Even I might be able to figure this one out

I haven't read the whole article yet, but this very interesting refereed journal article by Dr. Loren Marks of LSU notes that when the APA released their meta-analysis of 59 studies claiming that the children of homosexual parents did not differ statistically from the children of heterosexual parents, 26 of those studies did not have a heterosexual comparison group or "control," and none of the studies had a sufficient sample size to find small differences.  Also worth noting is the prevalence of single parents as a control group--13 of 33 studies--when the research does clearly indicate degraded outcomes for the children of single parents, and the fact that in the remaining 20 studies, important data about the control group (marital status, income, race, etc..) is not listed.  In short, it would appear that in none of the studies cited by the APA was an actual, valid statistical test performed.

Now no matter where you fall on issues regarding homosexuality, this ought to be of concern.  More and more often, it appears that researchers are simply not following basic procedures in their research, and their colleagues are not calling them out on this.  Like in this case.  And since more and more public policy is being decided on the basis of studies like these, we all have a lot to lose.