I would like to ask if someone could help me to convert unit from OR per SD to 1mm for meta analysis.
for example The age-adjusted OR per SD cIMT for MMSE
I do not understand your abbreviations. I would normally assume OR is "odds ratio" and SD is "standard deviation" but that doesn't make since because OR is calculated on binary data and SD on continuous data. I do not know what cIMT is and I assume MMSE is the Mini-Mental State Exam.
Assuming you are really talking about beta logistic regression coefficients, then if beta (multiplicative increase in RR (OR) per unit change in cIMT which I assume is a continuous variable) then multiplicative change in RR(OR) for SD units would be exp (SDx beta) = exp(beta) SD so multiplicative increase in RR per unit cMIT = (multiplicative increase in RR per 1SD) 1/SD
you would have to know the SDs (of cMIT) for the individual studies.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis