Visual presentation of "Check all that apply" question that saves as a dummy-coded variable

[Calvin Lai]

Hi,

I am currently trying to create a "Check all that apply" question that will be used by many researchers. It currently looks like this.

You'll notice that there are weird spaces between each of the options. That's the issue we're trying to troubleshoot.  We don't want that space to be there.

The space appears between the options because this is a series of "selectOne" questions with "" as the stem. We set it up that way because that was the only way we could think to get a bunch of dummy-coded variables like:

multi_asian: 1

multi_black: 1

multi_hispanic: 1

multi_white: 0

multi_middleeastern: 0

...and so on.

Using the default selectMulti option would solve the visual presentation problem. However a multiracial person who is White, Asian, and Black would show up as a string variable that looks like ["Asian", "Black", "Hispanic"] in the recorded data. That would require some intermediate-level data-wrangling to parse the string for the information that a researcher needs. We anticipate that novice or early PhD students would face special trouble since they are typically not experienced with cleaning string data.

Is there a way to present this series of dummy-coded questions without space in between? Thanks!

For reference, here's our current code for one of those dummy-coded questions:

API.addQuestionsSet('singleChoice',{

inherit: 'basicQ',

type: 'selectOne', 

help: '<%= pagesMeta.number < 10 %>'

});

API.addQuestionsSet('multi_pacificislander',{

    inherit: 'singleChoice',

    booleanValues: true,

    name: 'multi_pacificislander',

    stem:"",

    answers: [

{text:'Native Hawaiian or Other Pacific Islander'}

]

});

Best,

Calvin

[Calvin Lai]

Ah, I see the image didn't show up properly. Here it is, pasted again:

[Mayan Navon]

Hi Calvin,

I still can't see the image..

Anyway, maybe it would be easier for students to process the data with numeric values? e.g.,  White = 1, Asian = 2, Black = 3, so the data for a multiracial White & Asian would be an array with the numbers 1,2.

If this doesn't work for you, could you please provide the url for the study, so I can see the problem and check all the relevant code?

Thanks!

Mayan

ב-יום רביעי, 28 באוקטובר 2020 בשעה 23:41:27 UTC+2, ckl...@gmail.com כתב/ה:

[Calvin Lai]

Ah, that's weird. I actually ended up sending an external link to the image and it seems to have been deleted.

Here's the relevant study URL: https://app-prod-03.implicit.harvard.edu/implicit/common/en-us/html/showfiles.jsp?user=ksaadatian&study=researchsitedemo

I don't think the numeric values approach would work unfortunately since the data would still be presented in a string like "[1, 2, 4, 6]". Creating the dummy variables by default continues to be the easiest approach from the researcher perspective.

Thanks!

Calvin

[Mayan Navon]

You can use stemCss: { 'display': 'none' }, on each 'multi'  question. For example:

    API.addQuestionsSet('multi_pacificislander',{
        inherit: 'singleChoice',
        booleanValues: true,
        name: 'multi_pacificislander',
        stem:""

,
        stemCss: { 'display': 'none' }, 
        answers: [
            {text:'Native Hawaiian or Other Pacific Islander'}
        ]
    });

​

This will hide the stem element, so the space will be hidden.

Here is how it looks:

Mayan