More Surveys, Enrollment Data, and Next Steps

These past few weeks have been mostly filled with trying to convince people to fill out our survey. I've been inputting a lot of data into spread sheets. We're not doing any analysis until we've collected all of the data, but here are just the numbers in terms of surveys collected so far. (I do not have Anthony's data included in this.)

	Freshman		Sophomore		Junior		Senior		(5th or 5th double)
	M	F	M	F	M	F	M	F	M	F

Gender Totals	88	31	14	0	47	7	19	2	1	1	210
Totals	119		14		54		21		2		210

Looks like we did pretty well administering surveys to the freshmen. I could not find the statistics on freshmen in SCS for 2009, but past data seems to indicate a fairly stable number of students enrolled in the program. It is my guess, though I will of course confirm this when I can get the real data, that we have completed surveys from 80-90% of the freshmen in SCS.

We haven't collected very much data from sophomores due to professors for many of the sophomore classes being unresponsive to e-mails requesting to administer the survey in their classes. I am planning on coordinating with Anthony to wait outside sophomore level classes to see if we can increase the number of completed surveys.

We are doing pretty well collecting data from juniors, and I know that Anthony had some schemes to specifically target seniors, so I am sure that we will be approaching about 50% participation rate of juniors and seniors for the survey. Unfortunately, we do not have that many women responding to our survey, however, I plan on asking women in W@SCS to fill out the survey, so hopefully we will have a more representative distribution.

In looking at enrollment data, trying to figure out the stats on SCS entering freshmen, I came across enrollment data for 2008. Here is a copy of the data with some statistics that I calculate from the data highlighted. When compared against other colleges within CMU, SCS has the highest percentage of male applicants, the highest percentage of offers of admissions to male applicants, and the highest percentage of male enrollment.

In addition, one would expect that once students have been admitted, a certain percentage will enroll. While this percentage may vary across schools as some schools and majors within CMU are more exclusive / have better reputations than other, we would not expect this percentage to vary drastically WITHIN a school across gender. For example, in H&SS (School of Humanities and Social Sciences), 22.70% of males who apply decide to enroll and 22.22% of females who apply decide to enroll. These percentages are very similar. This is pretty much true for most schools at CMU. The percentages - with the exception of SCS - differ by no more than 6%. When we look at SCS however, 36.00% of males versus 24.60% of females who were admitted, enroll.

This raises the question of why, if women are getting into SCS, are they not enrolling? Clearly the arguments of "women not being capable of succeeding in CS as it is" or "the need for 'pink' curriculum" cannot be called upon. Admissions is not based on gender, but merit - and the admissions committee felt that these women could succeed in SCS at CMU without changing the curriculum to be more "female-friendly" or "pink." And if it is not ability based or some intrinsic difference between men and women, it must be a cultural thing.

I think that the next research step that I would like to take is to develop interview / focus group questions aimed at a particular group. Anthony is focusing more on women and CS on an international level, while I focusing largely at CMU. One of the groups that I would like to target for interviews or focus groups is students who left SCS to major in something else. I don't know about the stats on men, but in my year, at least 4 women have decided to switch majors from SCS to something else (which is a significant percentage). It would be interesting to see why they switched, what major they switched to, why they thought they wanted to do CS, if they had any role models in the field of computing, etc. I would definitely be interested in seeing if men and women had the same responses for those types of questions and to see specifically what the stats were on numbers of women vs. numbers of men who switch out.

Survey

Besides reading Gender Differences, I have been doing a lot of work on getting the survey out there. We finalized the survey at our meeting last week and I've been contacting professors of core classes, asking if we can use the last 5 to 10 minutes of a class period to come in and do our survey. We believe that this method will give us a higher number of responses than just standing in a hallway trying to get people to answer our survey. No one wants to answer a survey when they're rushing off to class. I have contacted a number of professors, and of the one who have responded, all have said yes!

