Student Evaluations of College Professors: Are Female and Male Professors Rated Differently?

Abstract

Over 1,000 male and female college students of 16 male and female professors (matched for course division, years of teaching, and tenure status) evaluated their instructors in terms of teaching effectiveness and sex-typed characteristics. Male students gave female professors significantly poorer ratings than they gave male professors on the six teaching evaluation measures; their ratings of female professors were poorer than those of female students on four of the six measures. Female students also evaluated female professors less favorably than male professors on three measures. Student perceptions of a professor's instrumental/active and expressive/nurturant traits, which were positively related to student ratings of teaching, accounted for only a few of these gender-related effects. Student major and student class standing also played a role in the evaluation of professors. The importance of gender variables in teacher evaluation studies is discussed, and implications for future research are noted. (PsycINFO Database Record (c) 2012 APA, all rights reserved)

Full text

Journal of Educational Psychology
1987,
Vol 79, No. 3, 308-314
Copyright 1987 by the American Psychological Association, Inc.
0022-066 3/87/$OO. 75
Student Evaluations of
College
Professors: Are Female
and Male Professors Rated Differently?
Susan A. Basow and Nancy T. Silberg
Lafayette College
Over 1,000 male and female college students of 16 male and female professors (matched for
course division, years of teaching, and tenure status) evaluated their instructors in terms of
teaching effectiveness and sex-typed characteristics. Male students gave female professors
signif-
icantly poorer ratings than they gave male professors on the six teaching evaluation measures;
their ratings of female professors were poorer than those of female students on four of the six
measures. Female students also evaluated female professors less favorably than male professors
on three measures. Student perceptions of a professor's instrumental/active and expressive/
nurturant traits, which were positively related to student ratings of teaching, accounted for only
a few of these gender-related effects. Student major and student class standing also played a role
in the evaluation of professors. The importance of gender variables in teacher evaluation studies
is discussed, and implications for future research are noted.
Research since the 1960s
has
documented prejudice against
women, particularly if women violate gender stereotypes, for
example, by having gender-atypical characteristics or by par-
ticipating in gender-atypical professions (Etaugh & Riley,
1983;
Paludi & Bauer, 1983). Because college teaching is
considered a high-status male occupation (Touhey, 1974) and
because evaluations made by others influence advancement
in such a career, it is important to determine if any biases
exist in the evaluation of
college
professors.
Most investigations of bias in the evaluation of professors
have produced conflicting results. However, two variables
that appear to be important are professor sex and professor
sex typing. Although some studies have found relatively few
or no differences in the evaluations of male and female
professors on the basis of sex alone (Basow & Distenfeld,
1985;
Bennett, 1982; Elmore
&
LaPointe, 1974, 1975), others
have found a sex bias (Kaschak, 1978; Lombardo & Tocci,
1979).
However, the nature of this bias seems to depend on
student sex, the type of questions asked, and specific teacher
qualities. For example, Kaschak (1978) examined student
evaluations of professors' teaching methods as a function of
teacher sex, student sex, and academic field (masculine, fem-
inine, or neutral). She found that female students rated female
and male professors as equally effective, concerned, likable,
and excellent. In contrast, male students showed a consistent
bias in favor of male professors. Both sexes, however, were
more willing to take a course with the same-sex instructor.
Lombardo and Tocci also found that male instructors were
perceived as more competent than female instructors by male
Portions of this research were presented at the meeting of the
Eastern Psychological Association, Boston, March 1985. We would
like to express our gratitude to the participants in this study as well
as to the anonymous reviewers of this article for their helpful com-
ments.
Nancy T. Silberg is now at the University of Vermont.
Correspondence concerning this article should be addressed to
Susan A. Basow, Psychology Department, Lafayette College, Eas-
ton, Pennsylvania 18042.
students, whereas female students showed no sex bias. In a
second study, Kaschak (1981) found that most of the sex bias
disappeared when the professors were described as award
winning. However, male professors were still described as
more powerful and effective than female professors, and
female professors in "feminine" fields were rated as more
concerned and likable than male professors in those fields.
These results suggest that gender-stereotypic qualities play a
role in the evaluation of
college
teachers.
In an effort to understand better the influence of gender-
stereotypic qualities on teacher evaluations, researchers have
explored the effect of professor sex typing on student ratings
of teachers. The findings, however, are unclear because in
some cases professor sex typing interacts with student sex or
professor sex. Furthermore, studies have varied considerably
in methodology, with some using written descriptions and
some using videotaped or real-life classroom instructors. In a
study using written profiles of male and female teachers that
included adjectives described as stereotypical l y masculine or
feminine, Harris (1975) found that both male and female
students rated teachers with "masculine" teaching styles more
positively than those with "feminine" teaching styles on all
measures except warmth. In a second study using a similar
design, Harris (1976) reported that masculine-stereotyped
teachers were evaluated more favorably at all grade levels
(nursery school through college) on all measures except
warmth and superiority. For warmth and superiority, profes-
sor sex typing interacted with student sex. Students of each
sex preferred the teacher who possessed the traits stereotyped
as appropriate for their own sex.
