Managing CS major enrollment boom with a lottery: “A lottery, by definition, is fair.”
June 22, 2020 at 7:00 am 22 comments
I am excited to see that the University of California, San Diego is now managing their over-enrollment in the computer science major with a lottery — see the article here.
Instead of enrolling students holistically or based on GPA, the department selects at random — assuming they exceed the 3.3 CSE GPA threshold. With the lottery system, all students are equally considered despite differences in their experience, drive, and ability.
When asked about the implications of the new system — and possible disadvantage to high-performing students — CSE Chair Dean Tullsen explained, “a lottery, by definition, is fair.”
“I think there’s this false assumption that the students who work harder are the ones who are getting the 4.0s, that hard work directly translates to a higher grade. [The lottery system will] admit a lot of hard-working students who weren’t getting in before,” CSE Vice-Chair for Undergraduate Education Christine Alvarado added.
This is a much more fair system than simply allowing in the top GPA students. It probably doesn’t make Tech companies happier, but it’s not clear that it makes them less happy. They will still get lots of potential employees who are above the bar. Those employees will likely be more diverse than the graduates being produced from CS programs today. The students getting the top grades in the early classes are typically those with more opportunity to learn CS, more wealth, and more privilege. A lottery says that anyone who is prepared for the courses can take them.
Entry filed under: Uncategorized. Tags: BPC, broadening participation in computing, computing education, undergraduate education, undergraduate enrollment.
22 Comments Add your own
Leave a comment
Trackback this post | Subscribe to the comments via RSS Feed
Computing Ed Research - Guzdial's Take 
1.
Mike leech | June 22, 2020 at 7:18 am
A lottery is fair for those that play it. And the more tickets you buy the better you do. What’s unclear from this article is the actual shift that a lottery would likely bring about. Not to say that it won’t be better, but it would be good to see the data they are running off.
2.
alanone1 | June 22, 2020 at 7:25 am
Interesting issue!
I agree with the idea that grade points might be very misleading abstractions of a student’s potential.
A lottery seems as though it should be a last resort.
(Dave Evans looked at resumes to try to gauge whether a candidate was “interesting” (in part he meant: “not terribly typical”)).
This worked very well for the U of Utah grad school. Would it work for undergrads?
I wonder whether this would be a better heuristic before doing a lottery?
3.
Mark Guzdial | June 22, 2020 at 9:19 am
There is a movement towards “Holistic Evaluation” for admissions which might include a review of resumes: https://www.thoughtco.com/what-are-holistic-admissions-788426. It’s not how one avoids the same bias as grades.
4.
alanone1 | June 22, 2020 at 11:10 am
An interesting example is “awards”. Since some are sometimes given out to not so deserving recipients, one could propose doing a lottery for these also.
On the other hand, a lot of the people I know who have received awards really do deserve them and really are more able than most of the people who don’t get them.
I think Dave Evans’ idea was that people who do things and might be good candidates don’t just start when they get to college, most of them have been doing -something- all along.
I guess the question is whether there should be different levels of “awards” for people who do more ably, or not. I have to admit that I’m prejudiced towards able people, whether conventionally or unconventionally able.
I’m having a hard time warming up to the lottery idea as -the- alternative to using grades (which I also thinks misses pretty badly, especially for grad school).
On the other hand, I like the idea of giving second, third etc chances.
But on what basis? Maybe there should be a 10% category that is filled via a lottery?
5.
Mark Guzdial | June 22, 2020 at 11:28 am
Should access to CS education be an award? It’s super valuable these days. Is it too valuable just to be made available to people lucky enough to gain access to CS in school?
6.
alanone1 | June 22, 2020 at 11:54 am
Hi Mark
I don’t know how to answer that question. I think at some point in life that people who have done more should be able to benefit from their extra work. Is that an “award”? Is it a “reward”?
Also, I think immense efforts should be made to furnish equal opportunity of every kind, especially in the early years.
And, as I said, many chances should be given throughout life.
But “Biology is variation” and “life has randomness: so no matter what you do there will not be an even distribution of abilities and potentials.
