What can undergraduate courses do to improve diversity in computing?

February 22, 2016 at 7:26 am 12 comments

Dan is one of the best computer science teachers I know, and I strongly agree with the goals he describes below.  I’m not sure how much intro courses can do to recruit more diverse students.  At Georgia Tech, Media Computation has been over 50% female since we started in 2003, but that’s because of what majors are required to take it and the gender distribution in those majors.  I know that Harvey Mudd, Stanford, and Berkeley have grown their percentage of females, but their undergraduates get to choose their majors while on-campus.  At schools like Georgia Tech, where students have to choose their major on the application form, the decision is made off-campus.

One clear thing we can do in undergraduate courses is retain more diverse students.  In our BS in CS, we graduated 16% female BS in CS students in Spring 2015, which is pretty good.  Taulbee Survey says that the national average is only 14.1% (see report here).  But our enrollment in CS is 25% female.  We lose a LOT of women who decide to try CS.  I’ve talked about some of the reasons in past blog posts (see post here about bad teaching practices and here about my daughter’s experience in CS at Georgia Tech).

Dan Garcia says there’s another important issue: Once courses are created, educators must make sure they’re reaching a diverse audience. Women and minorities are grossly under-represented, not just in tech fields, but also in computer science classes.Teachers should shake the trees and reach out to more kinds of students, not just the student who’s doing well in math. And, he says, connect computer science to bigger, more controversial topics, Garcia says, because coding and data are connected to issues of power. With the persistent digital divide, he says, educators must ask, “What does that mean for equity? What does that mean for fairness? Privacy issues? Hopefully the curriculum brings equity as part of it,” he says.

Source: Adding ‘Beauty And Joy’ To Obama’s Push For Computer Science Teaching : NPR Ed : NPR

Entry filed under: Uncategorized. Tags: , , , .

Dagstuhl Seminar Poster: Critiquing CS Assessment from a CS for All Lens Review the Draft of the 2016 CSTA K-12 CS Standards

12 Comments Add your own

  • 1. Bonnie  |  February 22, 2016 at 9:19 am

    Our intro to CS courses run about 25% African American and about 25% Hispanic. The rest are split between white, Asian-American, and not categorized. Most of our students are first in their family to attend college, and are often kids of immigrants from countries like Uzbekistan, Laos, Honduras, and Nigeria. So our classes are about as diverse as you can get. Unfortunately, many of the students struggle. I think the biggest problems are language skills for some of the immigrant students, and reading skills/study skills for many. I think the answer may be very intensive coaching for the most at-risk students, such as CUNY’s ASAP program which seems to be getting results
    “The ASAP students were given a lot of attention. They met frequently with advisers: 38 times a year, compared with six times a year for non-ASAP students. Unlike typical CUNY advisers, who have a caseload of 600 to 1,500 students, ASAP advisers advised 60 to 80 students. The ASAP students also got more tutoring, attending an average of 34 sessions (compared with seven in the control group). ”

    I see a lot of discussion of intrusive advising too, which I think might help – the kind of nagging reminders that parents traditionally took on. I know I have altered my practices so that I now tell students immediately if an assignment is missing, and I send out email if a student has been missing from class a couple of times.

    I wonder, though, how much this kind of effort will cost. HIgher educations costs are already unsustainable. I would definitely like to see more research on intensive advising for at-risk students so we can decide which are the most cost effective methods to pursue.

  • 2. alfredtwo  |  February 22, 2016 at 9:39 am

    I don’t think you’ll see more diversity until a course with solic CS is required. There are a lot of reason for that including bad advice given to female students and students of color. Far too few of these students have enough exposure to CS to know if they’d like it or be good at it.
    At my high school we have seen a big increase in girls taking our real programming courses since we changed our required freshman course to include real CS including some programming. Of course good teaching and projects that are not sole focused on what boys like helps a lot.
    Next year we start requiring a full year of CS for graduation. We’re working on adding more courses but always with a focus on projects and teaching that is not all about the boys who are already hooked.

  • 3. alanone1  |  February 22, 2016 at 10:10 am

    Apologize in advance for off topic:

    At GaTech, an undergrad has to choose their major on the application form? (This seems quite wrong, and even a bit ridiculous …) What were they thinking (or not)?

    • 4. Mark Guzdial  |  February 22, 2016 at 11:13 am

      I can’t defend it (I agree that it’s wrong), but I do understand why. Some of our programs are so packed with required classes there are no elective classes in all 8 semesters of a four-year Bachelor of Science degree. If they don’t declare on admission, they might take the wrong class in the first semester, and have to go for additional semesters — from the first day of class. Having programs with no elective hours (i.e., no choices at all for students) is also wrong.

