In my opinion, you have to construct the complete questionnaire with the four constructs and you submit it to a pilot test with participants and with those results you can calculate Conbrach's alpha according to the scores obtained in each question of each construct.

Hello Kharlo,

Short answer: you would collect data from a sample of participants (using all created questions) and use this data to calculate Cronbach's alpha separately for each construct. That is, you will compute Cronbach's alpha four times - once for each construct.

Long answer: Four questions can be enough to create a reliable and valid measure of a construct. However, when creating a new measure, researchers usually start with a large number of questions for each construct, collect data using all the questions, and then, using a technique such as EFA, reduce the number of items (i.e., removing 'bad' items with low factor loadings and keeping 'good' items with high factor loadings). The challenge is you want to keep no more items than are necessary, but also enough questions to adequately measure the construct you're interested in (see Content validity). Thus, it is difficult to know whether four questions is sufficient to adequately measure a particular construct without first subjecting the items to a statistical analysis. Cronbach's alpha coefficient can help us determine whether our questions are reliable, but not necessarily how valid they are.

I wish I had a good paper on hand detailing the use of Exploratory/Confirmatory Factor Analysis for item reduction. Hopefully another commenter here has a readily available resource on this topic.

-Blaine

Dear Mr. Kharlo Subrio, statistically, you could compute Cronbach's alpha with only one variable. I agree 100% with Mr. Blaine Tomkins about content validity [the main point of your question] and the applications of EFA and CFA.

Cronbach's alpha requires unidimensionality in construct, so you compute for each latent variable. I agree 100% with Mr. Blaine Tomkins again. I suggest the command alpha in Stata (https://www.stata.com/manuals13/mvalpha.pdf), e,g, alpha mot1 mot2 mot3 mot4.

There are a lot of validated scales that measure motivation, like SIMS/EME/EMIET. But, I do not know the scales about Skills, Tasks, Practices, but I think that you can find it [Google Scholar, WoS], or maybe some researcher in this area could suggest here.

My suggestions about Content Validity:

- Haynes, S. N., Richard, D., & Kubany, E. S. (1995). Content validity in psychological assessment: A functional approach to concepts and methods. Psychological assessment, 7(3), 238.

- Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory. 3rd ed. New York [etc].: McGraw-Hill. [Seminal book in my opinion]

My suggestions for readings in motivation scales:

- Vallerand, R. J., Pelletier, L. G., Blais, M. R., Briere, N. M., Senecal, C., & Vallieres, E. F. (1992). The Academic Motivation Scale: A measure of intrinsic, extrinsic, and motivation in education. Educational and psychological measurement, 52(4), 1003-1017.

- Guay, F., Vallerand, R. J., & Blanchard, C. (2000). On the assessment of situational intrinsic and extrinsic motivation: The Situational Motivation Scale (SIMS). Motivation and Emotion, 24(3), 175-213.

Tremblay, M. A., Blanchard, C. M., Taylor, S., Pelletier, L. G., & Villeneuve, M. (2009). - Work Extrinsic and Intrinsic Motivation Scale: Its value for organizational psychology research. Canadian Journal of Behavioural Science/Revue canadienne des sciences du comportement, 41(4), 213.