All the factors above have to be heeded I think.
7.
Mark Guzdial | June 22, 2020 at 1:47 pm
Hi Alan,
We’ve spoken often about how science was invented to deal with the fact that humans are error-prone and biased. We’re not likely to figure out the relationship between temperature and pressure of a gas, or the nature of life, or how a vacuum works without careful measurement and experimentation. Shouldn’t humans be just as error-prone or biased when it comes to making judgments about each other? Our most natural bias is, “Someone who is going to be successful is someone who is like me, since I’ve been successful.” Sure, we might strive to seek out unexpected or unconventional abilities, but it’s possible that guessing at unconventional ability is like inventing gods to explain the weather — it’s a creation on our part, mostly based on ourselves and what we can imagine, without a strong basis in empiricism. We’re better off inventing evaluation mechanisms that get around our biases, like blind auditions for orchestra seats: https://www.theguardian.com/women-in-leadership/2013/oct/14/blind-auditions-orchestras-gender-bias
I don’t have an answer to the best way to allocate CS classroom seats as a scarce resource. I do think that a lottery is a better way than humans making a judgement about who is likely to succeed. For many reasons (history, bias, wealth), some people get the chance to look like they’ll succeed, when others (who may be able to succeed just as well!) do not.
8.
alanone1 | June 22, 2020 at 2:46 pm
The theory of K-12 is “universal education” (one of the “UN Universal Rights, and there have been several recent court cases in the US that have upheld this idea).
The results in the US is very uneven and should be made more even. The unevenness in K-12 is carried into college, and even in the current not so equitable system, many students are doing very poorly (per the NAEP).
There’s no question that most students who go to college are not even getting the same level of general education that was required for HS some years ago (and for up to 8th grad more than 100 years ago).
So I think that college will wind up becoming a universal right and a public benefit in the US as K-12 now is.
In the 19th century the “10%” went past 8th grade, and the 10% could come up with teachers and professors for the 10%. When universal secondary education was made a right, there was neither a big plan for nor a push to fund where the teachers were going to come from.
The babyboom in colleges was enough of a swell to pretty much break the system for many aspects, especially outlook and quality.
One theory that was prevalent back in the 19th century was that part of the main goals of the 8 years of schooling was to not just be able to read, write, and cipher, but to learn how to learn from reading.
I think the current situation is that, even in college, most students really want to get their content orally in lectures (this seems like a really bad and also inefficient way to try to learn something).
It’s likely that we’ve already gotten to the point where a college education is more of a social badge than a learning experience. It’s possible that this is part of general movements towards equity, and that this is a good enough reason.
On the other hand, I don’t think we want our doctors or our airplane designers to be a general product of this process. There needs to be a place where the actual levels of knowledge vis a vis what the larger cultures have accumulated can be acquired.
Another one of Dave Evans’ theories could apply here. He over-admitted to a considerable degree, and because he didn’t trust either graders or even resumes very much. He treated the grad students like gold for two years, and then, unbeknownst to them, the faculty would meet to access each student at that mark, and decide whether they were “a real PhD”. If not, they went out one door with a Masters, and if so they went on to a PhD.
I found this out when I was on the faculty there for a year following my PhD. I thought it was a very good system. The overadmitting softened the past, and allowed some real misfits — like Jim Clark and a number of others who later became famous — to have their wonderful chance without knowing about the 2 year assessment. This worked very well I thought.
This could work in colleges today — but the colleges today are much more like businesses, and so they want retention, instead of bestowing an opportunity followed by a local assessment.
State universities used to have something like this: many of them had to admit any student who lived in the state and had graduated from high school. But they used the first two years (especially the first one) to “locally assess”.
Doesn’t this seem more equitable than a lottery?
9.
Mark Guzdial | June 22, 2020 at 3:28 pm
Hi Alan,
Unfortunately, I don’t. People make biased judgements. Without any mechanisms to reduce bias, people will tend to pick people that look and/or seem like them. I see no reason to believe that CS ability is to be found more in men than women at a ratio of 5:1, but that’s what the current selection mechanisms produce.
