I have results from the same regression on two datasets, namely t statistics and standard errors $t_1,t_2,SE_1, SE_2$, what method would be appropriate for a meta-analysis?
-
2$\begingroup$ Would you like to expand on this? What do you mean by saying you have values of t and their standard error? Do you mean you have the regression coefficients? And do you only have two data-sets to combine or was that just for illustration? $\endgroup$– mdeweyAug 15, 2016 at 17:52
1 Answer
I got the effect sizes from $\beta_i = t_i SE_i$, and followed these formulas: