hi i would be grateful if someone could advise me on how to combine results from a repeated measures analysis. i have five imputed datasets, run the usual complete case statistics thru spss and try to combine through the norm program.
with between subjects the output from the glm procedure, i can get parameter estimates (in addititon to the f-test results) from spss, thus can get a combined result from norm. however, for the within subjects and interaction analyses while i can produce the f-test results for the five imputations, no parameter estimates are available for me to put through the norm program. i'm not sure whether there is a way to combine the f-statistic and sig values, or another way to produce parameter estimates from spss (possibly through syntax). the other statistics that are produced from spss, that (i think) are amenable to combining are the differences in estimated mariginal means. These are produced within the pairwise comparisions section of output, which has mean differences and standard errors for the means. this would allow me to discern the significant specific differences between various times of the repeated measures and across groups. unfortunately, i am not sure if this is a correct process (i.e. the right way to do things). it does not provide me with a combined f- statistic for the within subjects or interaction terms, nor does it acount for the multiple comparisions through post hoc tests. i'm stuck at what to do and as to whether combining only the post hoc results would be appropriate. i would be grateful for any help available. thanks in advance shash Shashivadan Hirani University College London Medical School Department of Psychiatry and Behavioural Science 2nd Floor, Wolfson Building 48 Riding House Street London W1N 8AA Tel: 020 7679 9309 Mobile: 07736 129648
