Finding effect size from a research paper I am writing a review paper under a prof this summer, and I'm desperate to demonstrate that I am competent, despite being as inexperienced and unqualified as possible. I am working on a compiling the results of each paper in a table, but I am unsure of how to list the results with numbers.
I can write "increased," but I don't know how to find out the actual numbers to list.
I am supposed to be calculating effect size (d) from a collection of research papers, but I am struggling to find the proper numbers to use. I can't find actual results in any of these papers in the form of raw data, everything is published as F values, or P values, or sometimes n^2. I know this may be a silly question, but I'm completely lost and in need of guidance! Thanks so so much in advance
 A: There are many types of effect sizes, which depend on the nature of the study outcome, the study design, the research question and the statistical test or model used to answer this question.
For example, if a study compares two groups - an experimental group and a control group - with respect to a continuous outcome value (say, blood pressure) and the two groups are independent of each other, then a suitable effect size for the study could be Cohen's d, which is a standardized difference between the observed group means.
Once you clarify what specific effect you are interested in, you may need to compute it yourself from information included in your papers (not all authors report effect sizes in their papers). To this end, you may need to scour the papers you have for the appropriate information.
For the example provided above, let's say you need to compute Cohen's d. How you will perform this computation will depend on whether or not the two groups being compared have equal sizes.
The website https://www.psychometrica.de/effect_size.html shows you that, if the group sizes are equal (e.g., each group includes 20 subjects), you can compute Cohen's d from the observed means and standard deviations of the outcome variable in each group, as well as the total size of the two groups. For example:
    Group 1            Group 2 

   Mean = 15          Mean = 10
     SD = 3           SD   = 4
   N1   = 20          N2   = 20

The total size is N = N1 + N2 = 20 + 20 = 40. Plugging these numbers in the appropriate boxes of the website will help you produce the desired effect size estimate.  This is an example where the effect size could be computed entirely from descriptive statistics, which are typically reported in one of the early tables in a paper (e.g., Table 1 or 2).
The website gives you the possibility to compute other types of effect sizes - you can play with the various scenarios to get a sense for what each scenario may involve.  Ultimately, knowing what effect size you should report should give you clues as to what ingredients you need to collect from your each of your papers in order to compute this effect size.  While some effect sizes rely on descriptive statistics, others require information collated from a statistical test (e.g., t-test, ANOVA) or a statistical model (e.g., linear regression, logistic regression). In any event, you should not have to compute effect sizes from raw data.
Don't be afraid to ask your professor for further guidance - research is a collaborative process which entails communication between the parties involved. When asking guidance, it helps to explain what you might consider doing and where you are not sure you are on the right track.
