Cover Image


View/Hide Left Panel

•   Even though we have good data and should act on them, we nonetheless lack important data points and types of data that are needed to inform and guide future efforts.

Regarding the second point, most existing data pertain to individuals, and data on institutional context are much weaker. Running parallel to the individual/institutional problem is that of quantitative versus qualitative data. A good deal of quantitative data are available (although they are not complete), but several participants cited a lack of certain types of qualitative data that would elucidate key information about individuals’ choices and career patterns, and institutions’ climate, practices, and policies. For example, when the data show a drop in the number of women of color between high school graduation and college graduation, and between college graduation and completing a Ph.D., it is not known whether the “missing” individuals began graduate programs and dropped out or whether they did not enroll in the first place. Moreover, qualitative data gathering is necessary to reveal the nuances of individuals’ perceptions, choices, and experiences. The need for longitudinal data, in particular, was highlighted by numerous participants, as was the importance of periodic reassessment of metrics for productivity, advancement, and retention. These nuances could greatly enrich and inform institutions’ and organizations’ efforts to capture and retain top American talent. Some of the data needs identified by conference participants included:

•   Disaggregated data. Comments from the conference participants and written testimonies submitted by professional societies underscore the need for data disaggregated by race/ethnicity and gender.

•   Longitudinal data. There is a need for longitudinal data that are tied to multiple factors simultaneously: individuals in the training period (e.g., students, graduate students, and postdocs) as well as academic institutions (e.g., programs and policies; and rates of recruiting, enrolling, and supporting women of color).43

•   Qualitative data. Several presenters pointed to a need for more qualitative data that add nuance to the quantitative data currently existing or in the process of being gathered.

•   Better response rates from women of color and people in other potentially disadvantaged groups. In the breakout session “What Data Can and Cannot Tell Us,” the discussion pointed to the difficulties of gathering critical information from members (in the case of professional societies) or study populations (in the case of researchers). Women of color, for example, often may not respond to surveys or may choose not to provide identifying information—rank, race/ethnicity, or department—because they are concerned that they may be identified by doing so and their responses may become public. Trust, therefore, is a key component of obtaining higher response rates from women of color.44

•   The ability to determine the exact number of faculty working in STEM fields and their corresponding demographic information using data from the Integrated Postsecondary Education Data System (IPEDS).45 The commissioned paper from Hurtado underscored the challenge that the Higher Education Research Institute faces in


43 Appendix A-1: Education and Academic Career Outcomes for Women of Color in Science and Engineering.

44 See footnote 5 on the tension between acquiring more specific, actionable information and the need for confidentiality.


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement