Advice on coding SPSS Mixed

[J P-6]

Hello,
 
I have longitudinal data spanning 6 time periods (waves). For times 1-3 neither group experience the intervention. Prior to time 4 one group was given the intervention (training) and three more waves of data were collected. The goal of the analysis is to determine if the training had an impact on some measures of workplace climate.
 
What is the appropriate method for testing the effect of the training? The interaction between time and intervention, as usual? Or the main effect of the intervention alone?

In the data file Intervention for the comparison group is coded 0 for all waves, for the intervention it is coded 0,0,0,1,1,1.

Here is my SPSS syntax.

MIXED
S1 WITH Int_var BY wave
/FIXED =intercept Int_var2 wave Int_var*wave | SSTYPE(3)
/METHOD = ML
/PRINT = R SOLUTION TESTCOV
/repeated = wave | SUBJECT(HRID) COVTYPE(UN) . 
Advice is greatly appreciated.

Thank you,
John

[MaxJasper]

 

I have longitudinal data spanning 6 time periods (waves). For times 1-3 neither group experience the intervention. Prior to time 4 one group was
[mj] [....] Advice is greatly appreciated.

[mj] data file is greatly appreciated for analysis.

[J P-6]

Cannot send data file. But here is a simple example using four waves instead of 6.

 

ID     wave    score    intervention

1        1            3            0

1        2            4            0

1        3            5            0

1        4            6            0

2        1            3            0

2        2            4            0

2        3            5            1

2        4            6            1

          

Person 2 (ID 2) is in intervention group this intervention is coded 0 for first two waves, before the intervention.

 

 

 

---

From: MaxJasper <[hidden email]>
To: [hidden email]
Sent: Sat, February 27, 2010 2:02:00 PM
Subject: RE: Advice on coding SPSS Mixed

 

I have longitudinal data spanning 6 time periods (waves). For times 1-3 neither group experience the intervention. Prior to time 4 one group was
[mj] [....] Advice is greatly appreciated.

[mj] data file is greatly appreciated for analysis.