In this paper, the author discusses the role of methodology in assessing the effects of teacher qualifications on teacher effectiveness (also commonly referred to as the "quality of teaching"). The author focuses on some of the statistical problems that can arise in a non-experimental setting and the ways in which statisticians attempt to deal with them. Then, using data on teachers and students from a statewide administrative dataset from North Carolina, he shows how estimates of the impact of various teacher qualifications change depending on the type of statistical methodology applied to the data. The next section briefly reviews the literature on teacher effects and examines how one might think about findings on the impact of teacher qualifications on student achievement in light of the statistical issues can arise. (Contains 3 tables and 25 endnotes.)