An important target of education policy is to improve overall teacher effectiveness using evidence-based policies. Randomized control trials (RCTs), which randomly assign study participants or groups of participants to treatment and control conditions, are not always practical or possible and observational studies using rigorous quasi-experimental or comparison group methods must frequently be used. Each of the three studies in this dissertation examines a facet of teacher effectiveness that can be used to inform policy while also contributing to scholarship in the area of causal inference from observational data. First, estimating or measuring individual-level teacher effectiveness using student value-added requires models that are robust to observational data. Second, teacher participation in a teaching intervention does not necessarily imply compliance with the treatment. The factors that interfere in compliance are non-random such that understanding the role of the hypothesized process in an RCT requires observational methods. Third, if observational methods such as propensity score analysis offer credible evidence of causality, they may be useful for examining efficacious interventions that are taken to scale in non-random settings. These issues place observational designs and methods at the center of efforts to study evidence-based teaching and teaching practices. [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/en-US/products/dissertations/individuals.shtml.]