Gas station without pumps

2014 November 24

A seat at the table

Mark Guzdial in #Gamergate as a response to re-engineering: BPC as a conspiracy to change computing wrote

We in the Broadening Participation in Computing (BPC) community are aiming to achieve a similar kind of social engineering that the Gamergate supporters are complaining about. I am part of a vast, international (though maybe not particularly well-organized) conspiracy to change computing culture and to invade computing with many women and members of under-represented groups. We are “actively plotting to influence” computing. The Gamergate supporters argue that the conspiracy is about “artistic aspirations.” In BPC, we say that we’re about social justice, equity, and diversity. From the perspective of the “engineered,” the difference in purpose may not make much difference. One of the pushbacks on the call I shared to eliminate nerd culture was, “Can’t we just shape/change nerd culture?” Do the nerds want to be changed?

What might a response to BPC look like? Might well-prepared, privileged male and white/asian CS students complain about efforts to give seats in classes to women or under-represented minorities whom they may perceive as less-prepared?

I have no objection to giving seats in classes to anyone capable of  learning the material, but I believe that this needs to be done by increasing the number of seats, not taking them away from other students.  I’m all in favor of expanding the pipeline, but not of holding back those who have already started on the path, so that others can “catch up”.

There’s a general awareness that there’s a problem, but there’s less conviction that it’s an important problem or that there’s an obvious way forward to fixing it.

I agree that the problem of gender imbalance and racial imbalance in CS is an important one, but I’m less convinced than Mark that there is an obvious, equitable way to fix the problem. He seems to think that lotteries are the way to go:

In NPR When Women Stopped Coding in 1980′s: As we repeat the same mistakes, Mark wrote

I understand why caps are going into place. We can’t support all these students, and there are no additional resources coming. What else can CS departments do? We might think about a lottery or using something beyond CS GPA to get those seats, something that’s more equitable.

I disagree with him strongly on this. I responded on Mark’s post with the following comment:

I’m not sure that I agree with “We can’t support all these students, and there are no additional resources coming. What else can CS departments do? We might think about a lottery or using something beyond CS GPA to get those seats, something that’s more equitable.”

Granting access to a limited resource to those whose prior achievement is highest seems to me to be highly equitable. Denying higher achievers because they are of the wrong race or gender does not.

Increasing the resources available for teaching, so that we don’t have to restrict who majors in a field seems like a good strategy, as does providing slower on-ramps for those who did not have good early training. But denying entrance to those who may have dedicated their lives to the field, just because others did not have (or did not take) the opportunity to reach that level of achievement—that does not seem “equitable” to me.

Note: I may be biased here, because my son is a white male majoring in computer science who has been doing recreational programming as a major activity since he was 10 years old. I would be very offended if he had to win a slot in the major by a lottery—college admissions alone is enough of a lottery these days.

Are we then to tell students not to form any intellectual passions in middle school or high school, because doing so will get them labeled as “privileged” and denied further opportunity? Or should they only form passions for things that no one cares about, so that no one will try to take their passions away from them?

Although I’m not fond of sports analogies, it is common for people to point out the absurdity of the lottery position by suggesting that the same be applied to sports teams. The football teams at the Big 10 schools should not consist of those privileged athletes who started young, got the best training, and had the best performance in high schools, but should be assigned by lottery to anyone who is interested in playing, even if they have never picked up a football in their lives. Why should only those who had the good fortune to be large, fast, and strong be allowed to play?

Michael S. Kirkpatrick countered my comment with

It’s often so hard to be objective when it comes to perceptions of equity. As Anatole France observed, “In its majestic equality, the law forbids rich and poor alike to sleep under bridges, beg in the streets, and steal loaves of bread.” The open question is whether those students truly are higher achievers, or if they are just starting from an advantageous position. In that case, would it not be more equitable to give the opportunity for students who did not have prior opportunities?

His argument makes the assumption that primary goal of college education is a social justice function—to provide opportunity for those who have not previously had it. While a generous impulse, this philosophy taken to extremes results in eliminating grad schools and upper-division courses to create more freshman courses, and even replacing freshman courses with remedial courses, resulting in college as very expensive high school (or, in the case of some athletes in scandal-ridden schools, grade school).  Increasing opportunity is a great thing, but it shouldn’t be allowed to kill off the other great things about universities: like the opportunity for people to stretch their minds to the limit, to share ideas with other intelligent and passionate people, and to advance the state of the art. While universities do serve an important role in aiding social mobility, it is not, in fact, their primary function in society.

A variant of Kirkpatrick’s argument has often been used to kill off gifted education in public schools (because of a correlation between socio-economic status and identification for gifted programs)—forcing the parents of gifted students to take on educating their children themselves, which only the wealthy (or upper middle class) can easily afford to do. This approach increases the disparity between the wealthy and the poor, as the gifted students with less wealthy parents get much more limited educations—defeating the original goals of “equity” that killed off the public programs for gifted students.

There are good reasons why many parents of gifted kids started referring to “No Child Left Behind” as “No Child Allowed Ahead”, as it was much easier for schools to reduce their achievement gaps by slowing down the students who were learning fastest than by speeding up those learning slowest. Guzdial’s approach to rationing CS education seems to be following the same model.

