The New York Times on cheating in university programming courses

Yesterday the New York Times published an article on cheating in university computer programming courses. To anyone who has taught a programming course, much of the narrative is probably all too familiar. Nonetheless, it is interesting to see it featured in mainstream media, it is helpful to hear about what is happening at different universities, and it may play a small part in refocussing the communities involved. This is important as the number of students taking computer science courses is increasing, as is their diversity (k-12, extracurricular, university non-majors, etc.). In addition, major efforts are under way to increase student diversity in terms of age, gender, ethnicity, socioeconomic background, and other ways. As a result of increasing student numbers and diversity, the numbers and diversity of teachers and institutions are also set to increase, making the issue of plagiarism one that is likely to continue to crop up in new and interesting ways.

The article goes into details on academic misconduct in programming courses, and methods of dealing with it, at Purdue, BerkeleyBrown, Stanford, Yale, and Harvard (including the popular CS50). It states that in some institutions, computer science courses have become some of the most common sources of academic honesty incidents, but also states that many computer science educators reject that their courses are any more prone to cheating than those in other disciplines. This view is supported by a quote from Alex Aiken, the head of computer science at Stanford who mentions the tendency for computer science educators to actually check assignment submissions for plagiarism as one of the reasons for the number of cases, along with citing recent high-profile cheating scandals in other disciplines that have received wide media attention.

The use of software-based methods of detecting plagiarism such as MOSS and Codio are discussed, and I suspect used fairly widely, but there was also mention of other more novel strategies to cope with plagiarism. One of these is the ‘regret-clause’ introduced in Harvard’s CS50 by David J. Malan, which lets students who cheat and admit it within 72 hours receive an unsatisfactory or failing grade on an assignment, and avoid further discipline – however repeat offenders don’t have this option.

Some of the other discussion surrounds the grey area between the collaborative ethos in programming where code sharing and reuse is commonplace, and policies on academic honesty. The problem is that many of the practices employed by software developers in the workplace violate academic honesty policies. So we are teaching students to work in highly collaborative environments, in educational environments where their collaborative abilities are restricted. A case in point is instructors encouraging students to work together, but demanding that assignment submissions be unique. This creates a grey area where guidelines on what is acceptable and what is not can become confusing to students. In the article, Paul North, a professor of German and chair of the executive committee at Yale acknowledges that compared to the guidelines “… the code to be written seems simple …”. It is my opinion that programming courses (or computer science departments) often need specific guidelines for their students. Trying to apply catchall policies and guidelines designed for any discipline is sure to lead to grey areas in computer science courses, particularly those that involve programming assignments.

It was interesting to read that many computer science professors are now delivering stern warnings at the start of each course – this is something that done at all of the institutions I have taught at. Although I didn’t think it was a unique practice it is interesting to hear that it is actually common. I wonder if computer science does this more often than other disciplines. My feeling would be yes, but I don’t know.

One thing is for sure – the ways that computer science educators deal with plagiarism need to be continuously adapting. Almost everything about computer science (particularly in the way it is taught and learned) is changing at a phenomenal pace and that isn’t showing any signs of abating.

Computer science in China: high employment, highest satisfaction and salaries

Each spring I teach on University College Dublin‘s Software Engineering and Internet of Things Engineering Degrees in Beijing. The degree is offered by the Beijing Dublin International College (Chinese language link), a recognised college of UCD and the Beijing University of Technology.

Back in October, I reported on China topping the list of best programming nations. As I am shamefully monolingual (course delivery is in English), my news consumption is limited to English, and it’s not every day that I see English language reports on the state of computer science in China. That report back in October was the last I think I came across until today, when I was pleased to learn that Computer Science is the second most fruitful degree in China in terms of employment prospects from this article.

For those with a bachelor’s degree in computer science, the employment rate was 93.9 percent in 2016 – slightly trailing electrical engineering’s 95.5 percent, according to a report published by MyCOS, a company specializing in higher education data analytics.€

Computer Science also came out on top in terms of income potential, with computer science graduates earning 5,452 yuan ($792) a month, half a year after graduation. This is over 800 yuan ($116) more than finance grads, who came in second. After three years at work, computer science graduates are earning 8,665 yuan ($1,257), almost 1,000 yuan more than finance graduates.

Refreshingly, Chinese computer science graduates are also most satisfied with their jobs, with 75% reporting job satisfaction six months after graduation.

In Ireland, computer science/ICT grads also earn the most of any discipline, with 62% earning €29,000 a year or more which equates to 18,629 yuan per month. This data is for 2014 graduates, so most comparable to the Chinese figure of 8,665 yuan per month meaning that Irish computer science grads are earning more than double their Chinese counterparts in China.

in the US, computer science grads with 1-4 years experience are earning $63,281 a year, or $5,273 (36,323 yuan) per month which is over four times the Chinese salary, and double the Irish salary.

Of course, this all only matters in when the cost of living is taken into account. I just had dinner and drinks with 5 people in Beijing and the bill was $8 each. However, the average rent in Dublin for a one bedroom apartment is well over $1,000 and in Beijing it is just under that (numbeo.com). Let’s hope that the graduates living in the big cities earn more than the average.

Irish computer science dropout rate falls

The number of students dropping out of Irish computer science courses before beginning their second year has fallen for the first time in several years.

This is welcome news, particularly as in recent years, Computer Science courses have had some of the worst progression rates of all courses. This year* their progression rate has risen from 80% to 84% for honours degree (level 8)** programmes. For university courses the figure is 88pc, compared with 80pc at institutes of technology.

The problem of high dropout rates in Computer Science courses is not limited to Ireland, and has been a big problem in many, if not most countries.

Overall, there has been a small decrease in drop-out rates, although more than 6,200 (nearly one in seven) first year students in 2013/14 did not continue into second year. Interestingly and encouragingly, students from farming backgrounds, and female students, are least likely to drop out before second year.

The source of these figures is a newly published report from the Higher Education Authority (HEA), A Study of Progression in Irish Higher Education, 2013/14 to 2014/15. Figures in the report show that 85% of first years in 2013/14 progressed to second year, up from 84% in the previous year.

HEA chief executive officer Dr Graham Love attributed the improved progression rates in computer science to additional funding for retention initiatives such as maths enabling courses, peer mentoring and tutorials.

*The report detailed here is for students who began their course in 2013/2014.

**For international readers, “honours degree, level 8” courses are ‘traditional’ degree courses, typically 4 year BSc courses. See here for more.