A Tale of Two Recruiters

October 28, 2014 Civis Analytics

Last Wednesday, Annie and I headed up to Carnegie Mellon University (my alma mater! Go Tartans!) for the Spark Start-Up Career Fair to find the next batch of Cives. Because we have a tendency to (over)analyze any data we can find, here are some interesting facts about our trip.

According to CMU’s Career and Professional Development Center, 400 students registered for the fair. As students came up to Civis’ table, they could either approach Annie or me, and almost everyone with whom we spoke left a resume, which was helpful for data-collection.

Because Civis is committed to building a community of women in STEM, we were curious about how two women of color would be received. While we’d prefer to run an experiment in which we randomly assigned who got sent to which fair, small sample sizes make that sort of thing hard to analyze; instead, we’re just going to offer some descriptive statistics and make some semi-speculative leaps.

Students we met, by gender

Much to our delight, we saw a rough parity in the number of female and male candidates, which matched the composition of the school.

Students we met, by race

The applicants we received were substantially more Asian than the school, which was notable in part because Annie and I are both Asian women–but we’re not sure how much of that was us versus the composition of the fair.

Indeed, the most interesting thing we learned wasn’t actually about how effective Annie and I were together; it was about how different the two of us were. For example, here’s the breakdown of resumes we received by recruiter.

Resumes received, by recruiter

I’d like to point out, for the record, that I actually missed 30 minutes of the fair (12.5%) to have lunch with a professor, so we’re not exactly apples-to-apples here. But even after you make that adjustment, Annie still talked to many more people than I did.

But I wasn’t just standing around! In fact, we suspect that my conversations were, on average, longer than Annie’s because students had more questions for me. Because I’m a CMU alumna, I got a special badge, and lots of students were particularly interested in how my time at CMU prepared me for Civis. Those were questions that Annie was never asked.

If we dig into this a little more, we see that not only did Annie and I talk to different numbers of people, we talked to different types of people.

Gender and age of students we met, by recruiter

Whereas Annie talked to twice as many men as women, I talked to twice as many women as men, and whereas Annie spoke mostly to undergraduates, I spoke mostly to graduate students. Maybe this was because Annie looked a little younger, so undergrads felt more comfortable approaching her. Or maybe because I was a female alum, women felt more comfortable talking to me.

Finally, a note on homophily, which refers to the tendency of similar individuals to cluster together: this explains how most of your friends are probably in the same racial and socioeconomic group as you and why you’re more likely to marry someone who shares your political beliefs. Homophily got a bit of a moment in the sun after researchers Fowler and Christakis published a paper on how obesity was “contagious” within a social network: researchers who re-examined their work argued that the “contagion” effect, which Fowler and Christakis argued caused weight gain, could be explained by mere homophily.

Annie and I may have another data point to add to this debate:

Departmental interests of students we met

I work in the Data Science R&D department. Annie works in Applied Data Science. Somehow, both of us managed to convince a disproportionate number of our candidates to express interest in our departments. Was this homophily?—maybe people who were interested in research approached me, and people who were interested in more applied work approached Annie. We’re not so sure. We didn’t wear or say anything to identify our department. Moreover, we’ve collaborated on projects before, so we use the same vocabulary and discuss similarly technical subjects (albeit in different contexts). Plus, we recruited (roughly) the same proportion of engineering candidates, so it wasn’t as if one of us seemed significantly more technical than the other.

So what’s going on? Our working theory is that we just both love our jobs and find it easy to get other people interested in our jobs too.

If you’re interested in working at Civis, you can apply online. Both Annie Wang and I would be happy to talk to you (though, statistically, it appears I’m less approachable).

Written by Elaine Lee. This post is co-authored with Annie, who is endlessly amused by the fact that CMU’s mascot is Scotty the Scottie Dog.

The post A Tale of Two Recruiters appeared first on Civis Analytics.

Previous Article
A Formula to Find the Uninsured Around the Country
A Formula to Find the Uninsured Around the Country

Enroll America is an organization with a big job: to find the uninsured around the country and persuade the...

Next Article
City Uses New Technology to Target Women, Encourage Mammography
City Uses New Technology to Target Women, Encourage Mammography

“CHICAGO ““ The Chicago Department of Public Health (CDPH) has joined Civis Analytics to conduct targeted o...