Hello,
How would using a reliable subscale (Cronbach's alpha =0.89) affect my research validity since I'm not using all the items of the valid scale?
Also how would this affect me publishing my dissertation?
I was told I can use it since it has its own reliability, and others said if I do, then I may not be able to publish it.
Thank you.
Did you collect data on all the items in the original published version of the scale? If so, why did you omit some of them from your current analysis?
Dear Amira,
Choosing to use only some of the items from a validated scale will affect its reliability rating. If your sample size is not large then you will not be able to recalculate this accurately.
Cronbach's alpha is only one way of measuring construct validity. There are a lot of common misconceptions about it.
Having re-read your question, it sounds like your subscale has been validated, in which case there is no problem using it. However, items included within a subscale may not load equally on the principal component. Therefore, a regression model from a principal component analysis is often more accurate.
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publisher