I am looking for examples or suggestions how to use the R package metafor to conduct a meta-analysis if no descriptive statistics are available, only Cohen's ds and Cohen's dz, alongside the sample size. A few papers relevant report sufficient descriptive statistics so that I can compute Cohen's d or dz, but many unfortunately only report the t-value and sample size, or directly Cohen's d or dz etc. Ideally, I would 'feed' a function with the effect and sample sizes and the function would compute the variance of the effect size, the summary statistics and would allow me to create a forest plot. I am aware that this is possible with the formulas provided in Borenstein et al's (2009) book and rmeta (for the forestplot), but I want to familiarise myself with metafor.

The examples I found assume the ideal case that the authors report all relevant descriptive statistics, which happens unfortunately not that often.

  • $\begingroup$ I would have thought back-calculating the standardised mean difference from t and df or t and p was just a line of code (and actually questions asking for code are off-topic here). Have I misunderstood? $\endgroup$ – mdewey Nov 29 '18 at 14:38
  • 1
    $\begingroup$ This might be helpful: metafor-project.org/doku.php/tips:assembling_data_smd $\endgroup$ – Wolfgang Nov 30 '18 at 8:49

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.