Expressiveness is another sex-typed trait that affects evalu-
ations of teachers, although research on teacher expressiveness
has sometimes tapped the "masculine" aspect related to dy-
namism and sometimes the "feminine" aspect related to
warmth. In general, the expressive teacher is rated more
positively than the nonexpressive teacher (Abrami, Leventhal,
&
Perry,
1982;
Basow
&
Distenfeld,
1985;
Elmore
&
LaPointe,
1975;
Marsh & Ware, 1982; Ware & Williams, 1975). How-
ever, a number of methodological issues arise from this line
308

STUDENT EVALUATIONS
309
of
research.
Not only has the definition of teacher expressive-
ness and the way in which the teacher is presented varied
from study to study, but most of
this
research has used male
instructors. In the few studies that have examined teacher
sex,
interactions between teacher expressiveness and teacher sex
have been found. However, the nature of these interactions
appears to depend on the precise definition of expressiveness/
warmth and the particular questions asked (Basow & Disten-
feld, 1985; Elmore & LaPointe, 1974, 1975). For example,
Basow and Distenfeld found that expressiveness was most
important for male teachers on two factors—organization and
stimulating student interest.
Another problem with studies that have examined the
influence of sex-stereotyped qualities on performance ratings
of teachers may be that these studies deal with only one
stereotyped quality at a time. Because people can be high in
warmth/expressiveness and either low or high in instrumen-
tal/active qualities (Bern, 1981), the influence of a combina-
tion of these personal qualities could be expected. In a labo-
ratory study, Basow and Howe (1987) addressed this issue
and found that both instrumental and affective qualities are
important in evaluating professors but that affective qualities
are more important. These results were supported in a recent
field study by Erdle, Murray, and Rushton (1985), who found
that both achievement-related personality traits (such
as
dom-
inance) and interpersonal-related personality traits (such as
supportiveness) were related to global student evaluations.
However, nearly all the faculty used in this study were male.
Bennett (1982) investigated the effect of both sex and sex
role variables on teaching evaluations in a study that used
student ratings of actual teachers. The results suggested that
sex role stereotypes influence evaluations of female professors
but not those of male professors. According t o Bennett's study,
greater demands for student contact and support are placed
on female professors than on male professors, and female
professors' higher ratings of w a r m t h , potency, or both account
for their more positive evaluations on interpersonal aspects
of teaching. In addition, a highly structured instructional
approach is more important for teaching performance ratings
of female professors than of male professors. Bennett did not
find evidence of direct sex bias in student evaluations of
teachers. Bennett's study is important because it provides
information concerning the effects of
sex
and sex-stereotypic
variables on teaching evaluation in an actual classroom situ-
ation. Nevertheless, methodological differences make com-
parisons with other research in
this
area difficult. For example,
unlike other research, Bennett's did not directly measure
professor sex typing. Moreover, results from Bennett's study
as well as those from other studies (e.g., Elmore & LaPointe,
1974,
1975) were based on students' responses to individual
items of a teacher rating form. It has been suggested that
because teaching effectiveness is multifaceted, evaluation in-
struments that have a well-defined factor structure will pro-
vide more information about teaching effectiveness than
would responses to individual items (Marsh, 1984).
Given the confusion in the literature, further examination
of the effects of sex and sex typing on ratings of college
professors is warranted. It is particularly important to clarify
how these variables operate in an actual classroom setting. To
this end, the present study used a field design to determine
the effects of professor sex and professor sex typing on male
and female students' evaluations of college professors.
On the basis of previous research, we hypothesized that
professor sex would interact with student sex (Kaschak, 1978;
Lombardo & Tocci, 1979) such that female professors would
be rated more poorly by male but not female students and
that professor sex typing would be a significant variable in the
evaluation of professors (Basow & Howe, 1987; Bennett,
1982).
We hypothesized that instrumental/active and expres-
sive/nurturant traits would affect teaching performance rat-
ings.
Method
Subjects
Over half the student body of a small selective private college in
the northeastern United States received the 1,293 questionnaires.
Because 213 questionnaires were eliminated due to incomplete re-
sponses, the final sample consisted of 553 men and 527 women.
Some students participated in the study more than once. The sex
ratio of the sample
(51 %
male and 49% female) was similar to that
of the college population (57% male), as were the class levels
(21%
freshman, 22% sophomores, 30% juniors, and 26% seniors). A full
range of major fields of study was present, with 48% of
the
students
majoring in the social sciences, 19% in the natural sciences, 14% in
engineering, 5% in the humanities, and 9% undecided. Engineering
majors were underrepresented, and social science majors were over-
represented in the sample. (At the college, about one third of the
students majored in engineering, and an equal number majored in
the social sciences). Men and women differed significantly in their
major field of study, x
2
(6, N = 1029) =
50.63,
p <
.001,
with more
humanities majors being female and more engineering majors being
male. This distribution of male and female students by major, was
characteristic of
the
college population.
Materials
Teacher
rating
form.
A teacher rating form (Leventhal, Perry, &
Abrami, 1977, adapted from Hildebrand & Wilson, 1970) was used
to measure students' perceptions of teacher appeal and teacher effec-
tiveness. The rating form is composed of 26 questions rated on a 5-
point Likert scale ranging from
well above average
(1) to
well below
average (5). Question 1 related to overall teaching ability ranging
from outstanding (1) to poor and inadequate (5). The remaining
questions were equally divided among five factors: Scholarship, Or-
ganization/Clarity, Instructor-Group Interaction, Instructor-Indi-
vidual Student Interaction, Dynamism/Enthusiasm.
Personality
inventory.
The short form of the Bern Sex-Role Inven-
tory
(BSR1;
Bern, 1981) was used to measure students' perceptions of
their instructors' instrumental/active traits ("masculine" traits, such
as assertive, dominant) and nurturant/expressive traits ("feminine"
traits,
such as warm, understanding) on two separate scales. Ratings
of the 30 items were made on a 7-point Likert scale ranging from
never or
almost
never
true (1) to
always
or almost
always
true (7).
Procedure
Each of the 22 female professors at the college who were at the
assistant professor rank or above and who had taught full-time at
least 1 year at the college was matched, if possible, with a male

310
SUSAN A. BASOW AND NANCY T. SILBERG
professor. Matches were made on the basis of rank, course division
(i.e.,
humanities, social science, or natural science), and years of
experience at the college to minimize differences between male and
female professors and prevent a selection bias. Of the 19 pairs of
professors whom
we
approached individually and asked to help us in
a research project on
students*
perceptions of teachers,
16
pairs (70%
of
the
female faculty and
13%
of the male faculty) agreed to partici-
pate.
There were no significant differences between the female and
male professors in the sample in terms of
age,
years at the college,
rank, or course division. The distribution of professors by the course
division in which they taught (28% in humanities, 47% in social
sciences,
and
25%
in natural
sciences) was
representative
o f the female
faculty but not of the
college:
Social science professors, who constitute
20%
of
the
faculty, were overrepresented, and engineering professors,
who constitute 20% of the faculty, were not represented at all. The
tenure status of the group
(25%
of each sex) was representative of the
tenure status of female professors at the college (26%) but not of the
male professors (60%).
Each professor was evaluated by his or her students during the 5th
or 6th week of a 14-week semester.
A
female experimenter distributed
the teacher rating form and BSRI in counterbalanced order to stu-
dents in 78 classes at the beginning or end of
class.
Each professor
was rated by an average of 33 students; only two professors (both
female)
were
rated by fewer than
10
students. The study
was
described
as an investigation of
students*
perceptions of their professors. Those
students
who
did not wish to participate or
who
had already evaluated
the particular professor were asked not to complete the question-
naires. Anonymity of
responses
was emphasized.
Results
A two-way multivariate analysis of variance
(MANOVA;
Teacher Sex
X
Student Sex) was performed by using individ-
ual student ratings on Question 1 (overall teaching ability)
and all five factors of the teacher rating form (see Table
1
for
summary table).
There was a significant overall Teacher Sex x Student Sex
interaction that is supported by all the univariate analyses
except for Instructor-Individual Student interaction. Because
all the dependent measures were intercorrelated (r ~ .S7-.88),
standard&ed discriminant function coefficients were calcu-
lated.
As
shown in Table
1,
Scholarship, Organization/Garity,
and Dynamism/Enthusiasm contributed most strongly to the
discriminant function for the Teacher Sex x Student Sex
interaction and in a direction opposite to that of Instructor-
individual Student Interaction. Despite statistical significance,
the effect sizes
(m
z
)
for these interactions are all less than .01,
indicating that less than
1%
of the total variance in the scores
is accounted for by the two-way interaction.
The means (see Table 2) indicate that male students gave
significantly different ratings as a function of professor sex,
whereas female students rated male and female professors
more similarly. Post hoc t tests on the interaction means
revealed that male students gave female professors signifi-
cantly (p < .05) less positive ratings than they gave male
professors on
all
dependent
measures.
Male students
also
rated
female professors significantly (p<
.01}
more negatively than
did female students on Scholarship, Organization/Clarity,
Dynamism/Enthusiasm, and overall teaching ability. How-
ever, female students rated female professors significantly
more negatively than they rated male professors on Instruc-
tor-Individual Student Interaction (p < .01), Dynamism/
Enthusiasm {p < .05), and overall teaching ability (p < .01).
The significant multivariate Teacher Sex x Student Sex
interaction clarified the significant multivariate main effect
for teacher sex (see Table I). The standardized discriminant
function coefficients indicate that Dynamism/Enthusiasm
and overall teaching ability contribute the most to group
separation and in an opposite direction than does Instructor-
Group Interaction, which
was
not significant in the univariate
analyses. Instructor-Individual Student Interaction, not sig-
nificant in the interaction test,
was
significant
as
a main effect.
Both male and female students gave male professors better
ratings on this measure. The effect sizes
(w
2
)
are all low, with
the strongest effects due to teacher sex occurring in scores on
overall teaching ability (accounting for 3.4% of the variance)
and Dynamism/Enthusiasm (3%).
Student sex was not a significant main effect in the multi-
variate analysis, although it was significant on the univariate
test of Scholarship (p < .05). This finding of better ratings
given by female students needs to be interpreted in the light
of the significant two-way interaction on this factor.
Because students* perceptions of their professors' instru-
mental/active and expressive/nurturant personality traits as
Table 1
Two-Way
MANOVA
Summary Table With
Effect
Sizes
(m
2
)
and Standardized
Discriminant Function Coefficients (Discrim.)
Multivariate*
Source
Teacher sex
Student sex
Teacher sex
x Student sex
F
10.40***
1.29
2.54*
F
<**
Discrim.
F
*
2
Discrim.
F
w
2
Discrim.
I
7.71**
.006
-.03
3.92*
.003
.92
9.67**
.008
2
10.63**
.009
.16
1.15
.000
27
9.33**
.008
-.44
Univariate"
3
1,44
.000
.63
0.45
.000
-.77
4.62*
.003
.09
4
9.H**
.007
.05
1.85
.001
.49
1.79
,001
.63
5
34.64***
.030
-.70
1.12
.000
-.40
8.62**
.007
-.56
Overall
39.71***
.034
-.81
3.72
.002
,79
7.10**
.006
_ 11
Note. Univariate analyses are on Scholarship, Organization/Clarity, Instructor-Group Interaction, Instructor-Individual Student Interaction,
Dynamism/Enthusiasm, and overall teaching ability,
MANOVA
= multivariate analysis of variance.
"4& = 6, 1071.
"4ft-
1,1076.
•p<.05.
**p<.0\.
•**/><•001.

STUDENT EVALUATIONS
311
Table 2
Mean Ratings and Standard
Deviations
of Male and Female
Professors
by Male and Female Students
Variable
Scholarship
Organ i z a t i o n / C l a r i t y
Instructor-Group Interaction
Instructor-Individual Student
Interaction
Dynamism/Enthusiasm
Overall teaching ability
Male student
Male professor
(if = 275)
M
11.7
10.7
11.8
10.7
9.6
2.2
SD
3.1
3.3
3.2
3.3
3.6
0.8
Female
professor
(n = 278)
M
12.9
12.1
12.5
11.6
11.6
2.6
SD
3.1
3.9
3.5
3.5
4.0
1.0
Female student
Male professor
(n = 284)
M
12.0
11.1
12.1
10.7
10.0
2.2
SD
3.1
3.4
3.2
3.6
3.8
0.8
Female
professor
(n = 243)
M
11.9
11.2
11.9
11.1
10.7
2.4
SD
3.2
4.2
3.6
3.7
4.1
1.0
Note.
Lower scores indicate more positive evaluations. Score range = 5-25, except for overall teaching ability, for which range = 1-5.
well as student class level all were significantly correlated (p
< .01) with the dependent measures, they were used as co-
variates in a two-way multivariate analysis of covariance.
Instrumental/active and expressive/nurturant scores signifi-
cantly correlated with all levels of the dependent variables (p
< .001) such that the higher the scores, the better the student
ratings. Class was significantly related to all dependent vari-
ables except Organization/Clarity (p < .087), such that stu-
dents gave increasingly better evaluations as they moved from
freshman to senior status. These covariates attenuated the
multivariate interaction effect (p = .081), although the uni-
variate interaction on Scholarship, Organization/Clarity, Dy-
namism/Enthusiasm, and overall teaching ability remained
significant atp <
.05.
The main effect for teacher
sex
remained
with the covariate analysis (p < .001), but only the univariate
analyses on Dynamism/Enthusiasm and on overall teaching
ability remained significant (p < .001 on both).