10.
alanone1 | June 22, 2020 at 3:55 pm
Seems to me that universal admittance is more fair than a limited one via a lottery. In both cases there will still be problems of assessment, etc. There will still be problems of “retention” (which in many cases amounts to students being able to purchase their certification). There will still be many kinds of bias once the students hit campus.
Still, I’m not against trying lots of different things.
But it seems to me that college/university itself is now much more flawed than it was, and it’s hard to see how being equitable and perhaps with questionable standards is going to help a systemic problem much.
But, sure, why not try to get a more balanced freshman class?
Here’s what worries me from observing at UCLA: the conglomeration of many factors that we’ve been discussing has made a “really good PhD” very difficult to do, from being able to get top students, to the paper counting by the Regents of the U of C, to the choice of topics, to the lots of little papers by the grad student, to the way the funding is gotten and allocated and judged, to the almost 65% overhead that UCLA takes off the top from e.g. NSF grants.
This amounts — in the so-called field of Computer Science — in a really disappointing — and essentially polluting — of the field that has throttled up the one place left for possibly improving the most pressing problems and issues of our not-quite-a-field.
I understand the clamor for “computing” and for “more equity in computing” — I’m all for improving this in every way we can. But there is also a tragedy here for all if the field itself is left to dilute itself into irrelevance.
11.
gasstationwithoutpumps | June 22, 2020 at 5:13 pm
Alan, the “lottery” system described is closer to what you propose than it is to a pure lottery. Students are admitted to the university and allowed to take lower-division courses. Those who do well enough (3.3 GPA) are eligible to continue in the CS major, but because capacity is capped, there are more of the eligible students than there is room for. The lottery is used to select among the eligible students. There is some difference in students between a 3.3 GPA (B+) and a 3.9 GPA (A), but not a lot, and it is often more correlated with previous education (and because of that, with race) than with effort or ability.
A pure GPA system would raise the stakes for grades in lower-division courses very high, resulting in substantially more cheating (which is already a serious problem in lower-division CS courses).
UCSC does something probably less fair than UCSD—no student can change their major to CS. They have to be admitted (as freshman or transfer) into the CS major. There are other options (like computer engineering, which is a better balanced degree at UCSC, CS: Game Design, and Biomolecular Engineering and Bioinformatics) that students can declare without being directly admitted to the major, but someone who really wants CS and is not admitted to it as a freshman is best advised to go to a community college and try again as a transfer student (the admission ratio is much higher for transfer students).
I am not convinced that the “holistic” admissions process at UCSC is fair. It is certainly not transparent—just like most admissions processes at most colleges, the biases of admissions officers have a huge weight and there is no way of determining why some students are admitted and others not.
12.
alanone1 | June 23, 2020 at 12:13 am
Hi Kevin
It seems that a lot of the current processes have gotten there via capacity problems (including perhaps, too many students to preclude a faculty review of the kind that Utah did).
The idea of having to figure out your major on entry as a freshman seems to verge on the ridiculous — and is against what “growth while in college” is supposed to mean.
I will admit to being quite touchy about this. None of the major universities I’ve been a professor at would have accepted me as a grad student on today’s terms. And I had no idea what I wanted to do at all when I was a freshman. For me, college was exploration and growth (so I’ll admit I’m prejudiced in this direction).
My GPA was bimodal, so: “interesting” if not averaged … Fortunately for me Dave Evans didn’t look at GPAs at all.
13.
gflint | June 24, 2020 at 11:57 pm
I have to agree with Alan. I graduated high school at 17 with a 2.2 GPA. High school was an absolute bore. I started college at age 25. My GPA was 3.8+ for three years. I did not find a major (Secondary Ed) until my junior year. High school GPA is an indicator but should not be a limiter. A student knowing their major as a new freshman is totally ridiculous.
14.
orcmid | June 29, 2020 at 12:03 pm
I confess to similar experience, although I managed to stay away from academic setting for a long time after dropping out as a college freshman. When I decided to complete a degree as an adult learner, I used external programs and did not study CS, challenging out of it. Nowadays, I rely on MOOCs.
