Hi,

I'm using exploratory factor analysis on a 5 point Likert Scale. The problem I'm facing is that the questions in my questionnaire are less (4 questions), so after conducting the analysis, I end up with 2 questions only. To retain the factors, I am considering a cut off of 0.3 and 0.4 for factor loadings and communality respectively. Also, if I calculate the cronbach's alpha for the 2 questions I get, it is above 0.7 but if I consider all 4, the alpha is between 0.4 and 0.5. What should I do? I read somewhere that the construct validity is affected by deleting the items.

Also, since I'm doing an exploratory research, is it okay to use a chronbach's alpha value between 0.4 and 0.5?