dear s-plus users,
i use version 4.5 on windows 2000.
i would like to know the effect of a predictor which is not constant in
time - e.g. quality of the therapist-patient-alliance - on therapy
outcome but am not sure how to interpret lme coefficients. my data looks
like this:
CODE IN.THER SC10M ABM
s. id time.in.therapy (days) outcome alliance
989001 -64 0.56666667 2.36363636
989001 169 0.95555556 2.18181818
989001 298 0.75555556 2.18181818
989001 493 1.02222222 1.27272727
989001 693 1.03333333 1.13131313
989002 -31 1.86666667 1.63636364
989002 -7 1.14444444 0.63636364
989002 142 1.44444444 2.00000000
989002 519 0.54444444 2.27272727
|