Multiple Response and Chi-Squares

[Yawo Kokuvi]

Hello,

I am following up on my previous post on multiple response variables (available here: ) with a different question.

Here is a scratch of the dataset I am using:

* Example generated by -dataex-. To install: ssc install dataex
clear
input int id byte(Bad_ReduceSexualityLibido Bad_AgainstGodWill Bad_PhysicalPain Bad_HandicapDisability Bad_Abusive Bad_HealthInfections Bad_MentalHealth Bad_NonSpecificGenerally Bad_PatriarchyGender Bad_CulturalPractice)
1 1 0 1 0 0 0 0 0 0 0
2 0 0 0 0 0 1 0 0 0 0
3 0 0 1 0 0 1 0 0 0 0
4 0 1 1 0 0 0 0 1 1 0
5 0 0 0 0 0 0 0 1 0 0
6 0 0 0 0 0 1 0 0 0 0
7 1 0 0 0 0 1 0 0 0 0
8 0 0 0 0 0 0 0 1 0 0
9 0 1 0 0 0 0 0 0 0 0
10 0 0 0 0 0 0 0 1 0 0
end
label values Bad_ReduceSexualityLibido labels29
label def labels29 0 "No", modify
label def labels29 1 "Yes", modify
label values Bad_AgainstGodWill labels30
label def labels30 0 "No", modify
label def labels30 1 "Yes", modify
label values Bad_PhysicalPain labels31
label def labels31 0 "No", modify
label def labels31 1 "Yes", modify
label values Bad_HandicapDisability labels32
label def labels32 0 "No", modify
label values Bad_Abusive labels33
label def labels33 0 "No", modify
label values Bad_HealthInfections labels34
label def labels34 0 "No", modify
label def labels34 1 "Yes", modify
label values Bad_MentalHealth labels35
label def labels35 0 "No", modify
label values Bad_NonSpecificGenerally labels36
label def labels36 0 "No", modify
label def labels36 1 "Yes", modify
label values Bad_PatriarchyGender labels37
label values Bad_CulturalPractice labels37
label def labels37 0 "No", modify
label def labels37 1 "Yes", modify

The following command gives me the aggregated number of times a bad reason is given in the data

/* Number of times a Bad Reason is Given - regardless of multiple indicators */
.
. tabstat Bad_*, s(sum) c(s)

variable | sum
-------------+----------
Bad_Reduce~o | 77
Bad_Agains~l | 24
Bad_Physic~n | 95
Bad_Handic~y | 57
Bad_Abusive | 18
Bad_Health~s | 154
Bad_Mental~h | 22
Bad_NonSpe~y | 128
Bad_Patria~r | 45
Bad_Cultur~e | 15
------------------------
[/CODE]



Now, I want to find out how this aggregated variable differs by my predictors, like gender, race, etc, and then produce a chi-square to test the differences.

is this feasible on aggregated variable such as the version I have?

Thanks - cY

[Clyde Schechter]

First, it is not possible since your data apparently does not contain information about race, gender, etc.

Assuming you had a data set that contained these, the next question is why you would want to do this. It seems odd. How would it help you to know if there were more Bad_whatever responses among men than women without taking into account how many men and women there are in the data set. Maybe there are more Bad_whatever responses among, say, men, just because there are more men, but both men and women are equally prone to choosing that response. If there is some special situation you have where this reasoning does not apply and knowing the totals without the denominators makes sense, explain what you are trying to understand from this data, and perhaps there will be a way to do what you need.

[Yawo Kokuvi]

Clyde: thanks very much for your assistance. I am sorry I didn't include the race and gender variables in my model data. I will update it after considering the research questions, and hypothesis about any gender and racial differences. With much appreciation. cY