two-way repeat measure anova problem with varying n


New Member
I am completing a two-way repeat measure anova. I am trying to assess whether we see any month to month variation in our data which is a time course. so main effect 1: time after glucose addition (5 levels) and main effect 2 is month (7 levels).

Because the data is coming from separate studies comparing control with treatment the number of individuals within a month varies. So I don't have missing values as a random issue rather I have different number of individuals assayed for each month. To me it is an unbalanced design.

the problem: even when I syntax for missing values spss only reads the lines where it has an entry for all cells. so for one month I have three readings whilst for another I have 25 but spss only takes in the first three readings

Any advice appreciated as I am rather stuck. Is it a coding issue or a problem inherent with this stastistical approach?

GLM T0_mar08 T15_mar08 T30_mar08 T60_mar08 T120_mar08 T0_april08 T15_april08 T30_april08
T60_april08 T120_april08 T0_oct08 T15_oct08 T30_oct08 T60_oct08 T120_oct08 T0_dec08 T15_dec08
T30_dec08 T60_dec08 T120_dec08 T0_jan09 T15_jan09 T30_jan09 T60_jan09 T120_jan09 T0_feb09 T15_feb09
T30_feb09 T60_feb09 T120_feb09 T0_mar09 T15_mar09 T30_mar09 T60_mar09 T120_mar09 T0_apr09 T15_apr09
T30_apr09 T60_apr09 T120_apr09
/WSFACTOR=month 8 Polynomial assaytime 5 Polynomial
/PLOT=PROFILE(assaytime*month month*assaytime)
/WSDESIGN=month assaytime month*assaytime.
Last edited: