This study was designed to identify the variables that lead to the retention and attrition of talented men and women in science. It is an outgrowth of a longitudinal investigation conducted by the principal investigator with 1983 winners of the Westinghouse Science Talent Search (currently called the Intel Science Talent Search; Subotnik, Duschl, & Selmon, 1993; Subotnik & Steiner, 1993; 1994). The results of three earlier collection points have established patterns of external and internal influences that lead to participation in the research sciences, into applied sciences (e.g. medicine, engineering), or out of the field altogether.
The Intel (formerly Westinghouse) Science Talent Search is the most prestigious American high school science award. The men and women included in this study were among the semifinalists and finalists of the 1983 cohort. During their senior year in high school, they became winners by submitting a technical paper on an innovative and rigorous mathematics or science project conducted on their own or under the auspices of a mentor. A panel of judges made up of respected scientists reviewed the papers and recommended the winners to Science Service, the agency that conducts the competition.
The subjects (N = 85) are now 34 years old. A disproportionate representation of the women, one-third of the winner population, left science majors or training. This article describes the results of a fourth round of interviews with 1983 Westinghouse winners at a time when most of the cohort was deeply involved in career endeavors.
Review of the Literature
Poor-quality instruction at the undergraduate level can serve as an obstacle to retention in science and mathematics, and women have responded more negatively than men to lack of feedback for their course assignments or recognition of their individual accomplishments (Lewis, 1991; Seymour, 1992). Even if undergraduates successfully negotiate these introductory courses and experience working in an active research laboratory, the lifestyle of the research scientist can appear to be overly isolated, requiring superhuman dedication and delayed gratification. Becoming a productive scientist also requires more than high-level knowledge and cutting-edge research skills. Success in the science enterprise means integrating the norms and expectations of the profession. The novice must learn what is valued by the community, what should be discussed, and what should be ignored (Zuckerman, 1996). This special "tacit," or informal knowledge needed to become a master, is learned through the mentor relationship and deep imme rsion in a field (Murray, 1987; Simonton, 1992; Steinberg, Forsythe, Hedlund, Horvath, Wagner, Williams, Snook, & Grigorenko, 2000).
Outreach directors for high school and some college mentorship programs create matches between academically talented or motivated students and working scientists. At the graduate or professional level, however, mentors scout out students who appear to be potential protegees, those most likely to benefit from the mentors attention and guidance and most likely to contribute to the work of the laboratory. Being sought out by a prominent researcher to study or work with him or her foreshadows a successful science career (Zuckerman, 1996). The master may provide his or her apprentice with role modeling, intellectual sparing, emotional support, or introductions to important people in the field and other professional advice (Jacobi, 1991). The influence of a mentor is even more powerful in maintaining women in the science pipeline (Lewis, 1991; Light, 1990; O'Brecht, Pihl, & Bois, 1989). For the talented woman, additional variables come into play, such as the number of young children in her family, her marital statu s, and the gender of the mentor (Long, 1990; Mahony, 1995). Socio-cultural values, as well as institutional contexts, may influence the opportunities for collaboration between mentor and protegee, as well as for career advancement (Crosby, 1991; Hackett, 1990; Lederberg, 1990). …