1.080868</td>\n",
       "      <td>-0.779334</td>\n",
       "      <td>0.779334</td>\n",
       "      <td>-0.128410</td>\n",
       "      <td>-0.990276</td>\n",
       "      <td>-0.940083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.754230</td>\n",
       "      <td>0.685954</td>\n",
       "      <td>1.571687</td>\n",
       "      <td>-1.12333</td>\n",
       "      <td>0.762554</td>\n",
       "      <td>-0.635273</td>\n",
       "      <td>0.274298</td>\n",
       "      <td>0.311447</td>\n",
       "      <td>0.856370</td>\n",
       "      <td>0.796035</td>\n",
       "      <td>...</td>\n",
       "      <td>1.12333</td>\n",
       "      <td>-0.584206</td>\n",
       "      <td>0.583227</td>\n",
       "      <td>0.952768</td>\n",
       "      <td>-0.925182</td>\n",
       "      <td>1.283146</td>\n",
       "      <td>-1.283146</td>\n",
       "      <td>0.789139</td>\n",
       "      <td>0.835782</td>\n",
       "      <td>0.897990</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.754230</td>\n",
       "      <td>0.685954</td>\n",
       "      <td>1.571687</td>\n",
       "      <td>-1.12333</td>\n",
       "      <td>0.762554</td>\n",
       "      <td>1.574127</td>\n",
       "      <td>0.274298</td>\n",
       "      <td>0.311447</td>\n",
       "      <td>0.856370</td>\n",
       "      <td>1.480778</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.89021</td>\n",
       "      <td>1.711724</td>\n",
       "      <td>-1.714599</td>\n",
       "      <td>0.952768</td>\n",
       "      <td>-0.925182</td>\n",
       "      <td>-0.779334</td>\n",
       "      <td>0.779334</td>\n",
       "      <td>0.789139</td>\n",
       "      <td>0.835782</td>\n",
       "      <td>0.897990</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1567</th>\n",
       "      <td>0.754230</td>\n",
       "      <td>0.685954</td>\n",
       "      <td>1.571687</td>\n",
       "      <td>-1.12333</td>\n",
       "      <td>0.762554</td>\n",
       "      <td>-0.635273</td>\n",
       "      <td>0.274298</td>\n",
       "      <td>0.311447</td>\n",
       "      <td>-0.086357</td>\n",
       "      <td>-0.915822</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.89021</td>\n",
       "      <td>-0.584206</td>\n",
       "      <td>0.583227</td>\n",
       "      <td>-1.049573</td>\n",
       "      <td>1.080868</td>\n",
       "      <td>-0.779334</td>\n",
       "      <td>0.779334</td>\n",
       "      <td>0.789139</td>\n",
       "      <td>-0.990276</td>\n",
       "      <td>0.897990</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1568</th>\n",
       "      <td>0.754230</td>\n",
       "      <td>-1.457823</td>\n",
       "      <td>1.571687</td>\n",
       "      <td>-1.12333</td>\n",
       "      <td>-1.311382</td>\n",
       "      <td>-0.635273</td>\n",
       "      <td>0.274298</td>\n",
       "      <td>0.311447</td>\n",
       "      <td>-0.086357</td>\n",
       "      <td>-1.258193</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.89021</td>\n",
       "      <td>1.711724</td>\n",
       "      <td>-1.714599</td>\n",
       "      <td>0.952768</td>\n",
       "      <td>-0.925182</td>\n",
       "      <td>-0.779334</td>\n",
       "      <td>0.779334</td>\n",
       "      <td>-0.128410</td>\n",
       "      <td>-0.077247</td>\n",
       "      <td>-0.021047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1569</th>\n",
       "      <td>0.754230</td>\n",
       "      <td>0.685954</td>\n",
       "      <td>1.571687</td>\n",
       "      <td>-1.12333</td>\n",
       "      <td>-1.311382</td>\n",
       "      <td>-0.635273</td>\n",
       "      <td>0.274298</td>\n",
       "      <td>0.311447</td>\n",
       "      <td>0.856370</td>\n",
       "      <td>1.138407</td>\n",
       "      <td>...</td>\n",
       "      <td>1.12333</td>\n",
       "      <td>-0.584206</td>\n",
       "      <td>0.583227</td>\n",
       "      <td>-1.049573</td>\n",
       "      <td>1.080868</td>\n",
       "      <td>-0.779334</td>\n",
       "      <td>0.779334</td>\n",
       "      <td>0.789139</td>\n",
       "      <td>0.835782</td>\n",
       "      <td>-0.021047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1570</th>\n",
       "      <td>0.754230</td>\n",
       "      <td>0.685954</td>\n",
       "      <td>1.571687</td>\n",
       "      <td>-1.12333</td>\n",
       "      <td>-1.311382</td>\n",
       "      <td>-0.635273</td>\n",
       "      <td>0.274298</td>\n",
       "      <td>0.311447</td>\n",
       "      <td>-0.086357</td>\n",
       "      <td>-0.231079</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.89021</td>\n",
       "      <td>-0.584206</td>\n",
       "      <td>0.583227</td>\n",
       "      <td>0.952768</td>\n",
       "      <td>-0.925182</td>\n",
       "      <td>-0.779334</td>\n",
       "      <td>0.779334</td>\n",
       "      <td>-0.128410</td>\n",
       "      <td>-0.077247</td>\n",
       "      <td>-0.021047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1571</th>\n",
       "      <td>0.754230</td>\n",
       "      <td>0.685954</td>\n",
       "      <td>1.571687</td>\n",
       "      <td>-1.12333</td>\n",
       "      <td>0.762554</td>\n",
       "      <td>-0.635273</td>\n",
       "      <td>0.274298</td>\n",
       "      <td>0.311447</td>\n",
       "      <td>-0.086357</td>\n",
       "      <td>0.453664</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.89021</td>\n",
       "      <td>-0.584206</td>\n",
       "      <td>0.583227</td>\n",
       "      <td>0.952768</td>\n",
       "      <td>-0.925182</td>\n",
       "      <td>1.283146</td>\n",
       "      <td>-1.283146</td>\n",
       "      <td>-0.128410</td>\n",
       "      <td>-0.077247</td>\n",
       "      <td>-0.021047</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1572 rows × 44 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      BAN_SEMI_AUTO  BAN_AMMO_CLIPS  BAN_HANDGUNS  SUPPORT_CONCEALED_CARRY  \\\n",
       "0         -1.325856       -1.457823     -0.636259                  0.89021   \n",
       "1          0.754230        0.685954     -0.636259                  0.89021   \n",
       "2          0.754230        0.685954     -0.636259                 -1.12333   \n",
       "3          0.754230        0.685954      1.571687                 -1.12333   \n",
       "4          0.754230        0.685954      1.571687                 -1.12333   \n",
       "...             ...             ...           ...                      ...   \n",
       "1567       0.754230        0.685954      1.571687                 -1.12333   \n",
       "1568       0.754230       -1.457823      1.571687                 -1.12333   \n",
       "1569       0.754230        0.685954      1.571687                 -1.12333   \n",
       "1570       0.754230        0.685954      1.571687                 -1.12333   \n",
       "1571       0.754230        0.685954      1.571687                 -1.12333   \n",
       "\n",
       "      ARMED_SECURITY  ARMED_TEACHERS  MORE_GUN_SAFETY  EXPANDED_SCREENING  \\\n",
       "0           0.762554        1.574127         0.274298            0.311447   \n",
       "1           0.762554       -0.635273         0.274298            0.311447   \n",
       "2           0.762554       -0.635273         0.274298            0.311447   \n",
       "3           0.762554       -0.635273         0.274298            0.311447   \n",
       "4           0.762554        1.574127         0.274298            0.311447   \n",
       "...              ...             ...              ...                 ...   \n",
       "1567        0.762554       -0.635273         0.274298            0.311447   \n",
       "1568       -1.311382       -0.635273         0.274298            0.311447   \n",
       "1569       -1.311382       -0.635273         0.274298            0.311447   \n",
       "1570       -1.311382       -0.635273         0.274298            0.311447   \n",
       "1571        0.762554       -0.635273         0.274298            0.311447   \n",
       "\n",
       "      SUPERNATURAL_EVIL  ATTENDANCE  ...  NOT_PARTNERED  CHILDREN  \\\n",
       "0              0.856370    1.480778  ...        1.12333 -0.584206   \n",
       "1              0.856370   -0.915822  ...        1.12333  1.711724   \n",
       "2             -1.029083    1.138407  ...        1.12333 -0.584206   \n",
       "3              0.856370    0.796035  ...        1.12333 -0.584206   \n",
       "4              0.856370    1.480778  ...       -0.89021  1.711724   \n",
       "...                 ...         ...  ...            ...       ...   \n",
       "1567          -0.086357   -0.915822  ...       -0.89021 -0.584206   \n",
       "1568          -0.086357   -1.258193  ...       -0.89021  1.711724   \n",
       "1569           0.856370    1.138407  ...        1.12333 -0.584206   \n",
       "1570          -0.086357   -0.231079  ...       -0.89021 -0.584206   \n",
       "1571          -0.086357    0.453664  ...       -0.89021 -0.584206   \n",
       "\n",
       "      NO_CHILDREN     URBAN     RURAL     SOUTH  OTHER_REGION     DEVIL  \\\n",
       "0        0.583227 -1.049573  1.080868  1.283146     -1.283146  0.789139   \n",
       "1       -1.714599  0.952768 -0.925182  1.283146     -1.283146  0.789139   \n",
       "2        0.583227 -1.049573  1.080868 -0.779334      0.779334 -0.128410   \n",
       "3        0.583227  0.952768 -0.925182  1.283146     -1.283146  0.789139   \n",
       "4       -1.714599  0.952768 -0.925182 -0.779334      0.779334  0.789139   \n",
       "...           ...       ...       ...       ...           ...       ...   \n",
       "1567     0.583227 -1.049573  1.080868 -0.779334      0.779334  0.789139   \n",
       "1568    -1.714599  0.952768 -0.925182 -0.779334      0.779334 -0.128410   \n",
       "1569     0.583227 -1.049573  1.080868 -0.779334      0.779334  0.789139   \n",
       "1570     0.583227  0.952768 -0.925182 -0.779334      0.779334 -0.128410   \n",
       "1571     0.583227  0.952768 -0.925182  1.283146     -1.283146 -0.128410   \n",
       "\n",
       "          HELL    DEMONS  \n",
       "0     0.835782  0.897990  \n",
       "1     0.835782  0.897990  \n",
       "2    -0.990276 -0.940083  \n",
       "3     0.835782  0.897990  \n",
       "4     0.835782  0.897990  \n",
       "...        ...       ...  \n",
       "1567 -0.990276  0.897990  \n",
       "1568 -0.077247 -0.021047  \n",
       "1569  0.835782 -0.021047  \n",
       "1570 -0.077247 -0.021047  \n",
       "1571 -0.077247 -0.021047  \n",
       "\n",
       "[1572 rows x 44 columns]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "q4 = pd.DataFrame()\n",
    "\n",
    "scaler = StandardScaler()\n",
    "\n",
    "q4[all_cols] = scaler.fit_transform(q2[all_cols])\n",
    "q4"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "59bf44c9",
   "metadata": {},
   "source": [
    "We will then run traditional linear regression on our data\n",
    "\n",
    "* Two models were run:\n",
    " * One containing only the focal variable and the religious variables (**model_a**)\n",
    " * One containing all the variables (focal + religious + mediator + control) (**model_b**)\n",
    "\n",
    "* The linear regression model with standardized coefficients is considered to be easier to interpret.\n",
    "\n",