I administered the survey at my Competition Programming class last night and everyone was willing to answer the survey. Even in administering the survey, before reviewing the findings, I am learning so much. It turns out that about 1/3 of the students in the Wednesday Competition Programming class are not primary CS majors. Most students who were not primary CS majors were either ECE students (most of them) or math majors or graduate students in the CSD. Most said that they were double majoring or minoring in CS. I actually got one ECE student who asked to see the survey and when he looked at the questions, said that only a few of the questions were specifically related to students in SCS and that he would be happy to answer the survey! Unfortunately, I had to turn him down as this is specifically for students in SCS.

More Contests Follow-Up

I have been doing some more research into high school and middle school CS / programming competitions. As Anthony found, the two really big ones are USACO and ACSL. I've been talking to a lot of people here at CMU and I've found that some of them have done USACO and fewer have participated in ACSL. One interesting thing that I've found is there are many, many local competitions hosted by local universities or organized by some other party. If ACSL or USACO was not offered at a student's high school, many were involved in local competitions.

A lot of the students I know who have been interested in programming from a young age participated in online contests like Google Code Jam or other programming competitions. Unfortunately, these competitions have a heavy focus on only programming rather than algorithms and other parts of Computer Science.

Other competitions that are not just CS competitions are Robotics competitions which are all over the United States. FIRST has Robotics competitions for students from elementary school through high school. While it is not strictly CS (a large portion of many of the contests in engineering, most do require problems solving and programming). One of our hypotheses was that if CS is associated with math instead of engineering, it will be more attractive to women. I would be interested in looking at whether participation in Robotics competitions which contain a major engineering component as well as a CS component had any effect on women's decisions to major in CS.

I would definitely like to put some questions related to these competitions on some of our surveys. Find out how many people participated in CS / programming competitions and of what kind. ACSL / USACO, local competitions, online, Robotics, or some other kind?

I would also be interested in examining whether students participated in math competitions. Almost everyone I know in CS participated in some sort of math competition or math olympiad in high school. I would also like to look at it from a high school perspective. What percentage of students who participate in high school math competitions go on to major in CS? What percentage of students who participate in high school CS / programming competitions go on the major in CS? Robotics competitions?

Contests Follow-Up

This week has been a bit more preparation work. We've been discussing the surveys a lot as we hope to have that finalized sooner rather than later. Especially since September is slipping away oh so quickly. I imagine we'll at least have a prototype done within the next two weeks. Anybody willing to do a delightful fall survey?

That being said, a lot of my focus this week has been on the state of computer science contests in the United States. I found five specifically targeting middle or high school students that I found of note and will enumerate below.

SuperQuest: Not to be confused with the group that presented at CS4HS. SuperQuest was a competition for high school students. Teams were tasked with conceiving of a computational problem they found interesting. They'd create a proposal and submit it where a panel would pick out the most interesting proposals. The winners were then given the opportunity to model their problem on a supercomputer and in some cases the results were presented at various conferences.

Unfortunately, information on the SuperQuest program is rather scarce. Outside of an old paper linked on the Wikipedia page. I can't find any official information on the program suggesting it may have gone extinct at this point. That said, I feel like this program is interesting for a couple of reasons. For one, it makes the math/computer science relationship concrete. Students can use their already verbose backgrounds in mathematics and transform it into a computer science project giving them a readily available in. Additionally, it fosters teamwork, creating the proposal wasn't a solo activity. Further there is plenty of room for interdisciplinary projects. Just off the top of my head I can imagine modeling biological, chemical, and economic systems as being viable projects. It's a shame that information on this program is so scarce as I find this to be a fascinating idea.

A second contest I looked at was the USA Computer Olympiad. A contest running since at least 1992, this contest prepares and selects high school students for the International Olympiad of Informatics. The qualification contests are conducted over the Internet. A practice I think I like given that it offers more students the ability to compete as traveling to a central location isn't always feasible. However, the contest appears to be purely a test of programming. I investigated the online training website and the problems seem to be in the vein of, here's some input data, some output data, make 'em match. While it's everybody can participate nature is beneficial and online training materials are valuable, I feel this contest doesn't help us to move away from the computer science == programming paradigm. Additionally, skimming over quickly it looks like only 2 out of 68 students sent to the international competition from the USACO were women.

