Understanding human behavior and reasoning is essential for developing successful instruction. Discipline-based education researchers have examined how students learn, informing the development of successful instructional strategies. Researchers have also identified barriers to the successful implementation of such strategies. This work utilizes two theoretical frameworks from psychology to further examine: 1) efforts to enact instructional change and 2) the effectiveness of instructional approaches to improve students' reasoning in physics. The Theory of Planned Behavior (TPB) is used to assess professional development supporting the successful implementation of evidence-based instructional strategies. The Dual Process Theories of Reasoning and Decision-making (DPToR) are used to model human reasoning and explain persistent inconsistencies in student responses. Guided by the TPB, an assessment instrument was created, validated, and implemented to evaluate instructor’s beliefs and intentions about active-learning methodologies. A semi-novel research methodology was also applied to address response-shift bias, a phenomenon common in professional development self-reported assessments. The validation of the instrument and the utility of the retrospective pretest methodology are reported, together with initial assessment results, demonstrating the value of both the TPB and the retrospective pretest in the context of professional development. The second half of this work discusses inconsistent student reasoning, where students correctly apply conceptual understanding in one context but fail to do so in similar situations. This phenomenon is examined using the Dual Process Theories of Reasoning, which describes reasoning in terms of two processes: a fast, automatic process 1 and a slow, resource-intensive process 2. Process 1 is quick but frequently inaccurate. Process 2 is analytical but time-consuming and effortful. Four reasoning hazards are identified and examined through the lens of DPToR. Three different types of interventions are implemented to help students develop skills to navigate reasoning hazards: 1) Collaborative exams are used to trigger socially-mediated-metacognition in a high-stakes environment, modeling process 2 activation through group reasoning, 2) a multi-stage guided individual intervention followed by a classroom discussion, and 3) explicit discussion of human reasoning modeled by DPToR. The impacts of these interventions are assessed by comparing results from the treatment (intervention) and controlled (alternative intervention) groups.