Exploring Moderators of Gender Differences:

Contextual Differences in Door-Holding Behavior1

J ANI CE D. YODER,~ MARY HOGUE, ROBERT NEWMAN,
LI NDA METZ, AND TONYA LAVIGNE
University ofAkron
The simple behavior of holding a door for another person can be interpreted as gender-
neutral, helpful, or benevolently sexist. Each interpretation leads to competing hypotheses
predicting no gender differences in door holding, consistent door holding by men as chiv-
alrous helpers, and door holding by men in the gender-salient context of dating but not in
everyday interactions, respectively. Observations of 769 college-aged, female-male dyads
found a strong pattern of male door holding in dating but not in everyday contexts. This
pattern highlights social role theory's emphasis on understanding contextual moderators of
apparent gender differences, changes in door-holding behavior in everyday contexts across
the past 20 years, and the benevolent sexism subtly conveyed by male dominance in door
holding in dating contexts.
Eagly and Wood (1 99 1) argued that an important step toward understanding
apparent gender differences in social behavior is to explore moderator variables
that alternately exaggerate and attenuate these differences. Their social role the-
ory proposes that the influence of gender-role expectations on gender differences
in social behavior depends on the degree to which the context in which the behav-
ior occurs makes salient or minimizes gender stereotypes. The importance of
understanding social contextual contributions also is highlighted in Deaux and
Major's (1987) exposition of gender enactments in social exchanges. The purpose
of the present paper is to examine how gender salience affects patterns of female-
male behavior. Specifically, we will explore door-holding behavior in the gender-
salient context of dating versus everyday life.
The social meaning underlying door-holding etiquette seems to be in flux.
One argument assumes a gender-neutral stance, proposing that door holding is a
simple form of courtesy. Following this line of reasoning, who holds a door for
whom is dictated by rules independent of gender such that women should hold
doors for men as readily as men extend this courtesy to women, regardless of
context.
'The authors thank Eva Fong and David Worrels for their help with data collection.
korrespondence concerning this article should be addressed to J anice D. Yoder, Department of
Psychology, University of Akron, Akron, OH 44325-4301.
Journal of Applied Social Psychology, 2002, 32, 8, pp. 1682-1 686.
Copyright 0 2002 by V. H. Winston & Son, Inc. All rights reserved.
EXPLORING MODERATORS OF GENDER DIFFERENCE
1683
Others associate the public ritual of door holding with gender such that men
hold doors for women and the reverse is regarded as deviant. Indeed, there is
evidence to support this contention (Ventimiglia, 1982). Although participants in
rituals of gender etiquette typically were found to be unaware of the motivational
meaning underlying the proscribed behavior (Parker, 1988), at least two gendered
scenarios have been explored.
The first relates door holding to helping behavior. In this scenario, men hold
doors for women because helping, at least as an act of chivalry, is linked to men
and masculinity (Goldman, Florez, & Fuller, 1981). There is strong meta-analytic
evidence to support the association between chivalry and being male (Eagly &
Crowley, 1986). However, the connection between this form of helping and door
holding is less solid. If, indeed, door holding is an expression of chivalry, then we
would expect the pattern of men as holders and women as recipients to remain
constant across different contexts.
A second gendered analysis of the male-as-holder pattern regards door hold-
ing as fundamentally sexist (Walum, 1974). This reasoning is developed more
fully by Glick and Fiskes (1996) work that distinguishes between openly nega-
tive, hostile sexism and benevolent sexism. Although seemingly paradoxical,
benevolent sexism refers to attitudes toward women that view women stereotypi-
cally and confine them to restricted roles but that, on the surface, appear favor-
able in tone. Arguably, door holding is a behavioral expression of these attitudes
because superficially it appears kindly, but at a deeper level might convey mes-
sages of male dominance and female passivity.
Furthermore, Glick and Fiske (1 996) empirically supported a model whereby
benevolent sexism was shown to be rooted in three components: protective pater-
nalism, complementary gender differentiation, and heterosexuality. Door holding
by men for women cuts across each of these three issues by invoking male domi-
nance and protection of women, complementary roles for men as active and
women as passive, and sexual attraction, respectively. This analysis suggests that
gendered patterns of men holding doors for women will be most pronounced in
contexts emphasizing paternalism, complementarity, and heterosexuality. The
corollary to this is that gendered patterns should be less evident when the context
in which door holding occurs minimizes these three factors.
A context that emphasizes gender stereotypes as well as protection, comple-
mentarity, and heterosexual attraction is heterosexual dating. Indeed, undergradu-
ate mens hypothetical and womens actual dating scripts included references to
courtly behaviors, such as door holding (Rose & Frieze, 1993). On the other end
of the continuum, a context that de-emphasizes these factors, but that draws on
the same population of participants, is door holding in everyday public places
(e.g., campus buildings, churches, fast-food restaurants).
The three analyses of door holding we explored offer competing hypotheses.
A gender-neutral stance predicts equal rates of door holding by women and men
1684 YODER ET AL.
across both contexts. The helping approach postulates consistent helping by men
for women across both contexts. Finally, speculation about sexism expects men to
more frequently hold doors for female dates than for women in everyday con-
texts. Only the final hypothesis predicts gender differences in door-holding pat-
terns across the two contexts.
Method
Three female and two male undergraduate students, working on a research
project for a Research Methods class, individually and unobtrusively observed
769 female-male pairs in which one person opened a door for the other at 16
public locations in the fall semester of 1999. The locations and times of data col-
lection were selected to maximize the likelihood of finding either dating or non-
dating (everyday) mixed-gender, college-aged dyads.
Seven everyday locations included a shopping mall, two large public universi-
ties in northeast Ohio, a church adjacent to one of the campuses, and three fast-
food restaurants. The remaining nine locations were chosen as common dating
environments attractive to college students: six sit-down, inexpensive restaurants
and three skating rinks. Observations at the everyday locations were conducted
on weekdays and involved 404 pairs; data collection at the likely dating sites was
done on Friday and Saturday nights and included 365 couples.
Female-male pairs were included only when they appeared to be traditional
college age, approached the targeted door as a dyad, and one held the door open
for the other to walk through first. Couples in the dating locations and times were
included only when they exhibited overt signs of courtship, such as handholding
or some other deliberate sign of affection. Likewise, couples exhibiting courtship
behaviors in the everyday context were excluded. In sum, dating and everyday
designations of pairs were distinguished by differences in location, time, and
dyad-specific behaviors. Although no systematic data were collected about the
race/ethnicity of participants, most of the participants were White.
Results
To test whether the context in which door holding occurs matters, an overall
chi-square test comparing the pattern of door holding between the dyad designa-
tions was conducted and was significant, x2(1, N = 769) = 37.70, p < .001. This
overall chi-square test was subsequently probed with follow-up chi squares com-
paring women and men door holders within each dyad designation. In everyday
dyads, 223 women (55.2%) held the door for men significantly more than vice
versa, x*( 1, N = 404) = 4. 37, ~ = .037. The pattern reversed in the dating context,
with more men (66.8%, n = 244) holding the door for women than vice versa,
X2(1,N=365)=41.45,p< .001.
EXPLORING MODERATORS OF GENDER DIFFERENCE 1685
Discussion
Since both the gender-neutral and helping predictions about door-holding
practices predicted no differences across contexts, neither was supported by the
present data. Rather, the observed pattern is consistent with an explanation draw-
ing on the concept of benevolent sexism. In the everyday context, door holding
tipped statistically significantly in the direction of women holding doors for men
(55.2% women holding doors for men, 44.8% men holding doors for women).
The skew in the pattern was much more pronounced in a dating context and
reversed such that dating men held the door for women in about two of every
three couples (66.8%). An even stronger test of the present speculation would be
to observe the same women and men across both contexts.
The pattern found here for everyday door holding stands in sharp contrast to
door holding reciprocity in a non-dating setting (an airport) about 20 years ago.
Goldman et al. (198 1) found that women returned door holding for a man who
previously had held a door for them only 33% of the time, compared to 48% rec-
iprocity for men. Consistent with the earlier findings but not the present study,
Ventimiglia ( 1982), using the everyday setting of libraries, found significantly
more confusion reported by participants who encountered a woman holding a
door for a man than vice versa. It appears that the meaning of the ritual of door
holding in day-to-day contexts has changed across the past 20 years, possibly
from chivalry (a masculine domain) to politeness (Maccoby, 1990) or caregiving
(Gerstel& Gallagher, 1994), both of which are feminized.
However, the present data suggest that door holding retains its gendered
meanings in the more gender-salient context of heterosexual dating. In addition
to and congruent with the contextual analysis offered here, Rickard (1989)
related door-holding behavior in dating relationships to individual differences in
feminist identity. Classifying participants feminist identity development using
Downing and Roushs (1985) model, Rickard found that men with low levels of
feminist identity development were more likely to report that they held doors
than were pro-feminist men, and that women were more likely to describe being
the recipients of this behavior when they displayed low levels of feminist identity
development. These individual-difference data combine with the contextual data
presented here to support speculation that door-holding behavior is indeed linked
to sexism, however benevolent.
At a broader levef, the present pattern of data illustrates the importance of
exploring moderator variables as predicted by social role theory (Eagly & Wood,
1991). In the dating context, where gender-role expectations are salient and pro-
scribed, the public ritual of door holding takes on a much more consistent pattern
than in everyday contexts. A clear gender difference in door-holding behavior in
a dating context becomes more muddled in the less gendered context of everyday
female-male relations.
1686 YODER ET AL.
References
Deaux, K. , & Major, B. (1987). Putting gender into context: An interactive model
of gender-related behavior. Psychological Bulletin, 94,369-389.
Downing, N. E., & Roush, K. L. ( 1985). From passive acceptance to active com-
mitment: A model of feminist identity development for women. The Counsel-
ing Psychologist, 13, 695-709.
Eagly, A. H., & Crowley, M. (1986). Gender and helping behavior: A meta-
analytic review of the social psychological literature. Psychological Bulletin,
Eagly, A. H., & Wood, W. (1991). Explaining sex differences in social behavior:
A meta-analytic perspective. Personality and Social Psychology Bulletin, 17,
Gerstel, N., & Gallagher, S. (1994). Caring for kith and kin: Gender, employ-
ment, and the privatization of care. Social Problems, 41, 519-539.
Glick, P., & Fiske, S. T. (1996). The Ambivalent Sexism Inventory: Differentiat-
ing hostile and benevolent sexism. Journal of Personality and Social Psychol-
Goldman, M., Florez, C., & Fuller, G. L. (1981). Factors affecting courteous
behavior. Journal of Social Psychology, 115, 169- 174.
Maccoby, E. E. (1990). Gender and relationships: A developmental account.
American Psychologist, 45, 5 13-520.
Parker, S. (1988). Rituals of gender: A study of etiquette, public symbols, and
cognition. American Anthropologist, 90, 372-384.
Rickard, K. M. (1 989). The relationship of self-monitored dating behavior to level
of feminist identity on the Feminist Identity Scale. Sex Roles, 20,2 13-226.
Rose, S., & Frieze, 1. H. (1993). Young singles contemporary dating scripts. Sex
Ventimiglia, J . C. (1982). Sex roles and chivalry: Some conditions of gratitude to
Walum, L. R. (1974). The changing door ceremony: Notes on the operation of
100, 283-308.
306-3 15.
ogy, 70,491-512.
Roles, 28,499-509.
altruism. Sex Roles, 8, 1107-1 122.
sex roles in everyday life. Urban Lifr and Culture, 2, 506-5 15.