Hello everyone!
When calculating meta-analysis in R, using package metafor, and computing the -|lnRR| (that is, the negative of the absolute value of the natural logarithm of the response ratio) to compare the significance of effects with 0 for different treatments, I am getting some instances of positive results when I was expecting only negative values.
Has anyone come across something like this?
Thank you in advance!
If you are quantifying effect as RR, an RR of 1 corresponds to the nul effect (ie lnRR = ln 1 = 0).
Responses of treatment < control (RR< 1) will have ln RR negative. I'm not sure why you are using mod ln RR which will alawys be positive and hence - mod ln RR will always be negative.. There must be an error somewhere if you are getting positive values for the negative of a modulus.
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Cell Biology
- Culture Cells
- Cell Line