The ultimate goal of the USACO of course is the International Olympiad for Informatics which has apparently been operating since 1989. A skim shows quite a few women dotted amongst the crowds of competitors although without official designation of the competitiors I'm unsure just how prevalent they are.

Close to home, Slippery Rock University hosts an annual programming contest modeled off of the official ACM Intercollegiate Programming contest. Specifically, this targets the team programming disciplines that is touted by Meszaros and Kahle '07, Hazzan and Dubinsky '06, and many others. Unfortunately, it is a PROGRAMMING contest meaning the importance of the actual algorithms and analysis sometimes gets lost. The fact that the strategies page has a humongous section on I/O and a couple bullets about worrying about the actual problems suggests that algorithms isn't necessarily where the focus lies.

One final competition I looked at was the American Computer Science League competition. The first thing I have to mention about this is the fact that they include a short answer portion. If nothing else this contest I feel is valuable because it gets computer science off of the computer. That may have something to do with the fact that this is the only competition for "computer science" as the others have been for programming/informatics but I don't think this is a trivial difference. The problems featured in the sample problems aren't the most fascinating questions in the world (e.g. what does this program do?) but it's a start. The competition also includes a programming portion although with much misfortune the programming portion seems to be a solo exercise.

So none of these programming contests perfectly fit the ideal model I had in my head for a computer science competition but the fact that something exists I do feel is something positive and each of the contests offers a little bit of insight. I appreciate that there are competitions taking advantage of the agile programming and teamwork models to show how computer science is a team building exercise. I appreciate that there are contests that offer a written portion to show the idea that computer science can happen just as easily away from the computer as it can on it. I appreciate the ability to model interdiscplinary problems to promote a computer science way of thinking. If there is someplace we can combine all three in a way that is interesting, fun, and valuable to college applications is a question I'm going to continue asking.

Math Contests -> CS Contests?

This week I dived into a couple papers included in the book "Reconfiguring the Firewall" by Burger, Cremer, and Meszaros.

The first paper "Revisiting Culture, Time, and Information Processing Factors in Connecting to Girls' Interest and Choice of an Information Technology Career at the Secondary Level" by Meszaros and Kahle helped to really set the tone for my entire thought process this week. Specifically the line that reached out the strongest to me was concerning young women's decision not to take CS in high school was"computer science was either not offered or not considered necessary for acceptance to highly selective colleges." (45) The reason this line speaks out so strongly is the consistent narrative present in a significant portion of college computer science students. Specifically, the computer science courses at the high school level are few and the courses that are robust or go beyond programming are even fewer. Competition for positions in elite collge programs are getting worse all the time. No longer is a 4.0 GPA enough, you have to go above and beyond and GPAs start to approach 4.5. You can't just be valedictorian of your school you have to be valedictorian of the state. At some point the top of the class starts to look like a monolithic block and it becomes difficult enough to tell anybody apart as it is.

In such an environment, the high school computer science curriculum has no place. Too many narratives of inadequate cs courses abound. Much credit is due to the computer science teachers who are working to change that and make the coursework a valuable part of a college application. At times this argument feels like a backlash to Margolis-Fisher due to the removal of computer science background from admissions criteria but at the same time, the high school computer science courses that were being offered before Margolis-Fisher weren't any more gender balanced. I don't believe reversing the critiera is going to suddenly raise the national quality of computer science courses. What this does make me wonder though is how we can make computer science a valuable part of high school student's college applications?

From this springs one item which is undoubtedly present on countless members of CMU's computer science class. The heralded math competition permeates many college students applications and are well regarded across the nation. From the school to the state to the national level math contests continue to shape primary through high school education. In 4th to 6th grade I personally competed in math olympiads, 7th to 9th I was invested in Math Counts, of which the national competition is broadcast on ESPN. 10th to 12th I was a regular top finisher in the state math contest. Not a year of my adolescence when by when I wasn't competing in some form of math competition.

That being said, what exists in the same vein for computer science? Certainly they're not as apparent as the math competitions for a variety of reasons but what is out there and what can we do to increase their visibility? By establishing competitions we get around the "full schedules" Meszaros and Kahle cite as a reason students don't take computer science, give students something they can do to beef up their college applications, and most importantly increase the visibility of computer science in those critical pre-college years.

This little light bulb has highlighted some questions I intend to look at in the upcoming weeks. What content would be on such an exam for middle schoolers? What about high schoolers? Can we insert the agile programming and pair programming methods cited repeatedly as instances of the need for teamwork in computer science into the competition? What computer science contests already exist locally? Nationally? Internationally? What is the interest level for such a competition? Should we set quotas to ensure teams are gender diverse? What about the current student body, have they participated in computer science competitions? Moreover, did they participate in math competitions? Are these competitions remembered fondly? Are they part of the reason they became computer scientists?

Progress on Background Research

Since our first research team meeting, last Thursday, I have been constructing a reading list and reading / taking notes on some papers for background research. I will post some thoughts on my readings in this blog, but maintain more detailed notes on copies of the papers. Here are some of my thoughts on some of the papers I have read over the past week.

I read "An Expanding Pipeline: Gender in Mauritius" by Adams, Bauer, and Baichoo in which the authors examine the trend of increasing percentages of women in CS at the University of Mauritius and hypothesize about cultural differences between Mauritius and other countries like the U.S. which have a trend of decreasing percentages of women in CS. I thought that one of the most interesting quotes from the paper was the following: "The growing percentage of female CSE students plus the low and declining percentage of female CSE instructors contradicts the hypothesis that women require female academic role models to be attracted to computer science." At our last meeting, we discussed the possible influence of role models in attracting students to computer science. The paper only discuss professors at the University of Mauritius as potential role models and not any women or men in industry who may be role models to pre-college students which may influence students' decisions to study CS. I would be interested to learn if there were any cultural role models in the field of CS in Mauritius.

Another paper that I read was "Gender Gap in Computer Science Does Not Exist in One Former Soviet Union Republic: Results of a Study" by Gharibyan and Gunsaulus which examined women's involvement in CS in Armenia. I did not find this paper as informative as other papers, but this may be due to the fact that this paper was based on only one part of a study with several parts left to be conducted.

The paper seemed very informal and not terribly scientific when analyzing cultural influences. For example, one sentence from the paper reads, "Armenians are very realistic and reasonable in almost all aspects of their lives, including the planning of their future." Besides quotations like that, there were interesting conclusions drawn from the research done by Gharibyan and Gunsaulus which support some of our hypotheses such as "CS is viewed as more closely related to Math than Engineering in countries where women's representation is higher in the field." Other interesting observations the authors had were that women and men had similar motivations and influences when choosing a major and absence of role models and male dominated fields are not intimidating to women.

One thing that I found very interesting was the fact that for all of the 1980's and 1990's the percentage of women in Computer Science never fell below 75% at Yerevan State University in Armenia. However, it has fallen to 44% in recent years. The authors attribute this to growing popularity with men and not falling interest with women. But my question is, why isn't Computer Science growing in popularity with women as well?

I have been reading various papers in Frontiers: A Journal of Women Studies, Volume 26, Number 1, 2006. It is a special issue on gender, race, and information technology. I have started reading "Unlocking the Clubhouse: Women in Computing" by Jane Margolis and Allan Fisher on Google books, but there is only a limited preview, so I am waiting for my copy to come in from Amazon to finish it! I have also been constructing a reading list (of more papers / books that I need to read), which I hope to put up here in my next post.

In addition to reading papers, I have been thinking a lot about the questions that we will want to put on our survey. We don't want the survey to be too long because it might discourage participation, but we also want it long enough that we can gain useful information.

At our meeting last week, we also discussed having a separate survey for freshmen. At this point in the school year, freshmen have only 2 weeks of experience to answer questions like, "Do you feel that you fit in at CMU, specifically within the CS department?" While they have limited experience with living / fitting in at college, freshmen can answer certain questions that will benefit our study. For example, in studying cultural attitudes towards computer science, a natural question to ask people in computer science is, "Why did you decide to major in Computer Science?" A freshman may provide a more unbiased answer to this question than a senior - while the senior may know why he or she has remained in the field, he or she may not remember as clearly as the freshman, exactly why he or she had initially decided to study computer science.