I am performing a meta-analysis where my effects are mainly one-sample test against the chance level. Let's say that we have n subjects that perform several trials with a binary response (correct, wrong) and then I compute an accuracy index e.g. 0.7. All papers report a t-test against the chance level (e.g. 0.5) and then a cohen's d. I think that a more appropriate effect size measure could be a cohen's h or more generally a difference between two proportions, in this case one sample proportion and a population (chance) proportion. However, I'm not able to find a formula to compute the variance (for the inverse-variance weight) in this particular situation.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.