Communications of the ACM

˲ Regular email updates, or Skype discussion sessions: successful groups typically designate a mentee who starts the email conversation on a certain day of each month, which generates a thread of updates.

˲Face-to-face meetings at conferences that most group members attend: successful groups typically schedule their own work sessions to give feedback on each other’s papers or grant proposals during a conference.

˲Substantive interaction around research, setting norms and expectations of continued input as careers progressed: more-productive members set a standard for the less-productive ones.

˲Procedural interaction around norms of the profession, explicitly communicating what is often unspoken.

˲Mentorship and sponsorship by senior mentors: effective senior mentors often introduce junior women to other senior colleagues and initiate workshop or seminar invitations.

Impact of CeMENT in Academia

The interim results of CeMENT sug-
gest that putting women in touch
with other women with similar re-
search interests across institutional
boundaries is helpful, and that this
type of mentoring program improves
women’s grant and publication re-
cords. As a result of the publication
of the CeMENT evaluation, which es-
tablishes a benchmark for the effects
of mentoring female assistant profes-
sors in the economic profession, sim-
ilar mentoring programs have been
established by various professional
associations, such as the American
Philosophical Association, the Chi-
nese Women Economists Network,
the Japanese Women Economists
Network, and Association for the Ad-
vancement of African Women Econo-
mists. Subsequent programs do not
use RCT to select a treatment and
control group. Instead all eligible ap-
plicants are treated.

Application to Academic
Computer Science

Women-to-women mentoring has been in the fabric of the computing field for a long time. Systers, an email listserv, was founded by Anita Borg in 1987 and is still active with over 6,000 members today. Its members include academics, computing professionals, and students. Since 2010, the Computing Research Association Women (CRA-W) has held early career mentoring workshops for women and since 2013 mid-career workshops for women. Mentoring is a major feature of the annual Grace Hopper Celebration of Women in Computing, sponsored by the Anita Borg Institute. One focus of the Academic Alliance in the National Center for Women and Information Technology (NCWIT) is the retention of women in academia. Many of the lessons learned in CeMENT are already practiced in the computer science community. It is encouraging to see that many of the practices employed in the computer science community to improve academic retention have been validated in the CeMENT controlled study.

References

1. Blau, F.D. et al. Can mentoring help female assistant professors? Interim results from a randomized trial. American Economic Review 110, 2 (May 2010),

348–352.

2. Blau, F. D., Ferber, M. and Winkler, A. The Economics of Women, Men and Work. 6th ed. Pearson/Prentice-Hall, Upper Saddle River, NJ, 2010.

3. Ginther, D.K. and Kahn, S. Women in economics: Moving up or falling off the academic career ladder?

Journal of Economic Perspectives 18, 3 (2004),

193–214.

4. McDowell, J.M., Singell, L.D., and Slater, M. Two to tango? Gender differences in the joint decision to publish and coauthor. Economic Inquiry 44, (2006),

153–168.

5. Zweben, S. and Bizot, B. Taulbee Survey. Computing Research News 28, 5 (2016), 2–60.

Yan Chen ( yanchen@umich.edu) is the Daniel Kahneman Collegiate Professor of Information at the University of Michigan, and Distinguished Visiting Professor of Economics at Tsinghua University.

The author thanks Rachel Croson, Janet Currie, Richard Ladner, and Laura Razzolini for their helpful comments.

Copyright held by author.

treatment effects in terms of the three main predictors of successful tenure in Economics, the total number of publications, the likelihood of top-tier publications, and the total number of federal grants in treated women relative to those in the control condition. 1

By three years after the intervention, participants in the treatment have two more publications than those in the control condition, and by the fifth year, those in the treatment group have three additional publications. When evaluating the magnitude of the treatment effects, one must bear in mind the differences in publication norms between economics and computer science. Unlike computer scientists who typically publish several short papers ( 5–10 pages) each year, an economist typically publishes 1–2 longer ( 40–50 pages) and more substantive papers each year. This thus represents an extra year of productivity for those who were mentored.

Another measure of career success in economics is top-tier publications, which refer to the top-five general-in-terest journals among more than 300 economic journals. At least one top-tier publication is often necessary for tenure promotion in many economics departments in research universities. Participants in the mentoring workshop were 20 percentage points more likely to have a top-tier publication after three years and 25 percentage points more likely after five years.

Lastly, obtaining a NSF or NIH grant is quite unusual among junior economists compared to computer scientists who are expected to fund their own Ph.D. students. By the fifth year, workshop participants had 0.4 more NSF or NIH grants than those in the control condition. 1

Lessons learned. Although junior women in the treatment group on average are more successful than their control counterparts, considerable heterogeneity in activity levels exists among different groups. We summarize several features that might have contributed to the success of the program and characterize some of the more active groups: