My study has 300 respondents which include 4 variables in which each subset has 4 to 7 questions. I want to ask from all of you that what happens when the Cronbach’s alpha for some subsets fall below 0.7. I want to increase the Cronbach’s alpha value, but I don’t want to reduce the number of questions. Please help me. Is there any other solution for this?
The Cronbach’s alpha below 0.7 is acceptable, stated in a very amusing paper by Lance (Lance CE, Butts MM, Michels LC. The Sources of Four Commonly Reported Cutoff Criteria: What Did They Really Say? Organizational Research Methods. 2006 Apr 1;9(2):202–20. )
I don’t think you should rely on Cronbach’s Alpha (CA), as many researchers rely more on the AVE and composite reliability values. Because Cronbach’s alpha doesn’t prioritise the indicators. But PLS prioritises the indicators according to their reliability.
Loewenthal, K. M. (2004). An introduction to psychological tests and scales (2 ed.); This reference talks about how the Cronbach alpha coefficient is acceptable. According to it, the alpha coefficient of 0.6 is also acceptable as it takes into account the number of items, constructs validity.
You can read from the following, for a better understanding:
Hair Jr, JF ; Hult, GTM ; Ringle, CM ; Sarstedt, M. A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM). Thousand Oaks, California : Sage Publications.
I don’t think reducing the number of questions is the right option. You should only reconstruct the questions after diagnosing the problem. Your internal coherence may be slipping because of the nature of your questions. Deleting may affect the validity and adequacy of your project.