    "* After running the regression models (a) and (b) for each of the independent variables, we get the following results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "7dc65688",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=====================================================================\n",
      "                           Standardized OLS                          \n",
      "=====================================================================\n",
      "\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  BAN_SEMI_AUTO\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                 -0.000   1.000         -0.000   1.000      \n",
      "SUPERNATURAL_EVIL         -0.130   0.000   ***   -0.051   0.118      \n",
      "ATTENDANCE                 0.058   0.086          0.063   0.043     *\n",
      "BIBLICAL_INERRACY         -0.018   0.579         -0.004   0.881      \n",
      "BIBLICAL_LITERALISM       -0.013   0.715          0.018   0.595      \n",
      "MAINLINE_PROTESTANT        0.029   0.390          0.021   0.508      \n",
      "BLACK_PROTESTANT           0.067   0.036     *   -0.014   0.695      \n",
      "CONSERVATIVE_PROTESTANT   -0.007   0.877          0.046   0.250      \n",
      "CATHOLIC                   0.059   0.120          0.052   0.138      \n",
      "OTHER_RELIGION             0.074   0.012     *    0.039   0.149      \n",
      "POLITICAL_IDEOLOGY           NaN     NaN          0.318   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN          0.069   0.040     *\n",
      "HISPANIC                     NaN     NaN          0.058   0.019     *\n",
      "OTHER_RACE                   NaN     NaN         -0.006   0.797      \n",
      "FEMALE                       NaN     NaN          0.223   0.000   ***\n",
      "LESS_THAN_HS                 NaN     NaN         -0.043   0.095      \n",
      "SOME_COLLEGE                 NaN     NaN         -0.037   0.248      \n",
      "COLLEGE_GRADUATE             NaN     NaN         -0.069   0.037     *\n",
      "POSTGRADUATE                 NaN     NaN          0.020   0.558      \n",
      "AGE                          NaN     NaN          0.166   0.000   ***\n",
      "INCOME                       NaN     NaN          0.064   0.028     *\n",
      "MARRIED                      NaN     NaN         -0.041   0.114      \n",
      "CHILDREN                     NaN     NaN          0.009   0.734      \n",
      "URBAN                        NaN     NaN          0.056   0.018     *\n",
      "SOUTH                        NaN     NaN         -0.067   0.004    **\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  BAN_AMMO_CLIPS\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                 -0.000   1.000         -0.000   1.000      \n",
      "SUPERNATURAL_EVIL         -0.118   0.001    **   -0.053   0.118      \n",
      "ATTENDANCE                 0.061   0.070          0.061   0.054      \n",
      "BIBLICAL_INERRACY         -0.017   0.613         -0.002   0.935      \n",
      "BIBLICAL_LITERALISM       -0.034   0.348         -0.021   0.527      \n",
      "MAINLINE_PROTESTANT        0.030   0.388          0.016   0.614      \n",
      "BLACK_PROTESTANT           0.076   0.019     *    0.011   0.767      \n",
      "CONSERVATIVE_PROTESTANT   -0.001   0.977          0.044   0.278      \n",
      "CATHOLIC                   0.041   0.272          0.040   0.258      \n",
      "OTHER_RELIGION             0.065   0.029     *    0.034   0.217      \n",
      "POLITICAL_IDEOLOGY           NaN     NaN          0.283   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN          0.031   0.357      \n",
      "HISPANIC                     NaN     NaN          0.018   0.470      \n",
      "OTHER_RACE                   NaN     NaN          0.005   0.830      \n",
      "FEMALE                       NaN     NaN          0.207   0.000   ***\n",
      "LESS_THAN_HS                 NaN     NaN         -0.008   0.756      \n",
      "SOME_COLLEGE                 NaN     NaN         -0.042   0.198      \n",
      "COLLEGE_GRADUATE             NaN     NaN         -0.056   0.094      \n",
      "POSTGRADUATE                 NaN     NaN          0.020   0.561      \n",
      "AGE                          NaN     NaN          0.188   0.000   ***\n",
      "INCOME                       NaN     NaN         -0.002   0.945      \n",
      "MARRIED                      NaN     NaN          0.007   0.795      \n",
      "CHILDREN                     NaN     NaN         -0.024   0.366      \n",
      "URBAN                        NaN     NaN          0.064   0.008    **\n",
      "SOUTH                        NaN     NaN         -0.009   0.720      \n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  BAN_HANDGUNS\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                  0.000   1.000          0.000   1.000      \n",
      "SUPERNATURAL_EVIL         -0.205   0.000   ***   -0.160   0.000   ***\n",
      "ATTENDANCE                 0.050   0.127          0.062   0.057      \n",
      "BIBLICAL_INERRACY         -0.002   0.942          0.012   0.706      \n",
      "BIBLICAL_LITERALISM        0.058   0.102          0.070   0.044     *\n",
      "MAINLINE_PROTESTANT       -0.015   0.646         -0.002   0.944      \n",
      "BLACK_PROTESTANT           0.144   0.000   ***    0.085   0.024     *\n",
      "CONSERVATIVE_PROTESTANT   -0.011   0.793          0.021   0.620      \n",
      "CATHOLIC                   0.103   0.005    **    0.070   0.054      \n",
      "OTHER_RELIGION             0.098   0.001    **    0.077   0.007    **\n",
      "POLITICAL_IDEOLOGY           NaN     NaN          0.195   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN          0.060   0.086      \n",
      "HISPANIC                     NaN     NaN          0.137   0.000   ***\n",
      "OTHER_RACE                   NaN     NaN          0.088   0.000   ***\n",
      "FEMALE                       NaN     NaN          0.053   0.030     *\n",
      "LESS_THAN_HS                 NaN     NaN          0.039   0.148      \n",
      "SOME_COLLEGE                 NaN     NaN         -0.067   0.046     *\n",
      "COLLEGE_GRADUATE             NaN     NaN         -0.018   0.594      \n",
      "POSTGRADUATE                 NaN     NaN          0.023   0.511      \n",
      "AGE                          NaN     NaN          0.090   0.001    **\n",
      "INCOME                       NaN     NaN         -0.044   0.152      \n",
      "MARRIED                      NaN     NaN          0.012   0.666      \n",
      "CHILDREN                     NaN     NaN          0.033   0.223      \n",
      "URBAN                        NaN     NaN          0.057   0.021     *\n",
      "SOUTH                        NaN     NaN         -0.047   0.052      \n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  SUPPORT_CONCEALED_CARRY\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                 -0.000   1.000         -0.000   1.000      \n",
      "SUPERNATURAL_EVIL          0.220   0.000   ***    0.152   0.000   ***\n",
      "ATTENDANCE                -0.017   0.615         -0.029   0.362      \n",
      "BIBLICAL_INERRACY         -0.067   0.039     *   -0.089   0.004    **\n",
      "BIBLICAL_LITERALISM       -0.016   0.644         -0.054   0.113      \n",
      "MAINLINE_PROTESTANT       -0.002   0.946         -0.006   0.861      \n",
      "BLACK_PROTESTANT          -0.107   0.001    **   -0.076   0.040     *\n",
      "CONSERVATIVE_PROTESTANT   -0.001   0.989         -0.043   0.297      \n",
      "CATHOLIC                  -0.100   0.007    **   -0.093   0.010     *\n",
      "OTHER_RELIGION            -0.081   0.006    **   -0.051   0.068      \n",
      "POLITICAL_IDEOLOGY           NaN     NaN         -0.262   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN         -0.001   0.971      \n",
      "HISPANIC                     NaN     NaN         -0.039   0.131      \n",
      "OTHER_RACE                   NaN     NaN         -0.061   0.010     *\n",
      "FEMALE                       NaN     NaN         -0.151   0.000   ***\n",
      "LESS_THAN_HS                 NaN     NaN         -0.045   0.086      \n",
      "SOME_COLLEGE                 NaN     NaN          0.050   0.130      \n",
      "COLLEGE_GRADUATE             NaN     NaN          0.043   0.206      \n",
      "POSTGRADUATE                 NaN     NaN         -0.046   0.182      \n",
      "AGE                          NaN     NaN         -0.069   0.010     *\n",
      "INCOME                       NaN     NaN         -0.042   0.163      \n",
      "MARRIED                      NaN     NaN          0.049   0.067      \n",
      "CHILDREN                     NaN     NaN         -0.042   0.107      \n",
      "URBAN                        NaN     NaN         -0.058   0.017     *\n",
      "SOUTH                        NaN     NaN          0.070   0.004    **\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------------------------------------------------------------------\n",
      "Independent Variable:  ARMED_SECURITY\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                 -0.000   1.000         -0.000   1.000      \n",
      "SUPERNATURAL_EVIL          0.244   0.000   ***    0.157   0.000   ***\n",
      "ATTENDANCE                -0.100   0.002    **   -0.091   0.005    **\n",
      "BIBLICAL_INERRACY          0.025   0.440          0.008   0.792      \n",
      "BIBLICAL_LITERALISM        0.112   0.001    **    0.057   0.101      \n",
      "MAINLINE_PROTESTANT        0.048   0.146          0.063   0.053      \n",
      "BLACK_PROTESTANT           0.059   0.058          0.048   0.202      \n",
      "CONSERVATIVE_PROTESTANT    0.058   0.174          0.037   0.374      \n",
      "CATHOLIC                   0.040   0.268          0.048   0.188      \n",
      "OTHER_RELIGION             0.028   0.333          0.058   0.041     *\n",
      "POLITICAL_IDEOLOGY           NaN     NaN         -0.152   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN          0.030   0.397      \n",
      "HISPANIC                     NaN     NaN          0.014   0.592      \n",
      "OTHER_RACE                   NaN     NaN         -0.022   0.364      \n",
      "FEMALE                       NaN     NaN         -0.005   0.829      \n",
      "LESS_THAN_HS                 NaN     NaN         -0.053   0.050      \n",
      "SOME_COLLEGE                 NaN     NaN         -0.078   0.021     *\n",
      "COLLEGE_GRADUATE             NaN     NaN         -0.077   0.026     *\n",
      "POSTGRADUATE                 NaN     NaN         -0.156   0.000   ***\n",
      "AGE                          NaN     NaN         -0.056   0.040     *\n",
      "INCOME                       NaN     NaN         -0.107   0.000   ***\n",
      "MARRIED                      NaN     NaN          0.056   0.041     *\n",
      "CHILDREN                     NaN     NaN         -0.004   0.877      \n",
      "URBAN                        NaN     NaN         -0.035   0.157      \n",
      "SOUTH                        NaN     NaN          0.074   0.003    **\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  ARMED_TEACHERS\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                 -0.000   1.000         -0.000   1.000      \n",
      "SUPERNATURAL_EVIL          0.134   0.000   ***    0.082   0.018     *\n",
      "ATTENDANCE                -0.064   0.055         -0.082   0.012     *\n",
      "BIBLICAL_INERRACY         -0.006   0.865         -0.018   0.567      \n",
      "BIBLICAL_LITERALISM       -0.002   0.957         -0.027   0.443      \n",
      "MAINLINE_PROTESTANT        0.001   0.970         -0.006   0.856      \n",
      "BLACK_PROTESTANT          -0.058   0.069          0.024   0.530      \n",
      "CONSERVATIVE_PROTESTANT    0.090   0.038     *    0.049   0.237      \n",
      "CATHOLIC                  -0.032   0.388         -0.044   0.227      \n",
      "OTHER_RELIGION            -0.026   0.376         -0.009   0.744      \n",
      "POLITICAL_IDEOLOGY           NaN     NaN         -0.249   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN         -0.070   0.046     *\n",
      "HISPANIC                     NaN     NaN         -0.038   0.143      \n",
      "OTHER_RACE                   NaN     NaN         -0.012   0.627      \n",
      "FEMALE                       NaN     NaN         -0.146   0.000   ***\n",
      "LESS_THAN_HS                 NaN     NaN         -0.014   0.601      \n",
      "SOME_COLLEGE                 NaN     NaN         -0.003   0.922      \n",
      "COLLEGE_GRADUATE             NaN     NaN          0.001   0.970      \n",
      "POSTGRADUATE                 NaN     NaN         -0.049   0.166      \n",
      "AGE                          NaN     NaN         -0.072   0.009    **\n",
      "INCOME                       NaN     NaN         -0.020   0.514      \n",
      "MARRIED                      NaN     NaN          0.023   0.403      \n",
      "CHILDREN                     NaN     NaN          0.035   0.194      \n",
      "URBAN                        NaN     NaN         -0.083   0.001    **\n",
      "SOUTH                        NaN     NaN         -0.031   0.201      \n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  MORE_GUN_SAFETY\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                 -0.000   1.000         -0.000   1.000      \n",
      "SUPERNATURAL_EVIL         -0.033   0.341          0.005   0.883      \n",
      "ATTENDANCE                -0.038   0.257         -0.038   0.264      \n",
      "BIBLICAL_INERRACY          0.016   0.644          0.017   0.605      \n",
      "BIBLICAL_LITERALISM       -0.011   0.760          0.006   0.870      \n",
      "MAINLINE_PROTESTANT        0.026   0.449          0.035   0.316      \n",
      "BLACK_PROTESTANT           0.013   0.681         -0.015   0.699      \n",
      "CONSERVATIVE_PROTESTANT    0.029   0.514          0.062   0.160      \n",
      "CATHOLIC                   0.051   0.181          0.055   0.152      \n",
      "OTHER_RELIGION             0.025   0.395          0.012   0.689      \n",
      "POLITICAL_IDEOLOGY           NaN     NaN          0.127   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN          0.032   0.390      \n",
      "HISPANIC                     NaN     NaN          0.024   0.388      \n",
      "OTHER_RACE                   NaN     NaN         -0.028   0.280      \n",
      "FEMALE                       NaN     NaN          0.061   0.019     *\n",
      "LESS_THAN_HS                 NaN     NaN         -0.078   0.006    **\n",
      "SOME_COLLEGE                 NaN     NaN         -0.016   0.648      \n",
      "COLLEGE_GRADUATE             NaN     NaN         -0.028   0.437      \n",
      "POSTGRADUATE                 NaN     NaN          0.005   0.893      \n",
      "AGE                          NaN     NaN          0.049   0.094      \n",
      "INCOME                       NaN     NaN         -0.004   0.904      \n",
      "MARRIED                      NaN     NaN          0.025   0.393      \n",
      "CHILDREN                     NaN     NaN         -0.046   0.105      \n",
      "URBAN                        NaN     NaN          0.078   0.003    **\n",
      "SOUTH                        NaN     NaN         -0.031   0.234      \n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  EXPANDED_SCREENING\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                 -0.000   1.000         -0.000   1.000      \n",
      "SUPERNATURAL_EVIL         -0.065   0.062         -0.040   0.266      \n",
      "ATTENDANCE                 0.008   0.812          0.019   0.573      \n",
      "BIBLICAL_INERRACY          0.014   0.683          0.018   0.576      \n",
      "BIBLICAL_LITERALISM       -0.012   0.737         -0.008   0.834      \n",
      "MAINLINE_PROTESTANT        0.026   0.445          0.032   0.344      \n",
      "BLACK_PROTESTANT           0.038   0.236         -0.002   0.952      \n",
      "CONSERVATIVE_PROTESTANT   -0.044   0.314         -0.015   0.735      \n",
      "CATHOLIC                   0.051   0.181          0.045   0.240      \n",
      "OTHER_RELIGION             0.028   0.355          0.021   0.490      \n",
      "POLITICAL_IDEOLOGY           NaN     NaN          0.154   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN          0.038   0.304      \n",
      "HISPANIC                     NaN     NaN          0.071   0.010     *\n",
      "OTHER_RACE                   NaN     NaN         -0.010   0.681      \n",
      "FEMALE                       NaN     NaN          0.096   0.000   ***\n",
      "LESS_THAN_HS                 NaN     NaN         -0.031   0.268      \n",
      "SOME_COLLEGE                 NaN     NaN         -0.053   0.130      \n",
      "COLLEGE_GRADUATE             NaN     NaN         -0.060   0.097      \n",
      "POSTGRADUATE                 NaN     NaN         -0.024   0.518      \n",
      "AGE                          NaN     NaN          0.051   0.075      \n",
      "INCOME                       NaN     NaN         -0.007   0.816      \n",
      "MARRIED                      NaN     NaN          0.022   0.438      \n",
      "CHILDREN                     NaN     NaN         -0.056   0.046     *\n",
      "URBAN                        NaN     NaN          0.039   0.136      \n",
      "SOUTH                        NaN     NaN          0.013   0.614      \n",
      "\n"
     ]
    }
   ],
   "source": [
    "from statsmodels.formula.api import ols\n",
    "\n",
    "print(\"=====================================================================\")\n",
    "print(\"                           Standardized OLS                          \")\n",
    "print(\"=====================================================================\", end=\"\\n\\n\\n\")\n",
    "\n",
    "for col in dependent_variable_cols:\n",
    "    \n",
    "    religious_columns_formula = col + \" ~ \" + '+'.join(focal_var_col + religious_variable_cols)\n",
    "    all_columns_formula = col + \" ~ \" + '+'.join(focal_var_col + religious_variable_cols + mediator_col + control_cols)\n",
    "    \n",
    "    model_a = ols(formula=religious_columns_formula, data=q4)\n",
    "    model_b = ols(formula=all_columns_formula, data=q4)\n",
    "    \n",
    "    res_a = model_a.fit(disp=0)\n",
    "    res_b = model_b.fit(disp=0)\n",
    "    \n",
    "    print(\"---------------------------------------------------------------------\")\n",
    "    print(\"Independent Variable: \", col)\n",
    "    print(\"---------------------------------------------------------------------\")\n",
    "    \n",
    "    summary = pd.DataFrame()\n",
    "    summary.index = res_b.params.index\n",
    "    summary[\"coeff_a\"] = res_a.params.round(3)\n",
    "    summary[\"pval_a\"] = res_a.pvalues.round(3)\n",
    "    summary[\"sig_a\"] = np.where(summary[\"pval_a\"] < 0.001, \"***\", \n",
    "                                     np.where(summary[\"pval_a\"] < 0.01, \"**\", \n",
    "                                             np.where(summary[\"pval_a\"] < 0.05, \"*\", \"\")))\n",
    "    \n",
    "    summary[\"coeff_b\"] = res_b.params.round(3)\n",
    "    summary[\"pval_b\"] = res_b.pvalues.round(3)\n",
    "    summary[\"sig_b\"] = np.where(summary[\"pval_b\"] < 0.001, \"***\", \n",
    "                                     np.where(summary[\"pval_b\"] < 0.01, \"**\", \n",
    "                                             np.where(summary[\"pval_b\"] < 0.05, \"*\", \"\")))\n",
    "    print(summary, end=\"\\n\\n\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "22bc5046",
   "metadata": {},
   "source": [
    "In addition, the same linear regression models were run with unstandardized inputs to compare the results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "9cab7503",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=====================================================================\n",
      "                          Unstandardized OLS                         \n",
      "=====================================================================\n",
      "\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  BAN_SEMI_AUTO\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                  0.756   0.000   ***   -0.179   0.039     *\n",
      "SUPERNATURAL_EVIL         -0.059   0.000   ***   -0.023   0.118      \n",
      "ATTENDANCE                 0.010   0.086          0.010   0.043     *\n",
      "BIBLICAL_INERRACY         -0.019   0.579         -0.005   0.881      \n",
      "BIBLICAL_LITERALISM       -0.015   0.715          0.020   0.595      \n",
      "MAINLINE_PROTESTANT        0.039   0.390          0.027   0.508      \n",
      "BLACK_PROTESTANT           0.129   0.036     *   -0.027   0.695      \n",
      "CONSERVATIVE_PROTESTANT   -0.007   0.877          0.050   0.250      \n",
      "CATHOLIC                   0.066   0.120          0.059   0.138      \n",
      "OTHER_RELIGION             0.134   0.012     *    0.071   0.149      \n",
      "POLITICAL_IDEOLOGY           NaN     NaN          0.100   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN          0.107   0.040     *\n",
      "HISPANIC                     NaN     NaN          0.086   0.019     *\n",
      "OTHER_RACE                   NaN     NaN         -0.013   0.797      \n",
      "FEMALE                       NaN     NaN          0.218   0.000   ***\n",
      "LESS_THAN_HS                 NaN     NaN         -0.089   0.095      \n",
      "SOME_COLLEGE                 NaN     NaN         -0.039   0.248      \n",
      "COLLEGE_GRADUATE             NaN     NaN         -0.076   0.037     *\n",
      "POSTGRADUATE                 NaN     NaN          0.023   0.558      \n",
      "AGE                          NaN     NaN          0.005   0.000   ***\n",
      "INCOME                       NaN     NaN          0.019   0.028     *\n",
      "MARRIED                      NaN     NaN         -0.040   0.114      \n",
      "CHILDREN                     NaN     NaN          0.010   0.734      \n",
      "URBAN                        NaN     NaN          0.054   0.018     *\n",
      "SOUTH                        NaN     NaN         -0.066   0.004    **\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  BAN_AMMO_CLIPS\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                  0.784   0.000   ***   -0.039   0.652      \n",
      "SUPERNATURAL_EVIL         -0.052   0.001    **   -0.023   0.118      \n",
      "ATTENDANCE                 0.010   0.070          0.010   0.054      \n",
      "BIBLICAL_INERRACY         -0.017   0.613         -0.002   0.935      \n",
      "BIBLICAL_LITERALISM       -0.038   0.348         -0.024   0.527      \n",
      "MAINLINE_PROTESTANT        0.038   0.388          0.021   0.614      \n",
      "BLACK_PROTESTANT           0.141   0.019     *    0.020   0.767      \n",
