Open yhoriuchi opened 1 year ago
I now think that this is one of the top priorities. Currently, users need to specify both attributes and levels for a profile choice and a profile not-chosen. Too many things to set. The function will be simpler and allow users to specify any combination of attribute-levels (using any number of AND and OR) for choice-level analysis. What do you think, @thegaryking and @aaronrkaufman ?
I am not sure I understand this, but I'd like to talk through it!
Consider that you want to compare:
for non-white Republican respondents. In this case, you need to use two attributes (race, party). You also need to use "Asian OR Hispanic OR Black).
Currently, we cannot calculate the choice-level MMs, which involve more than one attribute.
Eventually, for choice-level analysis, we should allow users to specify any combination of attribute levels using AND and OR. For example, the arguments should be something like the following: