I need to calculate cronbach's alpha, two-way mixed ICC and corrected item total correlation. I have 2 raters rating 15 patients using 25 questions (value for each answer is between 1 and 5). The 25 questions can be grouped into 5 categories of 5 questions each. So far, I have made two columns on SPSS for rater 1 and rater 2, and the rows are the 15 patients. Each cell is the mean value of 5 questions in each category, so I have 5 datasheets. I used analyse>scale>reliability analysis and found the cronbachs alpha and average ICC. My question is if this calculation is right and how I would go about finding the corrected item total correlation for each question.