      "CONSERVATIVE_PROTESTANT   -0.001   0.977          0.046   0.278      \n",
      "CATHOLIC                   0.046   0.272          0.044   0.258      \n",
      "OTHER_RELIGION             0.113   0.029     *    0.060   0.217      \n",
      "POLITICAL_IDEOLOGY           NaN     NaN          0.086   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN          0.047   0.357      \n",
      "HISPANIC                     NaN     NaN          0.026   0.470      \n",
      "OTHER_RACE                   NaN     NaN          0.011   0.830      \n",
      "FEMALE                       NaN     NaN          0.197   0.000   ***\n",
      "LESS_THAN_HS                 NaN     NaN         -0.017   0.756      \n",
      "SOME_COLLEGE                 NaN     NaN         -0.043   0.198      \n",
      "COLLEGE_GRADUATE             NaN     NaN         -0.061   0.094      \n",
      "POSTGRADUATE                 NaN     NaN          0.023   0.561      \n",
      "AGE                          NaN     NaN          0.005   0.000   ***\n",
      "INCOME                       NaN     NaN         -0.001   0.945      \n",
      "MARRIED                      NaN     NaN          0.006   0.795      \n",
      "CHILDREN                     NaN     NaN         -0.025   0.366      \n",
      "URBAN                        NaN     NaN          0.060   0.008    **\n",
      "SOUTH                        NaN     NaN         -0.008   0.720      \n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  BAN_HANDGUNS\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                  0.466   0.000   ***    0.022   0.797      \n",
      "SUPERNATURAL_EVIL         -0.087   0.000   ***   -0.069   0.000   ***\n",
      "ATTENDANCE                 0.008   0.127          0.010   0.057      \n",
      "BIBLICAL_INERRACY         -0.002   0.942          0.011   0.706      \n",
      "BIBLICAL_LITERALISM        0.063   0.102          0.075   0.044     *\n",
      "MAINLINE_PROTESTANT       -0.019   0.646         -0.003   0.944      \n",
      "BLACK_PROTESTANT           0.260   0.000   ***    0.153   0.024     *\n",
      "CONSERVATIVE_PROTESTANT   -0.011   0.793          0.021   0.620      \n",
      "CATHOLIC                   0.110   0.005    **    0.075   0.054      \n",
      "OTHER_RELIGION             0.167   0.001    **    0.131   0.007    **\n",
      "POLITICAL_IDEOLOGY           NaN     NaN          0.058   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN          0.087   0.086      \n",
      "HISPANIC                     NaN     NaN          0.190   0.000   ***\n",
      "OTHER_RACE                   NaN     NaN          0.182   0.000   ***\n",
      "FEMALE                       NaN     NaN          0.049   0.030     *\n",
      "LESS_THAN_HS                 NaN     NaN          0.076   0.148      \n",
      "SOME_COLLEGE                 NaN     NaN         -0.066   0.046     *\n",
      "COLLEGE_GRADUATE             NaN     NaN         -0.019   0.594      \n",
      "POSTGRADUATE                 NaN     NaN          0.026   0.511      \n",
      "AGE                          NaN     NaN          0.002   0.001    **\n",
      "INCOME                       NaN     NaN         -0.012   0.152      \n",
      "MARRIED                      NaN     NaN          0.011   0.666      \n",
      "CHILDREN                     NaN     NaN          0.034   0.223      \n",
      "URBAN                        NaN     NaN          0.051   0.021     *\n",
      "SOUTH                        NaN     NaN         -0.044   0.052      \n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  SUPPORT_CONCEALED_CARRY\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                  0.332   0.000   ***    1.022   0.000   ***\n",
      "SUPERNATURAL_EVIL          0.103   0.000   ***    0.071   0.000   ***\n",
      "ATTENDANCE                -0.003   0.615         -0.005   0.362      \n",
      "BIBLICAL_INERRACY         -0.071   0.039     *   -0.094   0.004    **\n",
      "BIBLICAL_LITERALISM       -0.020   0.644         -0.064   0.113      \n",
      "MAINLINE_PROTESTANT       -0.003   0.946         -0.008   0.861      \n",
      "BLACK_PROTESTANT          -0.212   0.001    **   -0.151   0.040     *\n",
      "CONSERVATIVE_PROTESTANT   -0.001   0.989         -0.048   0.297      \n",
      "CATHOLIC                  -0.117   0.007    **   -0.108   0.010     *\n",
      "OTHER_RELIGION            -0.151   0.006    **   -0.095   0.068      \n",
      "POLITICAL_IDEOLOGY           NaN     NaN         -0.085   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN         -0.002   0.971      \n",
      "HISPANIC                     NaN     NaN         -0.058   0.131      \n",
      "OTHER_RACE                   NaN     NaN         -0.138   0.010     *\n",
      "FEMALE                       NaN     NaN         -0.152   0.000   ***\n",
      "LESS_THAN_HS                 NaN     NaN         -0.097   0.086      \n",
      "SOME_COLLEGE                 NaN     NaN          0.054   0.130      \n",
      "COLLEGE_GRADUATE             NaN     NaN          0.049   0.206      \n",
      "POSTGRADUATE                 NaN     NaN         -0.056   0.182      \n",
      "AGE                          NaN     NaN         -0.002   0.010     *\n",
      "INCOME                       NaN     NaN         -0.012   0.163      \n",
      "MARRIED                      NaN     NaN          0.049   0.067      \n",
      "CHILDREN                     NaN     NaN         -0.048   0.107      \n",
      "URBAN                        NaN     NaN         -0.058   0.017     *\n",
      "SOUTH                        NaN     NaN          0.071   0.004    **\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------------------------------------------------------------------\n",
      "Independent Variable:  ARMED_SECURITY\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                  0.264   0.000   ***    0.844   0.000   ***\n",
      "SUPERNATURAL_EVIL          0.111   0.000   ***    0.072   0.000   ***\n",
      "ATTENDANCE                -0.016   0.002    **   -0.015   0.005    **\n",
      "BIBLICAL_INERRACY          0.025   0.440          0.008   0.792      \n",
      "BIBLICAL_LITERALISM        0.129   0.001    **    0.066   0.101      \n",
      "MAINLINE_PROTESTANT        0.063   0.146          0.084   0.053      \n",
      "BLACK_PROTESTANT           0.114   0.058          0.093   0.202      \n",
      "CONSERVATIVE_PROTESTANT    0.063   0.174          0.040   0.374      \n",
      "CATHOLIC                   0.046   0.268          0.055   0.188      \n",
      "OTHER_RELIGION             0.050   0.333          0.105   0.041     *\n",
      "POLITICAL_IDEOLOGY           NaN     NaN         -0.048   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN          0.046   0.397      \n",
      "HISPANIC                     NaN     NaN          0.021   0.592      \n",
      "OTHER_RACE                   NaN     NaN         -0.048   0.364      \n",
      "FEMALE                       NaN     NaN         -0.005   0.829      \n",
      "LESS_THAN_HS                 NaN     NaN         -0.110   0.050      \n",
      "SOME_COLLEGE                 NaN     NaN         -0.081   0.021     *\n",
      "COLLEGE_GRADUATE             NaN     NaN         -0.085   0.026     *\n",
      "POSTGRADUATE                 NaN     NaN         -0.184   0.000   ***\n",
      "AGE                          NaN     NaN         -0.002   0.040     *\n",
      "INCOME                       NaN     NaN         -0.031   0.000   ***\n",
      "MARRIED                      NaN     NaN          0.054   0.041     *\n",
      "CHILDREN                     NaN     NaN         -0.005   0.877      \n",
      "URBAN                        NaN     NaN         -0.034   0.157      \n",
      "SOUTH                        NaN     NaN          0.074   0.003    **\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  ARMED_TEACHERS\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                  0.143   0.000   ***    0.781   0.000   ***\n",
      "SUPERNATURAL_EVIL          0.057   0.000   ***    0.035   0.018     *\n",
      "ATTENDANCE                -0.010   0.055         -0.013   0.012     *\n",
      "BIBLICAL_INERRACY         -0.005   0.865         -0.017   0.567      \n",
      "BIBLICAL_LITERALISM       -0.002   0.957         -0.029   0.443      \n",
      "MAINLINE_PROTESTANT        0.002   0.970         -0.007   0.856      \n",
      "BLACK_PROTESTANT          -0.105   0.069          0.043   0.530      \n",
      "CONSERVATIVE_PROTESTANT    0.092   0.038     *    0.050   0.237      \n",
      "CATHOLIC                  -0.035   0.388         -0.047   0.227      \n",
      "OTHER_RELIGION            -0.044   0.376         -0.016   0.744      \n",
      "POLITICAL_IDEOLOGY           NaN     NaN         -0.073   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN         -0.102   0.046     *\n",
      "HISPANIC                     NaN     NaN         -0.053   0.143      \n",
      "OTHER_RACE                   NaN     NaN         -0.024   0.627      \n",
      "FEMALE                       NaN     NaN         -0.134   0.000   ***\n",
      "LESS_THAN_HS                 NaN     NaN         -0.028   0.601      \n",
      "SOME_COLLEGE                 NaN     NaN         -0.003   0.922      \n",
      "COLLEGE_GRADUATE             NaN     NaN          0.001   0.970      \n",
      "POSTGRADUATE                 NaN     NaN         -0.054   0.166      \n",
      "AGE                          NaN     NaN         -0.002   0.009    **\n",
      "INCOME                       NaN     NaN         -0.005   0.514      \n",
      "MARRIED                      NaN     NaN          0.021   0.403      \n",
      "CHILDREN                     NaN     NaN          0.036   0.194      \n",
      "URBAN                        NaN     NaN         -0.075   0.001    **\n",
      "SOUTH                        NaN     NaN         -0.029   0.201      \n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  MORE_GUN_SAFETY\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                  0.949   0.000   ***    0.771   0.000   ***\n",
      "SUPERNATURAL_EVIL         -0.008   0.341          0.001   0.883      \n",
      "ATTENDANCE                -0.003   0.257         -0.003   0.264      \n",
      "BIBLICAL_INERRACY          0.008   0.644          0.009   0.605      \n",
      "BIBLICAL_LITERALISM       -0.007   0.760          0.004   0.870      \n",
      "MAINLINE_PROTESTANT        0.018   0.449          0.024   0.316      \n",
      "BLACK_PROTESTANT           0.014   0.681         -0.016   0.699      \n",
      "CONSERVATIVE_PROTESTANT    0.017   0.514          0.036   0.160      \n",
      "CATHOLIC                   0.031   0.181          0.033   0.152      \n",
      "OTHER_RELIGION             0.024   0.395          0.012   0.689      \n",
      "POLITICAL_IDEOLOGY           NaN     NaN          0.021   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN          0.026   0.390      \n",
      "HISPANIC                     NaN     NaN          0.019   0.388      \n",
      "OTHER_RACE                   NaN     NaN         -0.032   0.280      \n",
      "FEMALE                       NaN     NaN          0.032   0.019     *\n",
      "LESS_THAN_HS                 NaN     NaN         -0.086   0.006    **\n",
      "SOME_COLLEGE                 NaN     NaN         -0.009   0.648      \n",
