In most cases, we may obtain odds ratios or relative risks from the given beta coefficients if the DV is categorical to estimate the pooled effect sizes.
What if the DV is continuous and report beta coefficients with corresponding 95% CI or p-values?
Are the focal explanatory variables dichotomous treatment indicator (or dummy) variables? Or are they quantitative? If the former, this chapter may give you some ideas:
- https://mason.gmu.edu/~dwilsonb/downloads/esformulas.pdf
Pay attention to sections 3.7 and 3.14. HTH.
For a continuous independent variable, the logistic regression coefficient represents the multiplicative increase in RR per unit of the independent variable.
one way to get an overall effect would be use the mean level of "high" and mean level of "low" and calculate the ratio of the corresponding relative risks
I use psychometrica for converting effect sizes into a desired method. Check it once, they have extremely good website and conversion will be in seconds
- Badges
- Science topic
- Similar topics
- Correction