Cronbach's coefficient alpha in questionnaire research

[habitusantos]

HI,

I intended to verify the effectiveness of a questionnaire in the SAS Guide through PROC CORR but I do not know how to start could give me some hint since this questionnaire has several answers that can go from scales from 0 to 5, such as answers of agree I do not agree .

Is attached a part of the questionnaire in an attempt to be more accurate in the question.

Thanks in advance

[Reeza]

Which portion do you need help with, methodology or coding? You first need to determine the methodology and then the code is pretty easy to develop.

Cronbach Alpha is via PROC FREQ, and using PROC CORR for data that's categorical can be problematic.

[habitusantos]

Hi Reeza,
Thanks for answering me . My urgent need it's essentially about the methodology because i have categorical data and it's not simple to convert it in numerical data.
Thanks in advance

[Reeza]

Recoding character to numeric is easy enough.

https://stats.idre.ucla.edu/sas/modules/creating-and-recoding-variables-in-sas/

If you need to do it for many:
https://stats.idre.ucla.edu/sas/seminars/sas-arrays/

Personally I'm a fan of formats, but that may be more complicated at the moment. You also need to show us what you need/want. Right now you've posted some data and a fairly vague question so if it's not specific enough you won't get answers.

I have a macro I use for survey/categorical data that may be helpful? Only if you have a single variable you want to analyze, otherwise PROC FREQ or SURVEYFREQ or TABULATE are what you want to be looking into.

https://gist.github.com/statgeek/0ff46bbd76bcc101e40945816b232303

[habitusantos]

Hi Reeza again,
thank you for helping me you have given me several clues to my problem but I think that the macro for the survey with categorical data would be very useful. Could you me pass? Thank you very much one more time

[Reeza]

It's the last link in the previous message (GitHub) above.

[habitusantos]

Thanks a lot for your help