      "COLLEGE_GRADUATE             NaN     NaN         -0.017   0.437      \n",
      "POSTGRADUATE                 NaN     NaN          0.003   0.893      \n",
      "AGE                          NaN     NaN          0.001   0.094      \n",
      "INCOME                       NaN     NaN         -0.001   0.904      \n",
      "MARRIED                      NaN     NaN          0.013   0.393      \n",
      "CHILDREN                     NaN     NaN         -0.027   0.105      \n",
      "URBAN                        NaN     NaN          0.040   0.003    **\n",
      "SOUTH                        NaN     NaN         -0.016   0.234      \n",
      "\n",
      "---------------------------------------------------------------------\n",
      "Independent Variable:  EXPANDED_SCREENING\n",
      "---------------------------------------------------------------------\n",
      "                         coeff_a  pval_a sig_a  coeff_b  pval_b sig_b\n",
      "Intercept                  0.953   0.000   ***    0.747   0.000   ***\n",
      "SUPERNATURAL_EVIL         -0.017   0.062         -0.011   0.266      \n",
      "ATTENDANCE                 0.001   0.812          0.002   0.573      \n",
      "BIBLICAL_INERRACY          0.008   0.683          0.011   0.576      \n",
      "BIBLICAL_LITERALISM       -0.008   0.737         -0.005   0.834      \n",
      "MAINLINE_PROTESTANT        0.020   0.445          0.025   0.344      \n",
      "BLACK_PROTESTANT           0.043   0.236         -0.003   0.952      \n",
      "CONSERVATIVE_PROTESTANT   -0.028   0.314         -0.009   0.735      \n",
      "CATHOLIC                   0.034   0.181          0.030   0.240      \n",
      "OTHER_RELIGION             0.029   0.355          0.022   0.490      \n",
      "POLITICAL_IDEOLOGY           NaN     NaN          0.029   0.000   ***\n",
      "BLACK_AFRICAN_AMERICAN       NaN     NaN          0.034   0.304      \n",
      "HISPANIC                     NaN     NaN          0.061   0.010     *\n",
      "OTHER_RACE                   NaN     NaN         -0.013   0.681      \n",
      "FEMALE                       NaN     NaN          0.055   0.000   ***\n",
      "LESS_THAN_HS                 NaN     NaN         -0.038   0.268      \n",
      "SOME_COLLEGE                 NaN     NaN         -0.033   0.130      \n",
      "COLLEGE_GRADUATE             NaN     NaN         -0.039   0.097      \n",
      "POSTGRADUATE                 NaN     NaN         -0.017   0.518      \n",
      "AGE                          NaN     NaN          0.001   0.075      \n",
      "INCOME                       NaN     NaN         -0.001   0.816      \n",
      "MARRIED                      NaN     NaN          0.013   0.438      \n",
      "CHILDREN                     NaN     NaN         -0.036   0.046     *\n",
      "URBAN                        NaN     NaN          0.022   0.136      \n",
      "SOUTH                        NaN     NaN          0.008   0.614      \n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(\"=====================================================================\")\n",
    "print(\"                          Unstandardized OLS                         \")\n",
    "print(\"=====================================================================\", end=\"\\n\\n\\n\")\n",
    "\n",
    "for col in dependent_variable_cols:\n",
    "    \n",
    "    religious_columns_formula = col + \" ~ \" + '+'.join(focal_var_col + religious_variable_cols)\n",
    "    all_columns_formula = col + \" ~ \" + '+'.join(focal_var_col + religious_variable_cols + mediator_col + control_cols)\n",
    "    \n",
    "    model_a = ols(formula=religious_columns_formula, data=q2)\n",
    "    model_b = ols(formula=all_columns_formula, data=q2)\n",
    "    \n",
    "    res_a = model_a.fit(disp=0)\n",
    "    res_b = model_b.fit(disp=0)\n",
    "    \n",
    "    print(\"---------------------------------------------------------------------\")\n",
    "    print(\"Independent Variable: \", col)\n",
    "    print(\"---------------------------------------------------------------------\")\n",
    "    \n",
    "    summary = pd.DataFrame()\n",
    "    summary.index = res_b.params.index\n",
    "    summary[\"coeff_a\"] = res_a.params.round(3)\n",
    "    summary[\"pval_a\"] = res_a.pvalues.round(3)\n",
    "    summary[\"sig_a\"] = np.where(summary[\"pval_a\"] < 0.001, \"***\", \n",
    "                                     np.where(summary[\"pval_a\"] < 0.01, \"**\", \n",
    "                                             np.where(summary[\"pval_a\"] < 0.05, \"*\", \"\")))\n",
    "    \n",
    "    summary[\"coeff_b\"] = res_b.params.round(3)\n",
    "    summary[\"pval_b\"] = res_b.pvalues.round(3)\n",
    "    summary[\"sig_b\"] = np.where(summary[\"pval_b\"] < 0.001, \"***\", \n",
    "                                     np.where(summary[\"pval_b\"] < 0.01, \"**\", \n",
    "                                             np.where(summary[\"pval_b\"] < 0.05, \"*\", \"\")))\n",
    "    print(summary, end=\"\\n\\n\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fa1c2ddc",
   "metadata": {},
   "source": [
    "The standardized coefficients allow us to compare more easily the effect of different predictors, as they are all examined in the same scale.\n",
    "\n",
    "Contrary to the paper's results, regarding concealed carry, although the effect of gender ($-0.15$) and the effect of college degree ($0.5$) were found to be similar to the paper, the effect of supernatural evil ($0.15$) was similar to the effect of gender , while political ideology was found to have a way larger effect ($0.26$).\n",
    "\n",
    "However, we notice that the unstandardized coefficients in fact yield the results described in the paper. That is, for concealed carry, the effect of supernatural evil ($0.07$) is indeed roughly the same as the effect of having a college degree ($0.05$) and political ideology ($0.085$), while the effect of gender is far more important ($-0.15$). \n",
    "\n",
    "Without the use of standardized coefficients, we cannot be sure about the real effect of each variable, as the variables compared do not have the same scale. Therefore, the standardized coefficients lead to much safer conclusions about the relationship of the variables.\n",
    "\n",
    "Having a look to the the standardized coefficients again, a different picture exists while examining different independent variables. Regarding the semi auto ban, the effect of supernatural evil ($0.05$) is significantly smaller than the effect of gender ($0.22$). Political ideology still remains the factor with the largest effect ($0.32$). We should note that the effect of age is also significant ($0.17$).\n",
    "\n",
    "While examining the rest of the standardized coefficient models, a similar picture exists: \n",
    "* The effect of political ideology is always the one with the largest effect.\n",
    "* Gender is usually more important than belief in supernatural evil when determining the attitude towards gun use.\n",
    "* Having a college degree is plays a less important role compared to the above variables.\n",
    "* The effect of age is also important and especially regarding the ban of semi-auto guns and high capacity ammi clips plays a much heavier role than supernatural evil. In the other models age remains an important factor but has a smaller magnitude compared to the belief in supernatural evil.\n",
    "\n",
    "Therefore, we can conclude that despite the fact that belief in supernatural evil may be a strong predictor regarding pro-gun beliefs, political ideology, gender and sometimes the effect of age have a larger effect in determining one's attitude towards guns. The effect of college degrees remains important but is less intense.\n",
    "\n",
    "\n",
    "### Q5: Compare with Decision Trees / Random Forest Estimators\n",
    "\n",
    "\n",
    "For the analysis that follows, we need to select the columns used in the decision threes. These include the variables used in the analysis of q2 and q3 but also the dummy variables not included in the regression:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "62c27074",
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_cols = [\"SUPERNATURAL_EVIL\", \"ATTENDANCE\", \"BIBLICAL_LITERALISM\", \n",
    "    \"BIBLICAL_INERRACY\", \"BIBLE_HUMAN_ERROR\", \"CONSERVATIVE_PROTESTANT\", \"MAINLINE_PROTESTANT\", \"BLACK_PROTESTANT\", \"CATHOLIC\", \n",
    "    \"OTHER_RELIGION\", \"NO_AFFILIATION\", \"POLITICAL_IDEOLOGY\",\"AGE\",\n",
    "    \"MALE\", \"FEMALE\",\"WHITE\", \"HISPANIC\",\"BLACK_AFRICAN_AMERICAN\",\"OTHER_RACE\",\n",
    "    \"LESS_THAN_HS\",\"HIGH_SCHOOL_OR_EQUIVALENT\", \"SOME_COLLEGE\",\"COLLEGE_GRADUATE\",\n",
    "    \"POSTGRADUATE\",\"INCOME\",\"MARRIED\", \"NOT_PARTNERED\",\"CHILDREN\", \"NO_CHILDREN\" ,\"URBAN\", \"RURAL\",\n",
    "    \"SOUTH\", \"OTHER_REGION\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "730a5ddc",
   "metadata": {},
   "source": [
    "* Firstly, we need to determine the best tree to use for our analysis. This will be done by comparing the following classifiers:\n",
    " 1. DecisionTreeClassifier (simple decision tree)\n",
    " 2. BaggingClassifier on the DecisionTreeClassifier (bagging for the simple decision tree)\n",
    " 3. RandomForestClassifier (random forests)\n",
    " 4. BaggingClassifier on the RandomForestClassifier (bagging for the random forests)\n",
    " 5. ExtraTreesClassifier (extremely randomized trees)\n",
    " 6. AdaBoostClassifier on the DecisionTreeClassifier (boosting on the simple decision tree)\n",
    " 7. XGBoostClassifier\n",
    " 8. LightGBMClassifier\n",
    " \n",
    " \n",
    "* For running all the models, some assumptions had to be made and some parameters have to be kept the same:\n",
    " * Random state is set to 0 in all models\n",
    " * Where applicable, entropy was selected as the preferred criterion as it yielded more accurate results\n",
    " * All the trees were set to have a maximum depth of 5. This prevented overfitting and led to significantly improved results.\n",
    " * We assume that the accuracy determines how well a decision tree performs. It will be used to measure the performance of every decision tree model examined. MORE_GUN_SAFETY and EXPANDED_SCREENING variables have very high percentage of positives (1) therefore accuracy may not be the best metric to use. However, this effect is mediated from the other six variables and as a result we assume that accuracy is a good estimator.\n",
    " * We will be using Classifiers as our variable take distinct values.\n",
    " \n",
    "We proceed by creating a dataframe where we will be keeping all the accuracies for the different variables regarding gun use:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "d7fdef99",
   "metadata": {},
   "outputs": [],
   "source": [
    "cv_scores = pd.DataFrame()\n",
    "cv_scores.index = dependent_variable_cols"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4d411f4b",
   "metadata": {},
   "source": [
    "First, we will be creating a simple Decision Tree:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "37f696e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "scores = []\n",
    "\n",
    "gun_use_tree = DecisionTreeClassifier(random_state=0, criterion='entropy', max_depth=5)\n",
    "for col in dependent_variable_cols:\n",
    "    cv_score = cross_val_score(gun_use_tree, \n",
    "                               q2.loc[:, feature_cols], \n",
    "                               q2[col])\n",
    "    scores.append(cv_score.mean())\n",
    "cv_scores[\"SimpleTree\"] = scores"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "55799d6f",
   "metadata": {},
   "source": [
    "We create a bagging classifier for our initial simple Decision Tree:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "922d7332",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import BaggingClassifier\n",
    "\n",
    "scores = []\n",
    "bagging_classifier = BaggingClassifier(gun_use_tree,\n",
    "                                       n_estimators=10, \n",
    "                                       n_jobs=-1,\n",
    "                                       random_state=0)\n",
    "for col in dependent_variable_cols:\n",
    "    cv_score = cross_val_score(bagging_classifier, \n",
    "                               q2.loc[:, feature_cols], \n",
    "                               q2[col], \n",
    "                               n_jobs=-1)\n",
    "    scores.append(cv_score.mean())\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "47fe21e6",
   "metadata": {},
   "source": [
    "We create a Bagging Classifier for our initial simple Decision Tree:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "213a258f",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import BaggingClassifier\n",
    "\n",
    "scores = []\n",
    "bagging_classifier = BaggingClassifier(gun_use_tree,\n",
    "                                       n_estimators=50, \n",
    "                                       n_jobs=-1,\n",
    "                                       random_state=0)\n",
    "for col in dependent_variable_cols:\n",
    "    cv_score = cross_val_score(bagging_classifier, \n",
    "                               q2.loc[:, feature_cols], \n",
    "                               q2[col],\n",
    "                               n_jobs=-1)\n",
    "    scores.append(cv_score.mean())\n",
    "cv_scores[\"BagSimple\"] = scores"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b4d14211",
   "metadata": {},
   "source": [
    "We create a Random Forest classifier:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "78bd9341",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "scores = []\n",
    "forest_classifier = RandomForestClassifier(n_estimators=50, \n",
    "                                               max_depth=5,\n",
    "                                               min_samples_split=2, \n",
    "                                               n_jobs=-1,\n",
    "                                               random_state=0,\n",
    "                                               criterion='entropy')\n",
    "for col in dependent_variable_cols:\n",
    "    cv_score = cross_val_score(forest_classifier, \n",
    "                               q2.loc[:, feature_cols], \n",
    "                               q2[col],\n",
    "                               n_jobs=-1)\n",
    "    scores.append(cv_score.mean())\n",
    "\n",
    "cv_scores[\"RndForest\"] = scores"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6e7a814f",
   "metadata": {},
   "source": [
    "We create a Bagging Classifier for our Random Forest Classifier:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "7bd897b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "scores = []\n",
    "bagging_forest_classifier = BaggingClassifier(forest_classifier,\n",
    "                                       n_estimators=10, \n",
    "                                       n_jobs=-1,\n",
    "                                       random_state=0)\n",
    "for col in dependent_variable_cols:\n",
    "    cv_score = cross_val_score(bagging_forest_classifier, \n",
    "                               q2.loc[:, feature_cols], \n",
    "                               q2[col],\n",
    "                               n_jobs=-1)\n",
    "    scores.append(cv_score.mean())\n",
    "\n",
    "cv_scores[\"BagForest\"] = scores"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1a92278a",
   "metadata": {},
   "source": [
    "We create a Extremely Randomized Tree Classifier:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "07048367",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import ExtraTreesClassifier\n",
    "\n",
    "scores = []\n",
    "extremely_random_trees_classifier = ExtraTreesClassifier(n_estimators=50, \n",
    "                                                         max_depth=5,\n",
    "                                                         min_samples_split=2,\n",
    "                                                         n_jobs=-1, \n",
    "                                                         criterion='entropy',\n",
    "                                                         random_state=0)\n",
    "for col in dependent_variable_cols:\n",
    "    cv_score = cross_val_score(extremely_random_trees_classifier, \n",
    "                               q2.loc[:, feature_cols], \n",
    "                               q2[col],\n",
    "                               scoring='accuracy')\n",
    "    scores.append(cv_score.mean())\n",
    "\n",
    "cv_scores[\"RndTrees\"] = scores"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c619d5d9",
   "metadata": {},
   "source": [
    "We create an AdaBoost Classifier for our initial simple Decision Tree:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "826cc167",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import AdaBoostClassifier\n",
    "\n",
    "scores = []\n",
    "rgr = AdaBoostClassifier(gun_use_tree, \n",
    "                        n_estimators=10, \n",
    "                        random_state=0)\n",
    "for col in dependent_variable_cols:\n",
    "    cv_score = cross_val_score(rgr, \n",
    "                               q2.loc[:, feature_cols], \n",
    "                               q2[col],\n",
    "                               n_jobs=-1)\n",
    "    scores.append(cv_score.mean())\n",
    "    \n",
    "cv_scores[\"BoostSimple\"] = scores"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "340a30f6",
   "metadata": {},
   "source": [
    "We create an XGBoost Classifier:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "8f50e1b8",
   "metadata": {},
   "outputs": [],
   "source": [
    "import xgboost as xgb\n",
    "\n",
    "scores = []\n",
    "xgb_classifier = xgb.XGBClassifier(use_label_encoder=False, \n",
    "                                   verbosity=0, \n",
    "                                   random_state=0, \n",
    "                                   n_jobs=-1,\n",
    "                                   max_depth=5)\n",
    "for col in dependent_variable_cols:\n",
    "    cv_score = cross_val_score(xgb_classifier, \n",
    "                               q2.loc[:, feature_cols], \n",
    "                               q2[col],\n",
    "                               n_jobs=-1)\n",
    "    scores.append(cv_score.mean())\n",
    "    \n",
    "cv_scores[\"XGBoost\"] = scores"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "120c4cb5",
   "metadata": {},
   "source": [
    "We create an LightGBM Classifier:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "8feb1946",
   "metadata": {},
   "outputs": [],
   "source": [
    "import lightgbm as lgb\n",
    "\n",
    "scores = []\n",
    "lgb_classifier = lgb.LGBMClassifier(random_state=0, \n",
    "                                    n_jobs=-1,\n",
    "                                    max_depth=5)\n",
    "for col in dependent_variable_cols:\n",
    "    cv_score = cross_val_score(lgb_classifier, \n",
    "                               q2.loc[:, feature_cols], \n",
    "                               q2[col],\n",
    "                               n_jobs=-1)\n",
    "    scores.append(cv_score.mean())\n",
    "    \n",
    "cv_scores[\"LGBM\"] = scores"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b233e176",
   "metadata": {},
   "source": [
    "* We then plot the accuracies along with the mean accuracies.\n",
    "* The columns are sorted according to the average accuracy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "f0f47e60",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>BagForest</th>\n",
       "      <th>BagSimple</th>\n",
       "      <th>RndForest</th>\n",
       "      <th>LGBM</th>\n",
       "      <th>RndTrees</th>\n",
       "      <th>SimpleTree</th>\n",
       "      <th>XGBoost</th>\n",
       "      <th>BoostSimple</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>BAN_SEMI_AUTO</th>\n",
       "      <td>0.718815</td>\n",
       "      <td>0.714991</td>\n",
       "      <td>0.716259</td>\n",
       "      <td>0.714354</td>\n",
       "      <td>0.695905</td>\n",
       "      <td>0.701001</td>\n",
       "      <td>0.711182</td>\n",
       "      <td>0.648861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BAN_AMMO_CLIPS</th>\n",
       "      <td>0.726450</td>\n",
       "      <td>0.729008</td>\n",
       "      <td>0.720734</td>\n",
       "      <td>0.721365</td>\n",
       "      <td>0.716271</td>\n",
       "      <td>0.701041</td>\n",
       "      <td>0.693380</td>\n",
       "      <td>0.660283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BAN_HANDGUNS</th>\n",
       "      <td>0.716283</td>\n",
       "      <td>0.734737</td>\n",
       "      <td>0.714373</td>\n",
       "      <td>0.717565</td>\n",
       "      <td>0.709281</td>\n",
       "      <td>0.716283</td>\n",
       "      <td>0.707352</td>\n",
       "      <td>0.690846</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SUPPORT_CONCEALED_CARRY</th>\n",
       "      <td>0.670478</td>\n",
       "      <td>0.653313</td>\n",
       "      <td>0.657108</td>\n",
       "      <td>0.668582</td>\n",
       "      <td>0.651372</td>\n",
       "      <td>0.616403</td>\n",
       "      <td>0.634201</td>\n",
       "      <td>0.601146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ARMED_SECURITY</th>\n",
       "      <td>0.697828</td>\n",
       "      <td>0.688307</td>\n",
       "      <td>0.688288</td>\n",
       "      <td>0.669214</td>\n",
       "      <td>0.692100</td>\n",
       "      <td>0.684477</td>\n",
       "      <td>0.651410</td>\n",
       "      <td>0.622794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ARMED_TEACHERS</th>\n",
       "      <td>0.722643</td>\n",
       "      <td>0.725834</td>\n",
       "      <td>0.725187</td>\n",
       "      <td>0.716922</td>\n",
       "      <td>0.720105</td>\n",
       "      <td>0.718208</td>\n",
       "      <td>0.682560</td>\n",
       "      <td>0.676844</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MORE_GUN_SAFETY</th>\n",
       "      <td>0.930025</td>\n",
       "      <td>0.930025</td>\n",
       "      <td>0.930025</td>\n",
       "      <td>0.922396</td>\n",
       "      <td>0.930025</td>\n",
       "      <td>0.925577</td>\n",
       "      <td>0.914759</td>\n",
       "      <td>0.886133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EXPANDED_SCREENING</th>\n",
       "      <td>0.911578</td>\n",
       "      <td>0.910943</td>\n",
       "      <td>0.911578</td>\n",
       "      <td>0.899488</td>\n",
       "      <td>0.911578</td>\n",
       "      <td>0.903951</td>\n",
       "      <td>0.896951</td>\n",
       "      <td>0.868962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MEAN</th>\n",
       "      <td>0.761763</td>\n",
       "      <td>0.760895</td>\n",
       "      <td>0.757944</td>\n",
       "      <td>0.753736</td>\n",
       "      <td>0.753330</td>\n",
       "      <td>0.745868</td>\n",
       "      <td>0.736474</td>\n",
       "      <td>0.706984</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         BagForest  BagSimple  RndForest      LGBM  RndTrees  \\\n",
       "BAN_SEMI_AUTO             0.718815   0.714991   0.716259  0.714354  0.695905   \n",
       "BAN_AMMO_CLIPS            0.726450   0.729008   0.720734  0.721365  0.716271   \n",
       "BAN_HANDGUNS              0.716283   0.734737   0.714373  0.717565  0.709281   \n",
       "SUPPORT_CONCEALED_CARRY   0.670478   0.653313   0.657108  0.668582  0.651372   \n",
       "ARMED_SECURITY            0.697828   0.688307   0.688288  0.669214  0.692100   \n",
       "ARMED_TEACHERS            0.722643   0.725834   0.725187  0.716922  0.720105   \n",
       "MORE_GUN_SAFETY           0.930025   0.930025   0.930025  0.922396  0.930025   \n",
       "EXPANDED_SCREENING        0.911578   0.910943   0.911578  0.899488  0.911578   \n",
       "MEAN                      0.761763   0.760895   0.757944  0.753736  0.753330   \n",
       "\n",
       "                         SimpleTree   XGBoost  BoostSimple  \n",
       "BAN_SEMI_AUTO              0.701001  0.711182     0.648861  \n",
       "BAN_AMMO_CLIPS             0.701041  0.693380     0.660283  \n",
       "BAN_HANDGUNS               0.716283  0.707352     0.690846  \n",
       "SUPPORT_CONCEALED_CARRY    0.616403  0.634201     0.601146  \n",
       "ARMED_SECURITY             0.684477  0.651410     0.622794  \n",
       "ARMED_TEACHERS             0.718208  0.682560     0.676844  \n",
       "MORE_GUN_SAFETY            0.925577  0.914759     0.886133  \n",
       "EXPANDED_SCREENING         0.903951  0.896951     0.868962  \n",
       "MEAN                       0.745868  0.736474     0.706984  "
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_scores.loc[\"MEAN\"] = cv_scores.mean()\n",
    "cv_scores.reindex(cv_scores.mean().sort_values(ascending=False).index, axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "20dc9cfd",
   "metadata": {},
   "source": [
    "* We can see that all the decision trees perform well, but the ones that have a particularly high accuracy are the Simple and Random Forest Classifiers combined with Bagging, with a mean accuracy of just above $76\\%$.\n",
    "\n",
    "* Therefore, we will be selecting the **random forests with bagging** to proceed.\n",
    "\n",
    "* In order to obtain the feature importances of the bagging classifier, instructions from this [stack overflow post](https://stackoverflow.com/questions/44333573/feature-importances-bagging-scikit-learn) were followed, as scikit learn does not have a built-in way of obtaining the feature importances for the bagging classifier. The mean feature importance of bagging estimators was calculated for each dependent variable.\n",
    "\n",
    "\n",
    "* This way, we get the feature importances for every dependent variable. Below we can see the 20 most important features according to the mean feature importance for all 8 gun related variables:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "e7a9b337",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>BAN_SEMI_AUTO</th>\n",
       "      <th>BAN_AMMO_CLIPS</th>\n",
       "      <th>BAN_HANDGUNS</th>\n",
       "      <th>SUPPORT_CONCEALED_CARRY</th>\n",
       "      <th>ARMED_SECURITY</th>\n",
       "      <th>ARMED_TEACHERS</th>\n",
       "      <th>MORE_GUN_SAFETY</th>\n",
       "      <th>EXPANDED_SCREENING</th>\n",
       "      <th>MEAN</th>\n",
       "      <th>STD</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>POLITICAL_IDEOLOGY</th>\n",
       "      <td>0.205330</td>\n",
       "      <td>0.189869</td>\n",
       "      <td>0.152070</td>\n",
       "      <td>0.205110</td>\n",
       "      <td>0.132639</td>\n",
       "      <td>0.213973</td>\n",
       "      <td>0.144500</td>\n",
       "      <td>0.164445</td>\n",
       "      <td>0.175992</td>\n",
       "      <td>0.027035</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AGE</th>\n",
       "      <td>0.109869</td>\n",
       "      <td>0.140372</td>\n",
       "      <td>0.102683</td>\n",
       "      <td>0.093378</td>\n",
       "      <td>0.095629</td>\n",
       "      <td>0.108311</td>\n",
       "      <td>0.128774</td>\n",
       "      <td>0.135897</td>\n",
       "      <td>0.114364</td>\n",
       "      <td>0.025219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SUPERNATURAL_EVIL</th>\n",
       "      <td>0.043089</td>\n",
       "      <td>0.045292</td>\n",
       "      <td>0.066669</td>\n",
       "      <td>0.076422</td>\n",
       "      <td>0.131411</td>\n",
       "      <td>0.056719</td>\n",
       "      <td>0.043802</td>\n",
       "      <td>0.064078</td>\n",
       "      <td>0.065935</td>\n",
       "      <td>0.015295</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>INCOME</th>\n",
       "      <td>0.045523</td>\n",
       "      <td>0.047309</td>\n",
       "      <td>0.070199</td>\n",
       "      <td>0.048583</td>\n",
       "      <td>0.099614</td>\n",
       "      <td>0.049029</td>\n",
       "      <td>0.072860</td>\n",
       "      <td>0.059765</td>\n",
       "      <td>0.061610</td>\n",
       "      <td>0.014017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ATTENDANCE</th>\n",
       "      <td>0.045266</td>\n",
       "      <td>0.049327</td>\n",
       "      <td>0.056277</td>\n",
       "      <td>0.049721</td>\n",
       "      <td>0.054600</td>\n",
       "      <td>0.055002</td>\n",
       "      <td>0.093758</td>\n",
       "      <td>0.053073</td>\n",
       "      <td>0.057128</td>\n",
       "      <td>0.010688</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FEMALE</th>\n",
       "      <td>0.090581</td>\n",
       "      <td>0.086458</td>\n",
       "      <td>0.018438</td>\n",
       "      <td>0.044560</td>\n",
       "      <td>0.016042</td>\n",
       "      <td>0.053555</td>\n",
       "      <td>0.029534</td>\n",
       "      <td>0.037754</td>\n",
       "      <td>0.047115</td>\n",
       "      <td>0.009316</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MALE</th>\n",
       "      <td>0.091956</td>\n",
       "      <td>0.084623</td>\n",
       "      <td>0.017975</td>\n",
       "      <td>0.046331</td>\n",
       "      <td>0.017434</td>\n",
       "      <td>0.046392</td>\n",
       "      <td>0.029946</td>\n",
       "      <td>0.035173</td>\n",
       "      <td>0.046229</td>\n",
       "      <td>0.007061</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>WHITE</th>\n",
       "      <td>0.016707</td>\n",
       "      <td>0.015227</td>\n",
       "      <td>0.074660</td>\n",
       "      <td>0.022770</td>\n",
       "      <td>0.012431</td>\n",
       "      <td>0.021826</td>\n",
       "      <td>0.014846</td>\n",
       "      <td>0.027538</td>\n",
       "      <td>0.025751</td>\n",
       "      <td>0.008054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>POSTGRADUATE</th>\n",
       "      <td>0.023433</td>\n",
       "      <td>0.020622</td>\n",
       "      <td>0.021917</td>\n",
       "      <td>0.035693</td>\n",
       "      <td>0.044597</td>\n",
       "      <td>0.023973</td>\n",
       "      <td>0.015770</td>\n",
       "      <td>0.015794</td>\n",
       "      <td>0.025225</td>\n",
       "      <td>0.003162</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RURAL</th>\n",
       "      <td>0.021395</td>\n",
       "      <td>0.015949</td>\n",
       "      <td>0.023733</td>\n",
       "      <td>0.024030</td>\n",
       "      <td>0.016967</td>\n",
       "      <td>0.029684</td>\n",
       "      <td>0.030772</td>\n",
       "      <td>0.026107</td>\n",
       "      <td>0.023580</td>\n",
       "      <td>0.007292</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>URBAN</th>\n",
       "      <td>0.015785</td>\n",
       "      <td>0.017894</td>\n",
       "      <td>0.022709</td>\n",
       "      <td>0.025003</td>\n",
       "      <td>0.021120</td>\n",
       "      <td>0.026571</td>\n",
       "      <td>0.039011</td>\n",
       "      <td>0.019884</td>\n",
       "      <td>0.023497</td>\n",
       "      <td>0.003027</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CONSERVATIVE_PROTESTANT</th>\n",
       "      <td>0.012794</td>\n",
       "      <td>0.015424</td>\n",
       "      <td>0.026241</td>\n",
       "      <td>0.023881</td>\n",
       "      <td>0.016729</td>\n",
       "      <td>0.036784</td>\n",
       "      <td>0.020442</td>\n",
       "      <td>0.028994</td>\n",
       "      <td>0.022661</td>\n",
       "      <td>0.009017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BIBLICAL_LITERALISM</th>\n",
       "      <td>0.011482</td>\n",
       "      <td>0.015259</td>\n",
       "      <td>0.011644</td>\n",
       "      <td>0.014410</td>\n",
       "      <td>0.043189</td>\n",
       "      <td>0.015269</td>\n",
       "      <td>0.015363</td>\n",
       "      <td>0.019994</td>\n",
       "      <td>0.018326</td>\n",
       "      <td>0.005448</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OTHER_REGION</th>\n",
       "      <td>0.019604</td>\n",
       "      <td>0.013552</td>\n",
       "      <td>0.015503</td>\n",
       "      <td>0.022976</td>\n",
       "      <td>0.025850</td>\n",
       "      <td>0.012179</td>\n",
       "      <td>0.021115</td>\n",
       "      <td>0.014973</td>\n",
       "      <td>0.018219</td>\n",
       "      <td>0.003602</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOUTH</th>\n",
       "      <td>0.018210</td>\n",
       "      <td>0.011800</td>\n",
       "      <td>0.013987</td>\n",
       "      <td>0.022334</td>\n",
       "      <td>0.028087</td>\n",
       "      <td>0.011717</td>\n",
       "      <td>0.017965</td>\n",
       "      <td>0.017662</td>\n",
       "      <td>0.017720</td>\n",
       "      <td>0.004710</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BIBLE_HUMAN_ERROR</th>\n",
       "      <td>0.013423</td>\n",
       "      <td>0.010888</td>\n",
       "      <td>0.014936</td>\n",
       "      <td>0.013156</td>\n",
       "      <td>0.033724</td>\n",
       "      <td>0.016266</td>\n",
       "      <td>0.016139</td>\n",
       "      <td>0.019291</td>\n",
       "      <td>0.017228</td>\n",
       "      <td>0.010254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOME_COLLEGE</th>\n",
       "      <td>0.012272</td>\n",
       "      <td>0.013195</td>\n",
       "      <td>0.022762</td>\n",
       "      <td>0.019533</td>\n",
       "      <td>0.013968</td>\n",
       "      <td>0.014136</td>\n",
       "      <td>0.021070</td>\n",
       "      <td>0.018440</td>\n",
       "      <td>0.016922</td>\n",
       "      <td>0.007208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MARRIED</th>\n",
       "      <td>0.016179</td>\n",
       "      <td>0.014552</td>\n",
       "      <td>0.015368</td>\n",
       "      <td>0.017674</td>\n",
       "      <td>0.013353</td>\n",
       "      <td>0.014729</td>\n",
       "      <td>0.016895</td>\n",
       "      <td>0.016855</td>\n",
       "      <td>0.015701</td>\n",
