Is it acceptable to use Cronbach's alpha to assess reliability for Yes/No questions?

#1
hi
Is it acceptable to use Cronbach's alpha to assess reliability for Yes/No questions?
if no, what should i do?!

there are 5 questions (with Likert-scale) and 6 questions (Y/N type)
these 11 questions measure a variable.
how can i assess reliability using SPSS?

thanks in advance !
 
Last edited:
#3
Re: Is it acceptable to use Cronbach's alpha to assess reliability for Yes/No questio

Have a look at this:

Raykov, T. (1997). Scale reliability, Cronbach's coefficient alpha, and violations of essential tau-equivalence with fixed congeneric components. Multivariate Behavioral Research, 32(4), 329-353.
 

spunky

Can't make spagetti
#5
Re: Is it acceptable to use Cronbach's alpha to assess reliability for Yes/No questio

safe?
emmmm
i am not sure what you mean ! :)
it means that (assuming the assumption of tau-equivalence is satisfied) there is no problem with using Cronbach's alpha on binary data. it just has a different name: the Kuder-Richardson 20 formula. but it's the same thing.
 
#6
Re: Is it acceptable to use Cronbach's alpha to assess reliability for Yes/No questio

it means that (assuming the assumption of tau-equivalence is satisfied) there is no problem with using Cronbach's alpha on binary data. it just has a different name: the Kuder-Richardson 20 formula. but it's the same thing.

got it

thank you :tup:
 

hlsmith

Less is more. Stay pure. Stay poor.
#7
Re: Is it acceptable to use Cronbach's alpha to assess reliability for Yes/No questio

You can also use it on a combination of binary and categorical questions, you just need to select the appropriate alpha (standardized or raw).