Because male and female students differed significantly in
their major field of study and because Basow and Distenfeld
(1985) found student major to affect some teacher ratings, an
exploratory three-way MANOVA (Teacher Sex x Student Sex
x Major) was calculated to throw further light on the Teacher
Sex x Student Sex interaction. Fifty-one students were elim-
inated because of their unspecified or unclassifiable majors.
Table 3 gives the number of students in each cell. Only one
cell, male humanities majors rating male professors, had less
than 10 subjects (see Table 3). A significant three-way inter-
action was found (Wilks's lambda = .96, p = .025) and
supported in two univariate analyses: Organization/Clarity,
Table 3
Sample Sizes as a
Function
of Student Sex and
Student Major
Male student
Female student
Variable
Humanities
Social sciences
Natural sciences
Engineering
Undecided
Male
professor
5
147
47
46
24
Female
professor
10
119
57
57
20
Male
professor
17
139
56
24
28
Female
professor
21
113
45
26
28
F\4,
1009) = 3.9, p = .004, and overall teaching ability, F\4,
1009) = 2.7, p =
.03.
The standardized discriminant function
coefficients for the first function indicate that Organization/
Clarity contributes the most
(-1.16)
to group separation and
in a direction opposite that of Dynamism/Enthusiasm (.66).
On the second function, overall teaching ability contributes
the most
(1.22)
to group separation and in a direction opposite
that of Organization/Clarity (-.83).
Table 4 gives the mean ratings and indicates that on the
two significant measures, male social science majors made
the largest distinction between male and female professors in
favor of the male professor. On Organization/Clarity, male
social science majors gave particularly positive ratings to their
male professors, and the most negative ratings received by
female professors were given by male engineering majors.
Female undecided majors also made a strong distinction
between male and female professors on this factor. On overall
teaching ability, male social science majors, as well as male
humanities and female engineering majors, made large dis-
tinctions between their male and female professors, rating
female professors more negatively. This interaction remained
significant even when student-perceived professor instrumen-
tal/active and expressive/nurturant scores and student class
were used as covariates in a three-way MANOVA (p = .016).
Also major was a significant main effect (Wilks's lambda =
.96,
p = .027), supported in the univariate analyses for Schol-
arship,
F\A,
1009) = 3.2, p = .012, and Instructor-Individual
Student Interaction, 7=1(4, 1009) = 3.5, p = .008. On all
measures, engineering majors gave the most negative evalua-
tions,
and humanities majors gave the most positive. Use of
the covariates, however, reduced the significance of major to
.076.
Discussion
Overall, results support predictions that professor
sex
would
interact with student sex on student evaluations of college
professors. As other research has found (Kaschak,
1978;
Lom-
bardo & Tocci, 1979), the significant interaction effect on all
measures except Instructor/Individual Student Interaction is
due to the consistently less favorable ratings of female profes-
sors given by male students. Nevertheless, the fact that all

312
SUSAN A. BASOW AND NANCY T. SILBERG
Table 4
Mean Ratings and Standard
Deviations
of
Professors
as a Function of Student Sex and Student Major
Variable
Scholarship
Humanities
Social sciences
Natural sciences
Engineering
Undecided
Organization/Clarity
Humanities
Social sciences
Natural sciences
Engineering
Undecided
Instructor-Group Interaction
Humanities
Social sciences
Natural sciences
Engineering
Undecided
Instructor-Individual Student Interaction
Humanities
Social sciences
Natural sciences
Engineering
Undecided
Dynamism/Enthusiasm
Humanities
Social sciences
Natural sciences
Engineering
Undecided
Overall teaching ability
Humanities
Social sciences
Natural sciences
Engineering
Undecided
Male student
Male
professor
M
11.2
11.4
11.6
12.4
12.0
11.6
9.7
11.1
12.4
U,3
12.4
11.3
11.1
13.3
12.3
11.8
10.0
10.5
11.8
12.3
12.0
9.1
9.3
10.6
10.2
2.0
2.0
2.1
2.4
2.5
SD
2.7
3.1
2.9
2.8
3,6
3.6
2.9
3.3
3.4
3.2
4.5
3.0
3.0
3.2
3.6
4.2
3.1
3.7
3.0
3.6
5.5
3.3
3.9
3.4
4.5
1.0
0.8
0.8
0.8
0.9
Female
professor
M
12.2
13.0
11.7
13.7
14.0
12.4
12.4
10.6
13.2
11.5
11.6
12,6
12.0
12.6
13.7
10.7
11.9
10.4
12.4
13.7
12.2
11.8
10.3
12.6
12.8
2,7
2.8
2.3
2.7
2.8
SD
1.7
3.0
3.0
3.4
4,1
3.1
3.8
4.4
3.5
4.0
3.1
3.4
3.7
3.5
4.2
3.6
3.6
3.5
3.2
4.2
4.7
4.2
3.5
3.9
3.2
0.9
1.0
1.0
1.0
1.0
Female student
Male
professor
M
10.8
12,1
12,2
12.8
11.4
9.6
11.5
11.6
ii.o
10.2
10.8
12.5
11.8
12.5
11.1
9.5
11.0
10.1
10.8
11.1
10.0
10.0
10.2
9.2
10.0
2.0
2.2
2.2
1.9
2.2
SD
2.8
3.0
2.8
2.8
4.0
3.7
3.2
3.6
2.4
3.8
4.0
2.9
3.3
2.9
3.2
3.4
3.5
3.8
3.9
3.2
4.9
3.5
4.4
3.4
4.1
1.0
0.7
0.7
0.6
0.9
Female
professor
M
11.4
11.5
12.2
12.6
12,7
10.1
11.0
11.1
11.3
12.6
10.7
H.6
12.6
12.5
12.2
9.8
1J.0
H.O
li.6
12.2
9.7
10.7
10.4
11.3
U.5
2.0
2.4
2.6
2.5
2.3
SD
2.5
3.3
3.0
2.8
3.4
3.6
4.3
4.0
3.5
4,7
3.4
3.5
3.6
3.2
4.2
3.9
3.2
4.2
3.5
4.2
3,8
4.3
3.8
3,4
4.3
0.7
1.0
0.9
0.9
0.8
Note. Lower scores indicate more positive evaluations. Score range = 5-25, except for overall teaching ability, for which range = 1-5.
professors were rated at least average on each factor suggests
that female professors are perceived as effective teachers,
despite being rated more negatively than male professors. The
pattern of results of this study is contrary to other field
research, which has found little evidence of differential ratings
of male and female professors (Bennett, 1982; Elmore &
LaPointe, 1974, 1975), perhaps because of methodological
differences. Unlike the present study, some studies did not
directly match their sample of male and female professors on
the basis of rank, course division, and years of experience at
the college {Elmore & LaPointe, 1974, 1975). Furthermore,
previous studies used different measures of teaching perform-
ance and surveyed different courses. For example, Bennett's
investigation was restricted to nonscience introductory
courses.
The fact that college teaching is considered a male occu-
pation may help explain why male students rate female pro-
fessors lower than they rate male professors. Less favorable
ratings of women are most likely to occur when women are
seen as not fitting gender stereotypes, in this case by partici-
pating in a sex-atypical profession (e.g., Etaugh & Riley,
1983).
If male students show more of a bias than do female
students, it may be because men tend to be more traditional
than women in terms of attitudes toward women and because
traditional attitudes toward women are associated with more
prejudicial attitudes (Holahan & Stephen, 1981).
Both male and female students, however, rated female
professors less favorably on Instructor-Individual Student
Interaction, which involves questions related to a professor's
availability to and contact with students. Because of gender
stereotypes, female professors may be expected to be more
accessible to students than are male professors (Bennett,
1982),
The lower ratings of female professors on Instructor-
Individual Student Interaction may be the result of their not
having conformed to this expectation. Another explanation
may be that female professors are in fact less accessible to
students than are male professors, although Bennett found no
evidence of this.

STUDENT EVALUATIONS
313
As
previous
research
h a s shown (Basow & Distenfeld, 1985;
Basow & Howe, 1987; Bennett, 1982; Erdle et al., 1985),
professor sex-typed personality characteristics are important
variables in the evaluation of
college
teachers. Both student-
perceived instrumental/active traits and student-perceived
ex-
pressive/nurturant traits significantly correlated with all the
dependent variables and attenuated the Teacher Sex x Stu-
dent Sex interactions somewhat. However, these traits did not
reduce the main effect for teacher sex, especially with regard
to Dynamism/Enthusiasm and overall teaching ability. The
results for Dynamism/Enthusiasm are surprising, because
Dynamism/Enthusiasm taps such qualities as dynamism, en-
ergy, enthusiasm, and confidence, which connote dominance
and authority and
are
similar t o the traits on the instrumental/
active scale. In the current study, male professors received
higher scores than did female professors from all students for
the instrumental/active traits, /(I, 1076) = 25.0, p < .001,
Ms
—
50.6 and 47.7, respectively. However, covarying the
effects of these traits did not eliminate the better ratings of
male professors compared with female professors on Dyna-
mism/Enthusiasm or any other measure. Thus, it may be that
the ratings on Dynamism/Enthusiasm, in particular, are
based not on these personality traits but on either differences
in gender-based expectations or differences in actual class-
room behavior. Erdle et al. (1985) documented the link
between different types of teacher behaviors and student
ratings but studied mainly male professors. It would be im-
portant to observe whether male and female instructors ac-
tually engage in different classroom behaviors and, further-
more, whether classroom behaviors relate to student ratings
in a different way depending on the sex of the professor and
the sex of the student. Even if there are actual sex differences
in classroom behavior related to dynamism and enthusiasm,
it should be recalled that male students still rated female
professors even lower on this variable than did female stu-
dents.
The lower ratings received by female professors, compared
with male professors, by both female and male students on
overall teaching ability may be attributed to the strong cor-
relation between overall teaching ability and Dynamism/
Enthusiasm (r = .88, p < .001). In fact, this correlation was
the strongest of all the intercorrelations among the dependent
measures. This explanation is supported by the Dr. Fox
studies (Abrami et al., 1982; Marsh & Ware, 1982; Ware &
Williams, 1975), which found that teacher expressiveness
strongly influences summary and global student ratings,
sometimes even more than does lecture content. Of course,
male professors may in fact be better overall teachers than
female professors. However, the evaluation of overall teaching
ability was based on one question, whereas ratings on the
other factors were based on five questions. Therefore, the
findings concerning overall teaching ability must be inter-
preted cautiously. In addition, male students seem to rate
male professors even more positively on overall teaching
ability than do female students.
Student major also has an effect on the evaluations of
professors, as Basow and Distenfeld (1985) found. Engineer-
ing majors gave the least positive ratings on each factor,
perhaps because no engineering professors were evaluated.
Because each discipline may vary with respect to the type of
teaching style used (Hudak & Anderson, 1984), engineering
majors may evaluate humanities, social science, and natural
science professors more negatively because they may be less
accustomed to their teaching styles or have less interest in
those courses than other students. Professor personality fac-
tors also may be important, because controlling for student-
perceived traits attenuated this main effect. However, the
covariates did not affect the significant three-way Teacher Sex
x Student Sex x Student Major interaction. On Organization/
Clarity and overall teaching
ability,
differential ratings of male
and female professors were given primarily by male social
science majors. On Organization/Clarity, male social science
majors gave their male professors particularly positive scores,
and male engineering majors gave their female professors
particularly negative ratings. Approximately
60%
of the social
science majors in this study were in economics/business.
Because economics/business and engineering are male-dom-
inated fields, students in these fields may have less interaction
with professional women, both at the college in this study and
in the business world. Men in these majors may also have
more traditional attitudes toward women, which may contrib-
ute to their lower evaluation of female professors, compared
with male professors (Holahan & Stephen, 1981). Further
research needs to clarify the effects of major on student
evaluations.
In sum, the present study supports the importance of sex
and sex role variables in the evaluation of
college
professors.
There are a number of cautions that should be noted, how-
ever. Although the differences found in this study were statis-
tically significant, the magnitude of the mean differences
actually was quite small and more easily discernible on the
composite scores and when broken down by student sex and
major. In practice, when student evaluations are used, student
information rarely is available, and scores are unlikely to be
summed into factor scores. Therefore, the subtle effect of
sex
bias may be hard to discern. Furthermore, the magnitude of
the effect sizes is quite small, indicating that sex of instructor
and sex of student account for only a small percentage of the
variance in student ratings. For example, the combination of
knowledge of teacher sex and student sex can predict only
about
4%
of the variance in scores on overall teaching ability.
Other factors, such as teacher personality traits and classroom
behaviors, are much more important (Erdle et al., 1985).
More research is needed before definite conclusions can be
drawn. First, replication in other settings is needed, because
the current sample comes from a small, selective, and private
undergraduate institution with only a 15-year history of co-
education. These students may be fairly traditional, as evi-
denced by their traditional division of majors by sex. As such,
they may show more sex bias and be more sensitive to sex-
stereotypic qualities of
a
professor than would less traditional
students. Second, the present study surveyed only a small
number of tenured professors and excluded professors in
engineering, a "male" discipline. Thus, the sample of profes-
sors was relatively young, relatively small, and not represent-
ative of the entire faculty, although it was representative of
the female faculty. Investigation of a more diverse faculty and
student body may lead to different findings. Third, a teacher's

314
SUSAN
A.
BASOW
AND
NANCY
T,
SILBERG
personality characteristics are important in student evalua-
tions,
and future research should incorporate teachers'
self-
ratings to determine the relation between students' percep-
tions and those of their professors. Fourth, because student
major appears to affect evaluations of male and female pro-
fessors, it would be important to ascertain if students in
different majors vary in their attitudes to ward
women.
Finally,
direct observation is needed to determine if female and male
faculty actually behave differently in the classroom. Once all
this information is in, investigators may be able to answer
with greater confidence the question of whether students
evaluate female and male professors differently.
References
Abrami,
P. C ,
Leventhal,
L., &
Perry,
R. P.
(1982), Educational
seduction.
Review
of
Educational
Research,
52,446-464.
Basow,
S. A., &
Distenfeki,
M. S,
(1985). Teacher expressiveness:
More important
for
males than females? Journal
of
Educational
Psychology,
77,
45-52.
Basow,
S. A.,
& Howe,
K.
(1987). Evaluations
of
college professors:
Effects
of
professors* sex-type
and
sex,
and
students' sex. Psycho-
logical
Reports,
60,671-678.
Bern,
S. L.
(1981). Bern Sex-Role
Inventory:
Professional
manual.
Palo Alto, CA: Consulting Psychologists Press.
Bennett,
S. K.
(1982). Student perceptions
of
and expectations
for
male
and
female instructors: Evidence relating
to the
question
of
gender bias
in
teaching evaluations. Journal
of
Educational
Psy-
chology, 74, 170-179.
Elmore,
P.
B.,
&
LaPoiitte,
K.
A. (1974). Effects
of
teacher sex, and
student
sex on the
evaluation
of
college instructors. Journal
of
Educational
Psychology,
66, 386-389.
Elmore,
P. B., &
LaPointe,
K. A.
(1975). Effects
of
teacher
sex,
student
sex, and
teacher warmth
on the
evaluation
of
college
instructors.
Journal
of Educational
Psychology,
67, 368-374.
Erdle,
S.,
Murray,
H. G., &
Rushtoo,
J. P.
(1985). Personality,
classroom behavior,
and
student ratings
of
college teaching effec-
tiveness:
A
path analysis. Journal of Educational
Psychology,
77,
394-407.
E l a u g h .
C ,
&.
Rilcy, S. (1983). Evaluating competence of women and
men: Effects
of
marital
and
parental status and occupational sex-
typing. Sex
Roles,
9,943-952.
Harris,
M. B.
(1975),
Sex
role stereotypes
and
teacher evaluations.
Journal of
Educational
Psychology,
67, 751-756.
Harris, M.
B.
(1976). The effects of sex, sex-stereotyped descriptions,
and institution on evaluations
of
teachers.
Sex
Roles,
2,
15-21.
Hildebrand,
M.
(
&
Wilson,
R,
(1970). Effective university teaching
and its
evaluation.
Washington, DC: U.S. Department
of
Health,
Education, and Welfare, Office of Education.
Holahan,
C K.,
& Stephen,
C.
W. (1981). When beauty isn't talent:
The influence
of
physical attractiveness, attitudes toward women,
and competence
on
impression formation. Sex
Roles,
7,
867-876.
Hudak, M. A., & Anderson, D. E. (1984), Teaching style and student
ratings.
Teaching
of
Psychology*
II,
177-178.
Kaschak, E. (197$). Sex bias in student evaluations of college profes-
sors.
Psychology
of Women
Quarterly,
3
t
235-243.
Kaschak, E. (1981). Another look
at
sex bias in students
1
evaluations
of professors:
Do
winners get the recognition
mat
they have been
given?
Psychology
of
Women
Quarterly,
5, 767-772,
Leventhal,
L.,
Perry,
R,, &
Abrami,
P,
(1977). Effects
of
lecturer
quality and student perception
of
lecturer's experience
on
teacher
ratings and student achievement. Journal of Educational
Psychol-
ogy, 69, 360-374.
Lombardo,
J.,
& Tocci, M. (1979). Attribution
of
positive and nega-
tive characteristics
of
instructors as
a
function of attractiveness and
sex
of
instructor and sex
of
subject.
Perceptual &
Motor Skills,
48,
491-494.
Marsh,
H. W.
(1984). Students' evaluations
of
university teaching:
Dimensionality, reliability,
and
validity, potential biases,
and
util-
\\y. Journal of
Educational
Psychology,
76, 707-754.
Marsh,
H. W., &
Ware,
J. E., Jr.
(1982). Effects
of
expressiveness,
content
coverage,
and incentive on multidimensional student rating
scales: New interpretations
of
the Dr. Fox effect. Journal of Edu-
cational
Psychology,
74, 126-134.
Paiudi, M. A., & Bauer, W, D. (1983). Goldberg revisited: What's
in
an author's name.
Sex
Roles,
9,387-390.
Touhey,
J. C.
(1974). Effects
of
additional women professionals
on
ratings of occupational prestige and desirability. Journal of Person-
ality and Social
Psychology,
28, 360-367.
Ware,
i. E.,
& Williams,
R. G.
(1975). The Dr. Fox effect
A
study
oflecturer effectiveness and ratings of instruction. Journal of Med-
ical
Education,
50
t
149-150-
Received July 26, 1986
Revision received February 5, 1987
Accepted March 1, 1987 •