Moore, Christopher2015-11-062015-11-062015-06https://hdl.handle.net/11299/175268University of Minnesota Ph.D. dissertation. June 2015. Major: Educational Psychology. Advisors: Ernest Davenport, Frances Lawrenz. 1 computer file (PDF); x, 150 pages.The Quality Compensation for Teachers (Q Comp) program provides up to $260 per student to Minnesota schools that adopt reforms to teacher pay and professional development. The reforms include an alternative salary schedule and observations of classroom instruction. Q Comp participants received about $419 million between 2006-2013, but voluntary participation and variability in implementation have made it challenging to evaluate Q Comp's overall impact on student achievement and identify its most effective reforms. This study applies spatial regression discontinuity (RD) and other quasi-experimental methods to estimate the effect of Q Comp participation and identify exemplars. Participation is estimated to significantly increase math and reading achievement by 0.0541 and 0.0247 standard deviation, respectively, compared to geographically neighboring districts that did not participate. The estimates are robust and diverge from a nonequivalent dependent variable. School distance from the Q Comp border is not a significant RD assignment variable. Five participating districts (Farmington, North St. Paul-Maplewood, Osseo, Spring Lake Park, and St. Francis) exhibited achievement gains that were significantly larger than expected, making them good candidates for qualitatively investigating which Q Comp reforms are most effective.enThesis or Dissertation