University CS graduation surpasses its 2003 peak, with poor diversity
April 10, 2017 at 7:00 am 4 comments
Code.org just blogged that we have set a record in the number of BS in CS graduates.
University CS graduates have set a new record, finally surpassing the number of degrees earned 14 years ago.With a 15% increase in computer science graduates (49,291 bachelor’s degrees), 2015 had the largest number of CS graduates EVER! The previous high point was over a decade ago, in 2003.
Source: University computer science finally surpasses its 2003 peak!
But look at the female numbers there — they are less than what they were in 2003. We are graduating 2/3 as many women today as in 2003. (Thanks to Bobby Schnabel for pointing this out.) We have lost ground.
My most recent Blog@CACM is on the new CRA “Generation CS” report, and about the impacts the rise in enrollment are having on diversity. One of the positive messages in this report is that departments that have worked to improve their diversity have been successful. As a national statistic, this doesn’t feel like a celebration when CS is becoming less diverse in just 12 years.
Entry filed under: Uncategorized. Tags: BPC, Code.org, NCWIT, women in computing.
Computing Ed Research - Guzdial's Take 
1.
Marcel Weiher | April 10, 2017 at 6:02 pm
In the meantime, diversity in overall college graduates continues to decrease, with males now making up around 40% of bachelor degrees pretty much across the board.
https://nces.ed.gov/fastfacts/display.asp?id=72
Considering that the totals skew strongly female, it follows that there must be other degrees that skew more female than CS skews male.
Maybe there are different preferences at work?
2.
Mark Guzdial | April 10, 2017 at 6:03 pm
For sure. I just responded to some of this issues in a comment on my Blog@CACM post on the “Generation CS” report: https://m.cacm.acm.org/blogs/blog-cacm/215245-generation-cs-when-undergraduates-realized-they-needed-computing/fulltext
3.
Michael Kirkpatrick | April 11, 2017 at 12:08 pm
It depends on what constitutes “different preferences.” One of my favorite visualizations comes from Randal Olson a couple of years ago: http://www.randalolson.com/2014/06/14/percentage-of-bachelors-degrees-conferred-to-women-by-major-1970-2012/
Compared to 1980, the percentages for women recipients of Bachelor’s degrees is higher in engineering (10% then, 18% now) physical sciences (23% then, 40% now), business (30% then, 48% now), and biology (40% then, 58% now). Math & statistics stayed approximately level in the low 40% range.
During that same span, CS dropped from 35% to 18%, with precipitous drops in the late 1980s and early 2000s. If CS had been consistently low for a half century, the case could be made that the field as a whole wasn’t necessarily appealing to girls and women for cultural or social reasons. But that’s not the case. Those drops aligned with rapid cultural changes in the field, often relating to gaming, hacking, and the Internet. This suggests it’s not about the girls; the field itself has changed.
4.
Marcel Weiher | April 30, 2017 at 6:49 am
I find it interesting where both of you draw your comparison from: the past, and a Sharia law country.
So where do you think women are freer to choose what they actually *want* to do? Modern western democracies or the past or Shariah law countries?
Another data point: Russia. http://www.bbc.com/news/business-39579321
From the article:
“… in Russia, even the very youngest were extremely focused on the fact that their future employment opportunities were more likely to be rooted in Stem subjects.”
So maybe higher CS enrollments by women reflect women not being free what they *want* to do, but rather doing what they feel they *have* to do.
Once people have the freedom to do what they like to do, the proportion of women in CS and other engineering subjects drops.