AbstractSeveral avenues are open to students who wish to study advanced science or mathematics in high school, which include Advanced Placement courses and teacher-designed courses unaffiliated with organized programs. We employ a retrospective cohort study of 4,691 na-tionally representative college students at 34 randomly selected, colleges and universities to examine the relationship between taking advanced high school courses and students' interest in pursuing a STEM career, while controlling for prior interests and experiences. We are able to distinguish between those students choosing to take an additional year in a science or math subject from those taking Advanced Placement (AP), which is most commonly taken as a second year course, but is increasingly taken as a first year course. We find that the number of years of a science or math subject taken in high school is associated with significant increases in STEM career interest, with results differing by subject. Taking AP courses in science or calculus appear to have no significant impact on STEM career interest over that of other advanced, non-AP courses. Taking calculus, a second year of chemistry, or one or two years of physics all predict large increases in STEM career interest. Additional years in biology and other subjects show no such relationship.Keywords: Advanced Placement, high school, career decisions, STEM pipelineIntroductionMaintaining a robust STEM (Science, Technology, Engineering, Mathematics) workforce is a matter with profound economic implications for the U.S. economy (Obama, 2009; National Research Council, 2005). Interest in a STEM career is often characterized as a pathway in which certain experiences serve to increase student interest in continuing on the STEM path (or joining the path), whereas other experiences reduce it (Blickenstaff, 2005; Kerr & Robinson Kurpius, 2004; Mitchell & Hoff, 2006; Riegle-Crumb, Moore & Ramos-Wada, 2011). While some students are exposed to STEM opportunities outside of school through their parents or other adults, or by engaging in informal science experiences, such as science museum visits or clubs (Dabney et al., 2012), their major exposure comes from school coursework (Jones, Howe & Rua, 2000; Hazari, Sadler & Tai, 2008).Advanced coursework in high school is an increasingly popular option for high school students. With reduced funding of dedicated programs for gifted students (Ward, 2005), many school systems opt for increasing students' opportunity to take advanced coursework. Three major outcomes are posited for students' taking advanced STEM coursework in high school:* gaining a head start on their college education with the possibility of reducing time to degree (National Research Council, 2002)',* making the learning in college STEM courses easier, owing to a stronger foundation (Federman, 2007; Sells, 1980, Sadler & Tai, 2007) , or easing the transition to college work from high school (Schwartz, Hazari & Sadler, 2008) ,* and increasing students' STEM interest and their persistence to a STEM career (Tyson, Lee, Borman & Hanson, 2007).2,3The rationale for this study is that while the first two major outcomes listed above have been studied extensively and reported on in the research literature, the impact of advanced coursework in high school on STEM interest and persistence has received less attention from researchers.The public primarily associates advanced high school coursework with the Advanced Placement program, which has expanded dramatically at a rate of 9.3% per year over the last two decades (Sadler, 2010). The popular US News & World Report's yearly ranking of the best U.S. high schools uses AP exam data. A succession of efforts to expand advanced science and mathematics offerings in high school call for training an additional 70,000 AP science and math teachers (National Research Council, 2005), federal and state subsidization of AP examination fees (Klopfenstein, 2004), and for offering cash incentives for students who pass AP exams and bonuses for teachers of students who do well on AP exams (Jackson, 2008). …
Stored only when both papers are in the corpus — 68 citations exist upstream.