I am writing my thesis on predicting response on ECT. My independent variables are the specific ECT variables (concordance, seizure length in seconds, postictal supression index etc.) and the HDRS questionnaire. All these variables are measured at two points in time, after the first and third ECT session. The dependent variable is categorical, with three categories: non-response, response and remission. This is only measured at the end of the ECT treatment. I use SPSS and was wondering how I can analyze the continuous independent variables with two measures on the categorical dependent variable with one measurement. Do I need to calculate the difference of the measurements of the independent variables, or can I use them both in an analysis? And what kind of analysis? I thought a multinominal regression analysis, but I am not sure.
Hi,
From my understanding, the dependent variable is on the ordinal scale on the HDRS scale. Ordinal regression analysis is possible if you are looking at a point in time. If over time changes in HDRS scores are measured then survival analysis with Cox's proportional Hazard model would be suitable.
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity