Hello, I am comparing two different groups for a study I have conducted. I was wondering, when calculating the internal consistency, do I calculate two separate alphas for each group or do I combine my sample and compute one total alpha?
How about if I'm using the multi-item scale for different sets of situations? Should I conduct a Cronbach Alpha for each set of situation? For example, I am measuring the attribute "complexity" and I have a six-item scale. I will use that scale to measure complexity of Product A (for a set of respondents who have used product A) and Product B (another set of respondents who have used product B).