Sex Roles, Vol. 53, Nos. 11/12, December 2005 (
Student Evaluations and Gendered Expectations:
What We Can’t Count Can Hurt Us
and Kelley Massoni
Does teacher’s gender impact students’ evaluations? We critically evaluated the research lit-
erature and concluded that the form gender bias takes may not be easily detectible by quan-
titative scales. To explore this possibility, we did a qualitative analysis of the words that 288
college students at two campuses used to describe their best- and worst-ever teachers. Al-
though we found considerable overlap in the ways that students talked about their male and
female teachers, we also saw indications that students hold teachers accountable to certain
gendered expectations. These expectations place burdens on all teachers, but the burdens on
women are more labor-intensive. We also saw signs of much greater hostility toward women
than toward men who do not meet students’ gendered expectations.
KEY WORDS: teaching evaluations; student evaluations; gender discrimination.
Administrators place increasing emphasis on
student evaluation of professors in making person-
nel decisions. For example, in a nationwide survey of
administrators of accredited, 4-year, undergraduate,
liberal arts colleges in 1998, 88.1% reported that they
“always used” systematic student ratings of teaching
in evaluating faculty; in earlier surveys, fewer admin-
istrators gave this response: 80.3% in 1988 and 54.8%
in 1978 (Seldin, 1999).
Reviews of the research literature on the stu-
dent evaluation of college teaching could easily give
the impression that gender is not a major factor in
the process (e.g., Aleamoni, 1999; Feldman, 1992,
1993; Fernandez & Mateo, 1997; Freeman, 1994;
Wheeless & Potorti, 1989). In fact, a respected
guide for administrators on how to evaluate teaching
speciﬁcally recommends that they not take gender
into account in interpreting student ratings; the
authors concluded that research has revealed that,
An earlier version of this paper was presented at the annual
meetings of the American Sociological Association, Chicago, IL,
Department of Sociology, University of Kansas, Lawrence,
To whom correspondence should be addressed at Department of
Sociology, University of Kansas, 716 Fraser Hall, Lawrence, KS
66045; e-mail: firstname.lastname@example.org.
if anything, students rate women higher than men
(Cashin, 1999). The only citation to support this
claim is two papers by Feldman (1992, 1993) that
report separate meta-analyses of experimental
studies and of studies based on students’ evaluations
of actual professors. Feldman concluded that direct
effects of gender on evaluation were minor, trivial
in size, and, in the case of evaluations of actual
professors, favored women over men.
We argue that a more careful reading of the re-
search literature reveals that the evidence is mixed.
Meta-analytic strategies such as Feldman’s (1992,
1993) may obscure more than they reveal. Miller and
Chamberlin (2000) argued that Feldman’s method-
ology may have depressed ﬁndings of gender effects
because he combined information from qualitatively
different sources. For example, he averaged corre-
lations between gender and student rating of the
professor that are quite disparate, and he aggregated
ratings from studies that vary by discipline, type of
institution, and unit of analysis.
In a review of the research, Aleamoni (1999) re-
ported that “a majority of studies” (p. 156) ﬁnd no re-
lationship between either teacher gender or student
gender and student ratings. But the claim of no gen-
der difference was weakly supported with a reference
to seven studies cited in an article published in 1971.
2005 Springer Science+Business Media, Inc.