fix: NaN bug

[wasita]

discovered 1/29/23

maybe comes from situations where someone gives an amount between 0 and 1 and it then gets rendered incorrectly as a nan? will investigate

[ejolly]

Oooh yea Dhaval mentioned thats when he noticed it. Basically if the total endowment amount was too small, then they don't need to move the slider very much and he found it hard to actually choose a value of $1 instead of $0. So I think there's two issues we need to fix:

1. Making sure we handle cases where people actually choose a value of 0, i.e. making sure the rest of the math/conversion works out and doesn't give NaNs cause Javascript...
2. Making an adjustment to the scale such that regardless of the range of values they're choosing between, the distance they have to move the scale is similar. That'll make it non-problematic for trials where the endowment amount is small like Dhaval's issue. It also handles issues for adapting this for fMRI studies where we don't want spurious motor activity that's systematically correlated with choice behavior (i.e. higher/lower value choice -> more/less movement). We may need to turn this into a prop and play with it a bit to think through the best way to normalize tho.

[wasita]

hehe yeah that all makes sense and i totally agree!