Americans are justly proud of their system of higher education. This is not surprising since it’s generally held in higher esteem than our K-12 schools whose performance, we are told constantly, lag those of peer countries. For instance, US institutions of higher learning in the United States generally dominate the top spots in most international rankings of universities. Something to cheer about, right?. Or is it?
The problem is that the metrics used by US News & World Report, the Times, the Shanghai Rankings et al to judge the quality of universities exclude the most fundamental one, namely, the quality of teaching, especially in subjects relating to Science, Technology Engineering and Math? The answer to how well they teach is arguably a better determinant of whether the world class reputation of US universities is deserved.
An international research team headed by Prashant Loyalka of the Stanford Graduate School of Education went in search of an answer, or a partial one anyway, insofar as it applies to computer science programmes from four countries which, together, produce 50% or more of the world’s computer science graduates annually. The results were presented in a paper published in the Proceedings of the National Academy of Sciences earlier this year. A summary by the lead researchers is here.
In brief, a statistically valid sample of final year undergraduate CS students in the US, China, India and Russia at a cross section of universities and colleges in each country, took a 2 hour standardized test developed by the Educational Testing Service, a non-profit testing and assessment organization. The test was designed to determine how well students understood computer science concepts and principles and was administered in the language of each student. In order that the US results were not skewed by international students attending its universities, the test was only administered to students who gave English as their sole or primary language.
The results were stunning to say the least. American students in regular college and university programmes far outperformed students from the other three countries also attending regular schools. In fact, these American students by and large performed on a par with Chinese, Indian and Russian students enrolled in elite schools. Meanwhile, the average American student enrolled in an elite CS programme in the US far outperformed students at Chinese, Indian and Russian elite schools. The difference in performance between students from the other three countries was statistically insignificant.
Even more impressive is that freshmen American students generally entered their CS programmes with less math and science preparation than their peers in China and Russia and about the same as Indian students. All of which suggests that US undergraduate computer science programmes hold a strong qualitative edge and add substantial value compared to those of the other nations. Interestingly, the worst performers were China’s universities, often flagged as emerging power houses, since their students started their CS programmes better prepared in math and science than the others yet performed no better than Indian and Russian students and well short of Americans by the end.
This is very good news, at least for the parents of US students who are studying computer science since it suggests they are getting their money’s worth. The study also has labour market implications since it indicates that American CS graduates from regular undergraduate programmes only have direct competition from students in the other three countries who attended elite schools.
However, some caveats apply. This is just one STEM subject and more research is needed to examine whether American students receive the same high quality instruction in other programmes (one of the subjects of the team’s continuing research). But for now at least, if not three cheers then certainly one and a half are warranted.