Teaching Evaluations: Nationality Discounts and Premia?

19 August 2006 at 7:46 am Nicolai Foss 3 comments

| Nicolai Foss |

As is, I suppose, the case with most of the readers of this blog, I am subject to the discipline of student evaluations. I tend to find them pretty useless because their information content is rather low and because the whole process is very noisy and biased, although I do admit that they are a powerful tool for getting rid of teachers who are placed at the left tail of the quality distribution (let me anticipate a possible misunderstanding: I am usually rated in the opposite end of the distribution).

Here is a possible example of bias: I have often observed, and so have many colleagues with whom I have discussed the matter, what seems to be a nationality premium.Specifically, my School often flies in foreign teachers, many from the US, and I seem to detect a pattern wherein US teachers in particular are rated higher than what their actual teaching performance would seem to warrant. Students may think that it is cool per se to be taught by an American, and thís may be reflected in evaluations. Conversely, there may be nationality discounts (although I cannot offer any examples, and of course doing so would be highly politically incorrect).

In principle, this is researchable. Data exist for the nationality of those teachers who are evaluated by students. Of course, inherent quality needs to be controlled for. This may perhaps be done by using rankings of B-schools and fitting the rank of the relevant’s teacher’s homeschool into the model.

Any apriori thoughts from our readers before I embark upon this ambitious project? ;-)

Entry filed under: - Foss -, Ephemera.

Roundup of Interesting Links Scientific Progress in Strategic Management Bleg

3 Comments Add your own

1. Jung-Chin Shen | 19 August 2006 at 10:45 am

As a reciprocity behavior (of not giving tough questions at AOM), I guess I should contribute a bit here. ;-)
It might be interesting to consider both beauty premia and national premia at the same time in order to explore how genetic and artificial factors affect people’s evaluation. It may contribute to both areas, and provide some interesting thought on an economist’s view of social construction. :-)
2. Hakan Ener | 19 August 2006 at 10:54 am

This is a research topic that may have potentially significant interpretations if, indeed, a national premium is found. For instanc, recruitment decisions may be altered as a result of the findings.
Therefore, I suggest that the research design contain multiple detailed aspects of classroom performance, such as the mother tongue of professor, the nature and quantity of real-world examples mentioned, number of words spoken by the professor, proportion of speaking done by students versus professor etc. Without these detailed metrics, the research design cannot tackle the endogeneity of who gets hired where befoire being reflected in teaching evaluations.
Ratings on visiting scholars with a control on the ranking of home institutions sounds like a good start, but needs to be complemented with many classroom metrics in order to isolate the true effect of potentially significant nationality premia.
3. Peter Klein | 19 August 2006 at 11:17 am

Here in the US, there tends to be a discount for foreign-born instructors — though I suspect that English-language proficiency, not nationality, is the determining factor.