As mentioned in 'How to use SPSS': Step-by-Step Guide to
Analysis and Interpretation' 4th Edition by Brian C. Cronk for calculation of Cronbach's Alpha assumptions are:
1. All the items in the scale should be measured on an interval or ratio scale.
2. In addition, each item should be normally distributed.
In a paper Fred; Avolio, Bruce J.; Avey, James B.; and Norman, Steven M., "Positive Psychological Capital: Measurement and Relationship with Performance and Satisfaction" (2007) while conducting CFA has mentioned that "Kolmogorov-Smirnov and Shapiro-Wilks tests of data normality were nonsignificant for each item". (Document attached for reference. See Results of CFA)
How do we check normality of each item?
Is establishing item normality necessary in construction and standardization of tools?
http://digitalcommons.unl.edu/cgi/viewcontent.cgi?article=1010&context=leadershipfacpub