Running log of DS ideas

[schanzer]

• add a find-by-id function that consumes a table and a value, and produces the first row where the first column matches the value. Introduce it in contracts-tables-displays to address Rows early, and replace row definitions in all starter files with the use of find-by-id
• Page 70 of fall2024 book ("Identifying Form and r-Values") should use a format more similar to page 76 ("Drawing Predictors")
• Contracts, Tables, Displays: introduce COE with numbers first
• add section or worksheet on "variability" to "visualizing the shape of the data"  #2085
• Add categorical column to US Income dataset
• Add reducers in some form
• Add a question to DR telephone pages in defining table functions that has students come up with their own testing tables for the functions.
• Add additional forms of data to Intro to DS lesson: Consider enhancing the initial question-asking activity with having students look at data visualizations or conclusions (news headlines?) to generate discussion on additional information and relationships, with the goal of reaching NGSS-HS-SEP1-2: Ask questions that arise from examining models or a theory, to clarify and/or seek additional information and relationships.
• show datasets with the same range but different IQRs to emphasize the need for more sophisticated measures of spread than just range.
• Reach out to [email protected]. re permission to use this image
• Data Cycle -
• Incorporate: "Here is the most common wrong answer from assessments. Ask students to write a question where it is the right answer."

OVERALL: I think this lesson needs to be refocused. The activities in this lesson well exceed the times given and in my opinion the focus on the question types detract from the Data Cycle. Another way to include those back in and for them to be actually helpful to students, would be to provide sentence stems for each type of question. That way there's less ambiguity and more structure for how to ask the question types (if students still feel that they are arbitrary). If that is provided elsewhere, I don't believe it is linked anywhere on the lesson plan.

• Standard Deviation - lots of heavy assertions here need @schanzer to weigh in
• Probability Inference Sample - consider rewriting the How to Spot a Scam section, which has been flagged as
  quite high level and unapproachable, especially if students have minimal probability background.
• edc sj data modules

Done

• research project slide deck improvements #1590
• Add a context for the playdough lesson that will justify ordering the bars in the histogram #1357
• Consider making a video of kinesthetic activities like play dough box plot #1349
• Consider whether we want to autotitle charts.
• Add to line-plots to exploring plotting lesson and contracts page. Actually explain that they are "used when the explanatory variable is time" in scatterplots, correlations and linear-regression
• The ethics, privacy and bias lesson should provide more rigorous discussion of those concepts!
• Include assessments (from Create assessments curr#441)
• Change image-scatter-plot lesson to "custom plots", make it more programming practice, and talk about Edward Tufte (add image- enhanced versions of histograms, bar- and pie-charts brownplt/code.pyret.org#456 and Use image-plot functions to replace/extend existing DS lessons #1145)
• Add stacked and multi bar charts #1217
• side-by-side bar charts: Show bar charts for fixed v. unfixed, old v. young ask "what might this mean?" (allows us to hit 8.SP.A.4)
• Change "questions and column descriptions" page to sentence stem: "I notice _____, so I wonder ____"
• Check in on Equity goals from Ideas from Equitable Math (equity) #456 and Recommendations from ASSESSING BIAS IN STANDARDS AND CURRICULAR MATERIALS  #629
• Rename random-samples appropriately (mentioned in audit lesson glossaries  #890)
• displaying-categorical-data should start with count of species, then move on to pie and bar charts (completed by new table-and-display-function lesson)
• Vet Launch activities for every lesson, and make sure they are substantive and engaging (from Stefanie Davison's feedback)
• Lying with Graphs (see shriram)
• Bugfix activity (from [DS] Bugfix activity curr#382)
• Make a bigger deal out of "what data? what variables? what display?"
• Specify tight core, with extensions for programming, stats, etc (from [DS] Specify tight core, with optional exercises for stats, CS, etc curr#403)
• Research paper rubric (from Joy)
• use f(x) notation instead of y= in Interpreting Regression Lines & r-Values
• in research paper, leave space for chart and the code that generates it

[schanzer]

Update from Joe: all of these are blocked, waiting on table-from-column to land as part of the Great Horizon Merge of 2020. If that happens in time for v1.5, we can do it. Otherwise, all of them may have to be punted.

[BootstrapJen]

Related to science-based extensions - One idea could be looking at the NGSS explicitly tied to Analyzing and Interpreting Data and determining if one or more of the topics would work well with Pyret.

https://www.nextgenscience.org/practices/analyzing-and-interpreting-data

[schanzer]

StandardDeviation.pptx

Nancy Pfenning has some ideas for standard deviation as well!