longitudinal dataset (vertical format): create tree diagram of change at timepoints

#1
Hello friends,

I have a very complex syntax question I believe may even stump the very best of you!

I have a longitudinal dataset measuring depression level on a 3 level scale (none, moderate, severe). The dataset is currently in vertical or 'stacked' format. I am to create a tree diagram of the changes at three time points.

For example, of those with severe depression at baseline, how many of those are severe, moderate and non-depressed at 3 months. Of each group at 3 months, how many are in each category at 6 months? I must do the same for all three categories of depression from baseline (see schematic attached for an example of what I mean).

Would DEFINITELY appreciate any help. As of now, am thinking I have to run MANUAL reports based on ID numbers, which will be very intense!

THANK YOU!!!!
 
#2
Re: longitudinal dataset (vertical format): create tree diagram of change at timepoin

Just in case anyone else needed a hand on this, I figured it out. The secret was I had to restructure the dataset back to horizontal format. Then it was easy....I could just split the cases by timepoint 1, and run simple frequencies for time 1 and time 2.