Percentage into d or r?

#1
Is it permissible to calculate an effect size from a percentage? If so, how might this be done?

For example, if 87% of participants experienced a change in N=30, is it possible to get an effect size number from that?

Thanks! :):confused:
 

Dason

Ambassador to the humans
#3
Yeah it's not entirely clear what you're asking.

We can put a confidence interval around the true proportion for the population (assuming the study in question meets a few assumptions). Is that what you mean?
 
#4
Sorry for the lack of clarity. I don't have this all figured out so it is even hard to ask the right questions. For a metaanalysis -- in examining different articles some articles report pre-test posttest data, others report their overall data in different ways. For example, in one study all that seems to be reported in % data. I need to convert that % data into a standardized score. Is it possible? If it is possible, what is the minimal data I need to convert the score and how might that be done.