Hi there fellow researchers,
I have a very important question to ask regarding likert scales.
I have previously conducted an experiment that is aimed at comparing the difference between two groups (IV) with regards to three constructs (DV). Data is collected on a questionnaire in the form of likert scales (1 - 7).
So, i intend to measure the reliability, validity and then go on to test the differences using MANOVA. My question is do i need to compute or get the weighted scores of the Likert scales before i conduct my analysis? (Eg. Identifying multivariate outliers, etc)
Thanks in advance!