general, the standard errors for subsample IV estimates are larger when the effect size is larger, and standard errors decrease as nX increases. Standard errors converge to the full analysis standard error of 0.063 for all effect sizes as nX approaches nY. Using a 2-sample IV approach, results are very similar (see Web Figure 1, available at http://aje.oxfordjournals.org/). Power for the 2-sample approach appears to be very slightly lower than that for the subsample approach, and standard errors do not converge to a common value as nX approaches nY.