At my start of college in 1957, the State universities had open enrollment and tuitions were pretty affordable (e.g., something far under $100 per quarter hour). Caltech was $300 per quarter. MIT wasn’t terribly more. Housing and books were the big expense.
All very anecdotal and doubtless unrepresentative. I strongly agree about choice of concentration/major though. I had no idea what I was doing and CS was not an option. But hanging out in an early university computer center and being quickly accepted and trusted to use the systems was amazing for me. I completely ditched any ambitions to complete university education and devoted myself to becoming what I now consider qualification as a CS. There were few entrance walls then, neither occupational nor academically. I take it there is little opportunity nowadays.
I am concerned with how CS Education is accomplished with the current technology and its complexity, yet the current collegiate situation is unimaginable for me.
15.
shriramkrishnamurthi | June 22, 2020 at 9:18 am
“Now”? This article is from 2017!
16.
Mark Guzdial | June 22, 2020 at 9:25 am
Sorry — new to me. I was talking to Leo Porter and we got to talking about their lottery, and this was the only article I found on it.
17.
Cyrus Omar | June 22, 2020 at 11:21 am
A lottery also introduces the anxiety of uncertainty and tragedies of chance. If I knew I wanted to major in CS and I was admitted to an otherwise great school that had a lottery system for admission into the major, I would seriously consider choosing a different school. What happens if I don’t get into the CS major at that school, but I still want a career in CS? Do I drop out and transfer to another school? What a disruptive and anxiety-inducing situation!
18.
Mark Guzdial | June 22, 2020 at 11:26 am
Sure, but let’s contrast with the student who spends four years trying to get in to the major but has little prior background and never gets the grades to gain entrance. https://cse.engin.umich.edu/stories/daetriana-burks-reed-fighting-for-minorities-in-computer-science
19.
BKM | June 22, 2020 at 4:47 pm
Are there any statistics on how well students persist in the major at this school compared to schools who use grade based standards to let students into the major?
20.
gasstationwithoutpumps | June 22, 2020 at 5:19 pm
Since this system is grade-based (minimum 3.3 GPA in lower-division courses), it isn’t clear what you are trying to contrast with. Also, a pure grade-based system with no other considerations probably doesn’t exist, as college admissions practices almost always include other information besides grades.
Comparisons between institutions, and even between cohorts, can be very difficult, as exactly what courses are included in the GPA calculation at the time of decision and how harshly they are graded could make a big difference.
21.
Megan | June 27, 2020 at 9:28 pm
This is an interesting point. If grades from the entry level courses can be transfered in from other schools, that is an entire layer of additional variability to control for in some way. If not, that’s a smaller denominator for Ss with transfer credit to work with.
I would assume that, were a program to adopt such a policy, the feeder classes would have to be coordinated to reduce variability between sections, instructors, and semesters, so that a 3.3 means a 3.3 for all students of a given cohort. I also have to wonder about slow starters, who repeat a course but then really catch on. What of their GPA? Is that grade replaced or averaged in? Is there, as mentioned elsewhere, opportunity for growth and learning in such a system?
I understand that a boom in CS has meant a real challenge for departments and that there is probably no one best way. Perhaps communication and collaboration with other universities and community colleges and shared tuition or other models, where students who are strives or want to pursue CS but don’t mKe the cut at these overflowing programs can go elsewhere without losing the time spent. Maybe a more global uni model than ailoed universities, with shared resources and student bodies? Harder in college towns and rural areas.
That said, once admitted on opaque requirements to a university, a lottery is inherently unbiased. The issue is the sunk cost for those who aren’t accepted but still meet requirements. Reciprocity with other schools or streamlined transfer for those students maybe could be options. But there will be humans who still make exceptions and pull strings, because that’s what humans do. That’s the flaw.
22.
Changing Computer Science Education to eliminate structural inequities and in response to a pandemic: Starting a Four Part Series | Computing Education Research Blog | July 13, 2020 at 7:00 am
[…] majors are racist in the ways that she describes. There are mechanisms that are better, like the lottery I described recently to reduce the bias in admission to the major. Amy’s points are inspiring this blog post […]