Correlations, trends and potential biases among publicly accessible web-based student evaluations of teaching: a large-scale study of data


- Estrelas (0)


Dono: Marcos Pedroso

Versão: 1.0

Última Atualiz.: 03-12-2018 13:14


Student evaluations of teaching are widely adopted across academic
institutions, but there are many underlying trends and biases that can
influence their interpretation. Publicly accessible web-based student
evaluations of teaching are of particular relevance, due to their widespread
use by students in the course selection process and the quantity of data
available for analysis. In this study, data from the most popular of these
websites,, is analysed for correlations between
measures of instruction quality, easiness, physical attractiveness, discipline
and gender. This study of 7,882,980 RateMyProfessors ratings (from 190,006
US professors with at least 20 student ratings) provides further insight into
student perceptions of academic instruction and possible variables in
student evaluations. Positive correlations were observed between ratings
of instruction quality and easiness, as well as between instruction quality
and attractiveness. On average, professors in science and engineering
disciplines have lower ratings than in the humanities and arts. When looking
at RateMyProfessors as a whole, the effect of a professor’s gender on rating
criteria is small but statistically significant. When analysing the data as a
function of discipline, however, the effects of gender are significantly more
pronounced, albeit more complex. The potential implications are discussed.

Comentários estão fechados.