This is a mixed-methods research project to explore what has and has not changed in the learning experiences of undergraduates in science, technology, engineering, and mathematics (STEM) majors since the original study by Elaine Seymour and Nancy Hewitt, "Talking about Leaving: Why Undergraduates Leave the Sciences (TAL-1)," published in 1997. The study is focusing on the impact on student persistence of a variety of factors. This is also the first large scale investigation of STEM field-switching since TAL-1 across multiple institutions that are changing teaching practices and forms of student support in foundational STEM courses. This study is exploring whether and how problems with the learning experience found to be major contributors to switching decisions have been addressed at the original sample of seven institutions, and whether such changes enable more students to graduate in these disciplines. The seven institutions provide a useful baseline to gauge the extent of changes in the teaching and learning environment. These seven also represent the full range of types of higher education institutions providing most STEM undergraduate education (except for community colleges). Many factors affecting student decisions are being examined and independent measures of faculty teaching practices are being made using classroom visits. The project has three inter-related component studies: (1) a study of STEM fields-switching patterns based on national and institutional data, (2) a switcher and non-switcher interview study that replicates and extends the original research, and (3) a three-part study of instructors teaching foundational STEM courses using course observation, instructor interviews, and online surveys of all students in those courses. Ethnographic data will be gathered from three sources: student interviews, instructor interviews, and students' write-in responses to the online survey. Instructor interviews will explore the processes that both enable and constrain uptake of available evidence-based teaching methods and materials. Structured classroom observation and recording will be undertaken using the Teaching Dimensions Observational Protocol (TDOP), and a custom version of the Student Assessment of their Learning Gains (SALG) online instrument will be administered to all students in the foundational courses. Intellectual Merit: Among a number of analytical focal points, the study will produce multi-level regression analyses to assess how student demographic and academic characteristics, SALG responses, and course characteristics such as instructional methods predict the probability of students switching from their major later in the study. It will identify variables that predict switching, and the relative weight of individual predictors. These analyses will shed light on links between particular teaching methods used in foundational STEM courses and their persistence outcomes in students. Broader Impacts: The dissemination plan for the last year of the project includes preparation of research briefs, peer-reviewed and popular articles, and a book. These will examine issues of persistence in the context of educational change. Findings will be presented at the seven research sites to audiences that include deans, directors, and chairs of STEM departments, colleges, and graduate schools; to funding agencies interested in STEM reform; to current and prospective STEM doctoral students; to faculty advisors; and to scholars of STEM education.