Transforming correlations and frequencies into Hedges g

Hi everyone,

I'm new to the R world and would very much appreciate your help!
I am trying to conduct a meta-analysis using the metafor-package. This is going well so far and I managed to add a new column to my data called "Hedges_g" which now contains all the hedges_g's I was able to compute via the escalc-function for the studies that provided means and standard deviations.
This is what I did:
data_hedge <- escalc(measure="SMD", n1i = N_Men, n2i = N_Women, m1i = Mean_Male, m2i = Mean_Female, sd1i = SD_Male, sd2i = SD_Female, data=data, var.names=c("Hedges_g", "vi"))

Now, there is still a lot of studies that only mentioned correlations or frequencies (in %) and did not measure my dependent variable as a continuous one, but as a binary. I need to include these studies, too, and therefore have to transform the correlations and frequencies into Hedges g as well and add them to the column I created before.

I looked around the internet and wasn't able to find anything that helped me with this.
Does anyone of you know any functions I could apply?

Thanks a lot in advance!!

Under the assumption all you need is the count, mean, and standard deviation to plug into your calculation here is a possible step forward.

Just a general comment/reminder. For frequency data if you know n (the total count of items) and p_class (the frequency of items in the class you care about) then you have mean = n * p_class and population variance is n * p_class * (1 - p_class). So one could try sqrt(n * p_class * (1 - p_class)) as a (biased) estimate of population standard-deviation (ref ). I guess you could also try to sample adjust it by using sqrt(n * p_class * (1 - p_class) * n / (n-1)). Hopefully some source or reference has what is considered an acceptable conversion.