Validation of a questionnaire

scale-validation

#1

Dear all,
maybe its a very stupid question but I am very desperate in the moment because nothing works… :frowning: but maybe you can help me :slight_smile:

i developed a questionnaire and now I want to prove the convergent validity. Therefore, I used an other questionaire. The people had to fill out my questionnaire and the other questionnaire. Now I want to check the convergent validity - does my questionnaire measures the same like the other one. For my understanding I need to get the correlation of both? Both questionnaire were rated on a 7 scale answer format.
Can you help me, which function do I need?

Thanks!!! :slight_smile:


#2

Hi Katharina,

Yes you are on the right trail. If you have two measures that are theoretically or designed to measure the same construct, you would estimate a correlation coefficient between the two scales. If each questionnaire has multiple items, you might add up the items and correlate the total scores. The justification for adding them up would be supported by computing a Cronbach alpha internal consistency coefficient.

These kinds of operations can be very simple. If you have your data in an Excel spreadsheet, you can compute sums and estimate correlations right within Excel. I don’t think Excel will give you an alpha. So if you are limited to Excel, forget it, or take the plunge into R.

This is just the beginning of a big wide and important world of measurement development and validation. It can get complicated quickly. Here are some recommendations on readings to get started

Hand DJ. (2017) Measurement: a very short introduction. NY: OUP.

Anastasi, A. (1976). Psychological Testing (4 ed.). New York: MacMillan Publishing Co., Inc.

Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York: McGraw-Hill College Division.

Good Luck!