  • 5. Ibrahim Albluwi  |  February 22, 2016 at 10:58 am

    I currently teach an algorithms course at a university in Jordan (Middle East). There are 87 students in my class all majoring either in Computer Science or Software Engineering. Females in this class make 44% of the students. I have taught other high-level CS courses also with females being more than the males.

    This high percentage of females in CS related majors in Jordan has always puzzled me given that other parts of the world usually complain about females being a minority. I have never heard of any serious campaigns or of any policies here in Jordan that encourage females to major in CS, and yet the percentage is relatively high.

    One of the reasons that I may think of is that many of the families I know here in Jordan look at Computer Science as a major that is “suitable for females”. They usually justify this saying that a married woman can work from home and look after the kids!

    I think that it is interesting to perform a cross-cultural study in order to understand why the number of females in CS-related majors in some cultures is high! Who knows … we may learn something interesting regarding how CS should be promoted among females!

    • 6. Mark Guzdial  |  February 22, 2016 at 11:15 am

      Yup, we see that in many classes in the MENA region — see my comments on Doha here.

      • 7. ialbluwi  |  February 22, 2016 at 5:18 pm

        Interesting read. It brings me memories from my BSc and MSc days, as I did them in the Emirates, which is quite similar to Qatar in many aspects. My batch had way more females than males … we were indeed a minority!

        By the way, it is a nice coincidence that the following Forbes article was shared on my Facebook by a friend:


        It poses the question of females in the US vs females in other parts of the world in CS.

        What is very interesting to me is that despite the high percentage of females in our CS departments, the number of females excelling in extra-curricular activities requiring sharp problem-solving skills is usually lower than the number of males. Take programming contests for example. I judge, organize and supervise students participating in many national and regional programming contests. The trend in these contests is quite clear. Males are mostly the champions. The top-ten teams are usually males. Females usually participate at the beginning and then quit.

        Does it have to do with the atmosphere of “competitiveness” as noted by some research papers? Or was the measure of enrolled female students in CS-related majors a false indicator of how well females are involved in the CS domain in the MENA region?

        It would be depressing if we find out that females enroll in CS majors for reasons that are not related to interest in CS (e.g. for cultural reasons). We will have to find a measure other than enrollment numbers. This is especially true knowing that changing majors after enrollment is usually looked at as a failure in our culture! I.e. Those who enroll in a major usually complete the major even if they don’t like it!

  • 8. kirkpams  |  February 22, 2016 at 12:43 pm

    I’m curious about this: “In our BS in CS, we graduated 16% female BS in CS students in Spring 2015, which is pretty good. But our enrollment in CS is 25% female. We lose a LOT of women who decide to try CS.”

    Are those numbers based on counting “enrollment” as anyone who takes a CS course or just majors? If it’s the former, it’s hard to see if you’re “losing them” if they never intended to stay; rather, it would seem to be more a failure to recruit them to switch in. In the latter case of counting enrollment as majors, I’m curious if there is any part of the curriculum that acts as a barrier, or if they leave in smaller amounts throughout the entire curriculum.

  • 11. gasstationwithoutpumps  |  February 22, 2016 at 1:48 pm

    I looked up UCSC’s numbers http://planning.ucsc.edu/IRPS/Degrees/2014-15/UndergraduateDegreesbyMajorandGender(DMF).pdf

    CS is graduating only 13.5% female, but this is a big improvement after spinning of the game design majors to a new department—the game design program was only graduating 9.7% female, despite approximate gender parity on the Computational Media faculty.

    Students don’t have to declare their major until the end of their sophomore year here, but if they want any of the engineering majors, they pretty much have to start first quarter of their freshman year (or take more than 4 years).

    • 12. Mark Guzdial  |  February 22, 2016 at 1:54 pm

      Thanks for sharing those.

      Whoa! Those Computational Media major numbers are awful, and so different from ours (see post here)! When I went to the OSTP event on games and CS Ed in December, there were lots of people saying, “The answer to more gender diversity in CS is to teach everything with game design!” And I thought, “That goes counter to all the data I’ve seen.” Your report support my perspective — game design classes are even more male than CS.

      I’m surprised that Robotics is so male. Here at GT, robotics has a significant female population. I’m surprised that it’s almost as male as computer engineering. I’d be interested in seeing national robotics numbers.

      The faculty gender diversity doesn’t actually have any influence on undergrad gender diversity. Thought experiment that supports that claim: You can easily find biology departments whose faculty are just as male as CS departments, yet the undergrad population will still be much more female.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Trackback this post  |  Subscribe to the comments via RSS Feed

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 9,057 other followers


Recent Posts

Blog Stats

  • 2,038,047 hits
February 2016

CS Teaching Tips

%d bloggers like this: