Hi All, I am currently conducting a longitudinal study of change in marital trust. I am assessing change in trust as part of a larger model by regressing a later measure of this variable onto an earlier one. Unfortunately, about 45% of scores for the second measure of trust are missing. To deal with this problem, I am using the FIML option in Amos. This approach is probably justified, since the data appear to be missing at random. However, I'm unsure about how well the FIML option will perform when such a large amount of observations are missing for my second measure of trust. Does anyone out there know of some references that might address this issue?
Paul _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
