7

Our (American) university has a lot of applications coming in from Asia- especially China, India, Japan and Korea.

Each of these seems to follow a different grading scheme and it is hard to do apples-to-apples comparisons of students scores across universities. We don't request WES transcripts.

What is a good way to normalize students across Asian universities? Which ranking schemes (global and local) are reliable? Do these ranking schemes provide fine-grained, easy-to-parse data that we can use for CS and ECE admissions (not just relative ranking)?

ADDENDUM: There seems to be an XY problem with the original question, as @FedericoPoloni points out. What we're interested in is indeed-- "What is a good way to compare transcripts for graduate admission, especially from Asian universities? Admittedly, this is a broad problem and solutions may be reductive or simplistic, but it seems unfair to not consider the quality of teaching, research and the cohorts worked with, when evaluating student transcripts.

The current answer and comments, while helpful, will leave admissions restricted to IIT, Tshingua, KAIST-- since we do not trust transcripts from other universities, and though exceptions can be one-off, this approach does not scale. I'd be grateful if you could share references to meaningful datasets and ideas to use the data without unfair bias.

Improve this question
11
  • 9
    The educational system in India is sooooo different from, say, China; I don't think any general "Asian" ranking will be useful to you. (Also see: Is it ethical to apply different criteria for graduate admissions based on country of undergraduate study?)
    – ff524
    Dec 21, 2016 at 20:18
  • 1
    @ff524, more local rankings would be great too, if they are reliable. At some point, we do have to normalize expectations for students from different countries. Even something simple like "acceptance rate" for particular programs-- say .0005% for Tshingua v/s 0.001% for IIT-B...
    – Jedi
    Dec 21, 2016 at 20:34
  • 4
    "Reliable" and "university ranking" are concepts that usually do not go along well. Maybe you should demote that aspect and ask directly "What is a good way to compare transcripts and student scores for graduate admission, especially regarding Asian universities?"
    – Federico Poloni
    Dec 21, 2016 at 21:12
  • 1
    @FedericoPoloni I'm fine if this question is read in that way. Alternatively, if anyone is involved in CS/Engg admissions for international students, "What approach and data do you use to admit fairly across nations?"
    – Jedi
    Dec 21, 2016 at 21:42
  • 2
    The approach we use for MS admissions in my ECE department is, "Looking at data from our enrolled students, identify characteristics of students from X that seem to predict success/failure in our program. Then for students from X, admit those with "success" characteristics and reject those with "failure" characteristics." Where X may be a university, a group of universities, or something else (depending on the data we have). For students whose background is unlike anything we've seen before, we make an educated guess as to whether they'll do well in our program.
    – ff524
    Dec 21, 2016 at 22:23

1 Answer 1

Reset to default
12
+50

I believe that using university ranking as a criteria for admission is fundamentally unethical.

Extremely good students end up going to middling or poor undergraduate institutions for any number of reasons. To downgrade somebody's application just because they went to a mediocre institution is actively going against the academic ideal that merit should be able to rise, and even poor institutions generally have at least a few students of excellent potential. Complementarily, pedigree is no guarantee of quality: rubbish students graduate from elite institutions all the time.

Now, there is, of course, a strong correlation between the quality of an institution and the quality of the students that it produces, and there are, of course, basic matters of accreditation: nobody should respect a degree from Trinity Southern University. In general, however, you should judge the students and not the institutions.

So, how do you judge a student from Asia? The same way that you judge a student from an unfamiliar university in Europe or North America, by the accomplishments for which they can give evidence. To judge classwork, one of the methods is to ask which textbooks were used (this is what some MIT departments do). To judge other work, look for the concrete products, such as research papers, code or web pages online, etc.

Students from universities you do not know will still be at a disadvantage, and doubly so for students from a very different environment and who may lack English communication skills, such as will often be the case for many Asian universities. But you can at least take steps to not prejudice yourself unduly by sorting their universities by their ability to game semi-arbitrary ranking systems.

Improve this answer
3
  • The goal is indeed to judge the student not the institution. You err in assuming that we are not evaluating research, their Github repositories, open source contributions etc. . However, a narrow subset of our student ranking also requires evaluating sheer academic performance in their field of study. All else being equal, would you prefer someone from the top of the CS class in Bombay University, or a top 20% student from KAIST?
    – Jedi
    Dec 22, 2016 at 1:40
  • 2
    @Jedi In your example, I would consider both students to have satisfied the basic requirement for academic performance. You can't really know much more than "appears qualified on paper" from pure academic ranking in any case. If their outside-the-classroom record is also qualifying, then the final differentiation and decision of one over the other would come from interviews or the like.
    – jakebeal
    Dec 22, 2016 at 2:00
  • 2
    Based on our experience we would not consider that both students will cope with the rigors of our university. Our experience shows that a 80th-percentile student from KAIST is significantly better prepared for grad school here compared to a 99th percentile student from Bombay University. However this has led to us restricting admissions to a few dozen "safe" universities across Asia.
    – Jedi
    Dec 22, 2016 at 15:05

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .