Increased use of achievement scores to measure the impact of school principals on student outcomes has led to increased attention on how best to generate estimates of principal effects. This study joins others in the field in attempting to estimate how K-12 principals affect test score outcomes of students in their schools. Toward this end, this study accomplishes two goals. First, it characterizes principal movement throughout the Colorado public school system for a period of ten years. Second, the study uses grade-level data from Colorado public schools to estimate principal effects on student outcomes. With respect to the effects estimation, this dissertation is a replication of an existing study: Grissom, Kalogrides, & Loeb (2015). The principal mobility analysis suggests that while some school and principal characteristics are statistically significant predictors of principal cross-school transfer, models overall have low predictive power. That is, principal transfers between schools are difficult to predict using publicly- available data on schools or principal administrative data. Like the work being replicated, this study examines three conceptual approaches for identifying these effects, explains associated empirical strategies for estimation, and then compares the results across model specifications, subjects, principals, and schools. The study also compares estimates to external measures of principal quality currently in use in the state. The results of this study demonstrate that different models give different impressions of average principal effects. School effectiveness models, which attribute school effects to principals, estimate principal effects as 0.17 standard deviations for both reading and math. However, models attempting to estimate principal effects as school improvement over time yield estimates of 0.01 standard deviations for math and approach zero for reading. Sensitivity analyses demonstrate that differences in estimates across models are not driven by differences in the samples used. Like the study being replicated, external measures are more highly correlated with estimates from school effectiveness models than other approaches that separate principal and school effects. The results support those from earlier studies in that the choice of conceptual approach and thus, estimation strategy, drives differences in principal effects estimates. Moreover, models producing estimates for the largest number of principals over the shortest period of time are those that are unable to disentangle principal and school effects, while models applying to smaller numbers of principals requiring a greater number of observations are better able to address the differences between the two. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com.bibliotheek.ehb.be/en-US/products/dissertations/individuals.shtml.]