Bonnie responded to Guzdial’s post with comments about what her college is doing to broaden participation, speaking both of successes and failures, and ending with

I just don’t know how we can make up for the poor education they received in K12. And that, I think, is where the true inequity lies.

Here I agree with Bonnie—if the problem is that some students get support early and others get support late, the solution is not to slam the door in the faces of those who got early support, but try to extend early support to more people. For that matter, I’m not in favor of slamming the door shut on anyone.  I don’t buy Guzdial’s assumption that this is a zero-sum game and that the only way we can have more women and URMs in CS is to have fewer white or Asian males. I think that there is plenty of room in the tent still for everyone who is interested and willing to work at learning the material.  We should not be rationing education, but providing enough education that everyone can get as much as they want.

In response to a different commenter, Guzdial wrote

We’re not talking about employees, Ian. We’re talking about seats at the table for students. If you get more women and under-represented minorities enthused about CS, there are still not enough seats at the table. If we’re going to allocate seats based on current ability, we have to get women and URM students to be better than privileged white boys. That’s a really high bar.

You may be under the misconception that computing is a meritocracy. It’s not. It’s not those with the most merit. It’s those with the most privilege.

It is almost certainly true the computer industry is not a meritocracy—but we should be trying to make it one, not rationing out education like butter in WW II. If there are not enough seats at the table, then buy a bigger table with more chairs! That will cost less in the long run than squabbling over who gets seated now.

2012 February 27

Scientists don’t test hypotheses

Filed under: Science fair — gasstationwithoutpumps @ 15:33
Tags: , , ,

On his Computing Education Blog, Mark Guzdial wrote about Nancy Nersessian’s work on how scientists really work: The Scientific Method is wrong: Scientists don’t test hypotheses, but build models.  He describes her idea as

Rather than test hypotheses, scientists do experiments to influence their models of how the world works.  The hypotheses they test come out of those models, …

That is hardly a new idea.  I’ve been trying to convince teachers for years that a hypothesis is not a guess, not even an educated guess, but the prediction of a model in a situation in which different models make different predictions. (See Science fair time again or Google science fair, for example).

I suppose that technically the term “hypothesis” should be used for the model, rather than for the prediction made from the model, because it comes from the Greek ὑπόθεσις (hypóthesis), meaning basis or supposition. But what gets stuck in the “hypothesis” box in science-fair forms is usually the prediction, not the model (if we should be so fortunate as to have a model rather than a wild-ass guess from the students).

Perhaps we should banish the term “hypothesis” from science fairs entirely, since it is used so badly. In its place we should ask students to provide the models that their experiment can distinguish among, and the predictions that would result from each model.  By making the models (always plural!) be the center of attention, rather than the prediction, I think we could correct a lot of the misunderstandings that abound about the scientific method.

2011 August 18

Mark Guzdial doubts AI course is real

Mark Guzdial has commented on the Stanford AI course on his Computing Education Blog: Stanford on-line AI course draws 58,000 — but is it real?. He contends that the courses are not going to be real courses—that they will be watered-down “outreach” classes with no content.

I believe that the courses are real intro courses, based on the syllabi on their web pages: Machine Learning, Introduction to Artificial Intelligence and Introduction to Databases. I looked at the machine-learning course in some detail while writing my blog post about the news. The course does require linear algebra, basic probability, and some programming as prerequisites—exactly what I’d expect for an undergrad machine learning course.

I don’t think that the courses will be greatly watered down, as Guzdial assumes, but  I do think that fewer than 10% of the people who sign up will complete the courses. Stanford is not giving credit for the on-line courses, but the same lectures attended in person do get credit, so the lectures at least will not be fluff.

Indeed, the whole advertising value of the courses would be defeated if the courses were too easy.  They want to establish that Stanford students know a lot and have high skill levels (selling students as their product) and that Stanford courses are meaty solid courses that are worth the $51,000 a year in tuition, room, and board.  Neither of these advertising goals would be met by fluffy courses.

I do wonder about programming assignments, though, as those are essential for a CS course at the proposed level, and I don’t see how meaningful feedback could be given in a huge class. Of course, I have seen CS classes in which programs are just checked automatically for producing the correct output for a given set of inputs, with no one reading the code to see if it is well-written.   In some cases, the students are even given the test input ahead of time, so that they only need to fix the bugs actually triggered by the test. I suppose that the Stanford professors could have set up the machine learning and AI courses with this sort of automatic grading even of programming assignments.

Automatic grading of programs explains, in part, the absolutely abysmal programming I’ve seen from some grad students—no one had ever looked at their code before. Even in classes with human grading, there is often little attention to programming style, because reading programs carefully takes time.  In my small grad classes, I was averaging 30–60 minutes per student per week grading programs and providing detailed feedback, a task that would not scale to 50 students, much less to 50,000.

Another possibility is that the real Stanford students will have programs graded, but the online masses will not.  I don’t think that Stanford anywhere promised that the free course was identical to the course students are paying about $5000 for—in fact, they may need to ensure that there is a noticeable value added for the students paying tuition, and just getting Stanford credit may not be worth that much.

%d bloggers like this: