I am currently conducting a research study comparing the long-term outcomes of a particular surgical procedure. I have pre-, 6 months, and 5-year postoperative AOFAS, VAS, and FFI scores for 23 patients. I also have 6-month and 5-year Karlsson-Peterson scores as well. What is the appropriate statistical analysis to be done on this data? I am new to research so any and all help would be greatly appreciated.
Respectfully
Hi Charles
You will probably need to use linear mixed effects models, since you have repeated measures. Best you try and find a local biostatistician to assist.
Adrian
I recommend you read about network analysis because it allows you to evaluate and compare long-term data as an evolutionary process and even determine the effectiveness of medical procedure.
The development of social network analysis
L Freeman - A Study in the Sociology of Science, 2004
Network analysis of 2-mode data
SP Borgatti, MG Everett - Social networks, 1997 - Elsevier
Network analysis of 2-mode data
SP Borgatti, MG Everett - Social networks, 1997 - Elsevier
Network analysis: methodological foundations
U Brandes - 2005
If your data are clearly not normaly distributed, you can use Friedman (aligned) ranks tests. But if they are normaly distributed, you can use repeated measures ANOVAs. See R script in attachment. This script includes Post Hoc pairwise comparisons between the Pre, 6 months and 5 years measures. (I fabricated data, e.g. for FFI. These most probably do not look like FFI scores. But, you can easily replace my fabricated data with your observed scores.)
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis