I need to ask you for help on what method to use to calculate the reliability and validity of a scale that I measured repeatedly on the same subjects during an experiment.
Let me explain: normally when you want to measure the reliability of a construct you do an EFA, a CFA, Cronbach's alpha, composite reliability, AVE and the square root of the AVE. Obviously, what has just been said is a simplification, but I want to be clear. However, this procedure can be used when scales are only administered once on a sample (and not more than once).
My situation is instead as follows: I administered a familiarity scale consisting of 4 items on 6 brands (familiarity means to what extent a subject knows each brand). Therefore, the same scale was administered 6 times. Of these 6 brands, 3 were chosen because they were unfamiliar (i.e. low familiarity), while the other 3 were chosen because they were well known/familiar (i.e. high familiarity). A Student's t-test confirmed that these two groups of brands (high and low familiarity) are statistically different. I ask you: what kind of reliability test (meaning internal consistency of the construct) should I use to measure the reliability of the familiarity scale? In fact, this scale was measured 6 times, i.e. 4 (items) x 6 (brands) for a total of 24 items.
Thank you for your attention.