       "      <td>0.005335</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NOT_PARTNERED</th>\n",
       "      <td>0.015657</td>\n",
       "      <td>0.015049</td>\n",
       "      <td>0.015287</td>\n",
       "      <td>0.017478</td>\n",
       "      <td>0.012000</td>\n",
       "      <td>0.015608</td>\n",
       "      <td>0.017215</td>\n",
       "      <td>0.016587</td>\n",
       "      <td>0.015610</td>\n",
       "      <td>0.004431</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OTHER_RELIGION</th>\n",
       "      <td>0.018120</td>\n",
       "      <td>0.019386</td>\n",
       "      <td>0.026216</td>\n",
       "      <td>0.017057</td>\n",
       "      <td>0.010052</td>\n",
       "      <td>0.008699</td>\n",
       "      <td>0.011048</td>\n",
       "      <td>0.013191</td>\n",
       "      <td>0.015471</td>\n",
       "      <td>0.006042</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         BAN_SEMI_AUTO  BAN_AMMO_CLIPS  BAN_HANDGUNS  \\\n",
       "POLITICAL_IDEOLOGY            0.205330        0.189869      0.152070   \n",
       "AGE                           0.109869        0.140372      0.102683   \n",
       "SUPERNATURAL_EVIL             0.043089        0.045292      0.066669   \n",
       "INCOME                        0.045523        0.047309      0.070199   \n",
       "ATTENDANCE                    0.045266        0.049327      0.056277   \n",
       "FEMALE                        0.090581        0.086458      0.018438   \n",
       "MALE                          0.091956        0.084623      0.017975   \n",
       "WHITE                         0.016707        0.015227      0.074660   \n",
       "POSTGRADUATE                  0.023433        0.020622      0.021917   \n",
       "RURAL                         0.021395        0.015949      0.023733   \n",
       "URBAN                         0.015785        0.017894      0.022709   \n",
       "CONSERVATIVE_PROTESTANT       0.012794        0.015424      0.026241   \n",
       "BIBLICAL_LITERALISM           0.011482        0.015259      0.011644   \n",
       "OTHER_REGION                  0.019604        0.013552      0.015503   \n",
       "SOUTH                         0.018210        0.011800      0.013987   \n",
       "BIBLE_HUMAN_ERROR             0.013423        0.010888      0.014936   \n",
       "SOME_COLLEGE                  0.012272        0.013195      0.022762   \n",
       "MARRIED                       0.016179        0.014552      0.015368   \n",
       "NOT_PARTNERED                 0.015657        0.015049      0.015287   \n",
       "OTHER_RELIGION                0.018120        0.019386      0.026216   \n",
       "\n",
       "                         SUPPORT_CONCEALED_CARRY  ARMED_SECURITY  \\\n",
       "POLITICAL_IDEOLOGY                      0.205110        0.132639   \n",
       "AGE                                     0.093378        0.095629   \n",
       "SUPERNATURAL_EVIL                       0.076422        0.131411   \n",
       "INCOME                                  0.048583        0.099614   \n",
       "ATTENDANCE                              0.049721        0.054600   \n",
       "FEMALE                                  0.044560        0.016042   \n",
       "MALE                                    0.046331        0.017434   \n",
       "WHITE                                   0.022770        0.012431   \n",
       "POSTGRADUATE                            0.035693        0.044597   \n",
       "RURAL                                   0.024030        0.016967   \n",
       "URBAN                                   0.025003        0.021120   \n",
       "CONSERVATIVE_PROTESTANT                 0.023881        0.016729   \n",
       "BIBLICAL_LITERALISM                     0.014410        0.043189   \n",
       "OTHER_REGION                            0.022976        0.025850   \n",
       "SOUTH                                   0.022334        0.028087   \n",
       "BIBLE_HUMAN_ERROR                       0.013156        0.033724   \n",
       "SOME_COLLEGE                            0.019533        0.013968   \n",
       "MARRIED                                 0.017674        0.013353   \n",
       "NOT_PARTNERED                           0.017478        0.012000   \n",
       "OTHER_RELIGION                          0.017057        0.010052   \n",
       "\n",
       "                         ARMED_TEACHERS  MORE_GUN_SAFETY  EXPANDED_SCREENING  \\\n",
       "POLITICAL_IDEOLOGY             0.213973         0.144500            0.164445   \n",
       "AGE                            0.108311         0.128774            0.135897   \n",
       "SUPERNATURAL_EVIL              0.056719         0.043802            0.064078   \n",
       "INCOME                         0.049029         0.072860            0.059765   \n",
       "ATTENDANCE                     0.055002         0.093758            0.053073   \n",
       "FEMALE                         0.053555         0.029534            0.037754   \n",
       "MALE                           0.046392         0.029946            0.035173   \n",
       "WHITE                          0.021826         0.014846            0.027538   \n",
       "POSTGRADUATE                   0.023973         0.015770            0.015794   \n",
       "RURAL                          0.029684         0.030772            0.026107   \n",
       "URBAN                          0.026571         0.039011            0.019884   \n",
       "CONSERVATIVE_PROTESTANT        0.036784         0.020442            0.028994   \n",
       "BIBLICAL_LITERALISM            0.015269         0.015363            0.019994   \n",
       "OTHER_REGION                   0.012179         0.021115            0.014973   \n",
       "SOUTH                          0.011717         0.017965            0.017662   \n",
       "BIBLE_HUMAN_ERROR              0.016266         0.016139            0.019291   \n",
       "SOME_COLLEGE                   0.014136         0.021070            0.018440   \n",
       "MARRIED                        0.014729         0.016895            0.016855   \n",
       "NOT_PARTNERED                  0.015608         0.017215            0.016587   \n",
       "OTHER_RELIGION                 0.008699         0.011048            0.013191   \n",
       "\n",
       "                             MEAN       STD  \n",
       "POLITICAL_IDEOLOGY       0.175992  0.027035  \n",
       "AGE                      0.114364  0.025219  \n",
       "SUPERNATURAL_EVIL        0.065935  0.015295  \n",
       "INCOME                   0.061610  0.014017  \n",
       "ATTENDANCE               0.057128  0.010688  \n",
       "FEMALE                   0.047115  0.009316  \n",
       "MALE                     0.046229  0.007061  \n",
       "WHITE                    0.025751  0.008054  \n",
       "POSTGRADUATE             0.025225  0.003162  \n",
       "RURAL                    0.023580  0.007292  \n",
       "URBAN                    0.023497  0.003027  \n",
       "CONSERVATIVE_PROTESTANT  0.022661  0.009017  \n",
       "BIBLICAL_LITERALISM      0.018326  0.005448  \n",
       "OTHER_REGION             0.018219  0.003602  \n",
       "SOUTH                    0.017720  0.004710  \n",
       "BIBLE_HUMAN_ERROR        0.017228  0.010254  \n",
       "SOME_COLLEGE             0.016922  0.007208  \n",
       "MARRIED                  0.015701  0.005335  \n",
       "NOT_PARTNERED            0.015610  0.004431  \n",
       "OTHER_RELIGION           0.015471  0.006042  "
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature_importances = pd.DataFrame()\n",
    "feature_importances.index = feature_cols\n",
    "\n",
    "for col in dependent_variable_cols:\n",
    "    bagging_forest_classifier.fit(q2[feature_cols], q2[col])\n",
    "    feature_importances[col] = np.mean(\n",
    "        [tree.feature_importances_ for tree in bagging_forest_classifier.estimators_], \n",
    "        axis=0)\n",
    "\n",
    "    \n",
    "feature_importances[\"MEAN\"] = feature_importances.mean(axis=1)\n",
    "feature_importances[\"STD\"] = np.std(\n",
    "    [tree.feature_importances_ for tree in bagging_forest_classifier.estimators_],\n",
    "    axis=0)\n",
    "top_feature_importances = feature_importances.sort_values(by=\"MEAN\", ascending=False).head(20)\n",
    "top_feature_importances"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4eeab9e5",
   "metadata": {},
   "source": [
    "To have a better look, we can plot the average feature importance for the 20 most important features:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "97d7bcb1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 3600x3000 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.style.use(\"bmh\")\n",
    "# setting the dpi for a higher resolution image\n",
    "plt.figure(figsize=(12, 10), dpi=300)\n",
    "plt.title(\"Feature importances\")\n",
    "pltdf = pd.DataFrame({\"Feature\": top_feature_importances.index, \n",
    "                      \"Importance\": top_feature_importances.MEAN})\n",
    "pltdf.plot.barh(ax = plt.gca(), \n",
    "               xerr=top_feature_importances.STD,\n",
    "               legend=None)\n",
    "plt.gca().invert_yaxis()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8604ee38",
   "metadata": {},
   "source": [
    "From the above graph, we can see that again, an individual's political ideology (importance $=17.6\\%$) (conservative/liberal) is a very important feature for predicting their attitude towards different gun policies. Political ideology therefore is a key feature for both our regression models and our decision trees.\n",
    "\n",
    "Age follows, with an average importance of $11.5\\%$. Age was also find to play an important role in the regression models (q3 & q4) but its effect was less intense compared to other parameters such as gender and belief in supernatural evil.\n",
    "\n",
    "Far behind, we can spot the SUPERNATURAL_EVIL, INCOME and ATTENDANCE parameters, while the FEMALE and MALE parameters are further behind. Compared to the regression models, the gender is less important. It also has lower importance compared to the superatural evil variables. **Belief in supernatural evil is far less important compared to the Political Ideology and Age parameters**.\n",
    "\n",
    "Finally, having a colege degree seems to be of little importance when predicting attitude towards firearm use, but having a postgraduate degree seems to have a more significant role, but still of minor importance compared to the aforementioned parameters.\n",
    "\n",
    "---\n",
    "To conclude, we can see that, although not always necessary, there were many similarities between the variables found to play an important role in our regression models and our random forest with bagging classifier parameters. \n",
    "\n",
    "Using two different approaches, we concluded that from the variables examined, an individual's political ideology is the stongest predictor of their attitude towards guns.\n",
    "\n",
    "Belief in supernatural evil, although not the most important factor, was found to also be a be a decent predictor of the attitude towards firerarms.\n",
    "\n",
    "Finally, the gender was not as important as described in the relevant [Economist article](https://www.economist.com/united-states/2021/11/06/belief-in-supernatural-evil-is-a-strong-predictor-of-pro-gun-beliefs)."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}