Content uploaded by Jonas R. Kunst
Author content
All content in this area was uploaded by Jonas R. Kunst on Dec 04, 2018
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=hmep20
Media Psychology
ISSN: 1521-3269 (Print) 1532-785X (Online) Journal homepage: http://www.tandfonline.com/loi/hmep20
Sexism, rape myths and feminist identification
explain gender differences in attitudes toward the
#metoo social media campaign in two countries
Jonas R. Kunst, April Bailey, Claire Prendergast & Aleksander Gundersen
To cite this article: Jonas R. Kunst, April Bailey, Claire Prendergast & Aleksander Gundersen
(2018): Sexism, rape myths and feminist identification explain gender differences in attitudes
toward the #metoo social media campaign in two countries, Media Psychology, DOI:
10.1080/15213269.2018.1532300
To link to this article: https://doi.org/10.1080/15213269.2018.1532300
Published online: 25 Oct 2018.
Submit your article to this journal
Article views: 209
View Crossmark data
Sexism, rape myths and feminist identification explain
gender differences in attitudes toward the #metoo social
media campaign in two countries
Jonas R. Kunst
a,b
, April Bailey *
b
, Claire Prendergast*
a
,
and Aleksander Gundersen
a
a
Department of Psychology, University of Oslo, Oslo, Norway;
b
Department of Psychology, Yale
University, New Haven, Connecticut, USA
ABSTRACT
On October 15, 2017, actress Alyssa Milano popularized the
#metoo campaign, which sought to expose the prevalence of
sexual harassment and assault in public domains by encouraging
victims to share their experiences on social media using the
hashtag metoo. The online campaign rapidly grew to a global
phenomenon, which was generally well supported. However,
some criticized the campaign online as a battle of the sexes,
which pits men against women. Our cross-cultural research inves-
tigated whether gender differences in attitudes and feelings
toward #metoo are due to underlying differences in ideologies
and experiences that only partly overlap with gender. We sur-
veyed respondents in the United States, where the campaign
began, and in Norway, a highly gender-egalitarian country. In
both countries, men expressed less positivity toward #metoo
than women and perceived it as substantially more harmful and
less beneficial. These gender differences were largely accounted
for by men being higher than women in hostile sexism, higher in
rape myth acceptance, and lower in feminist identification. The
results, hence, suggest that gender differences in attitudes to
social media campaigns such as #metoo might be best character-
ized as dimensional ideological differences rather than funda-
mental group differences.
In 2006, civil rights activist Tarana Burke introduced the phrase “me
too”to raise awareness about the widespread nature of sexual harass-
ment and assault (Ohlheiser, 2017). Over a decade later on October 15,
2017, American actress Alyssa Milano encouraged women who had
experienced sexual harassment or assault in public domains to share
their stories on social media by posting the term as a hashtag (i.e.,
#metoo). Milano’s suggestion sparked a global movement, and within a
day 85 million people had shared the hashtag in an effort to expose the
prevalence of sexual harassment and assault (CBS, 2017).
CONTACT Jonas R. Kunst j.r.kunst@psykologi.uio.no Postboks 1094, Blindern, 0317 Oslo, Norway.
*These authors contributed equally to this work.
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/hmep.
MEDIA PSYCHOLOGY
https://doi.org/10.1080/15213269.2018.1532300
© 2018 Taylor & Francis Group, LLC
Although the social media campaign was generally well received, some
criticized it. For instance, filmmaker Michael Haneke called the campaign a
“witch hunt”(T. Wright, 2018), and Texas attorney general aide Andrew
Leonie asked, “Aren’t you also tired of all the pathetic ‘me too’victim
claims?”(Astor, 2017). These negative statements by men might reflect
gender differences in reactions to the campaign. In general, men tend to be
more tolerant of sexual harassment than women (Russell & Trigg, 2004).
Indeed, a recent YouGov poll showed that even in Norway, one of the most
gender egalitarian countries in the world (UNDP, 2016; World Economic
Forum, 2016), more men than women believed that the campaign had gone
too far and would have no positive effect (Keldsen, 2018). Possibly because of
these more negative reactions by men, some characterized #metoo as a
“battle of the sexes,”which pits women against men (Fallon, 2018). Note
that in this article, we use the term gender to refer to individuals’self-
identified group membership, which often, but not always, overlaps with
their biological sex.
Men might oppose the campaign more than women due to group-level
processes. For instance, men and women might have an evolved tendency to
perceive social interactions between the sexes in inherently different ways
(Geary, 1998; Tannen, 1990). Thus, the same encounters might be judged by
women as sexual harassment, but by men as harmless flirtation. Even when
men and women judge the same encounter similarly, men might be incenti-
vized to oppose the #metoo campaign due to in-group favoritism and a
motivation to support other men (Tajfel, 1982). These explanations for
gender differences in reactions to #metoo thus emphasize intergroup varia-
bility and intragroup consistency.
In contrast to group-based perspectives that treat men and women as
fundamentally distinct and motivated by in-group favoritism, there stands
a more dimensional perspective on gender differences (Reis & Carothers,
2014; Wood & Eagly, 2012). In this view, men and women’s attitudes toward
#metoo might be explained largely by individual differences in ideological or
experiential dimensions that only partially align with gender. Suggestive of
this possibility, many individual men expressed support for #metoo
(Vagianos, 2017), and some individual women harshly criticized the cam-
paign (BBC, 2018).
In our research, we examined gender differences in attitudes and feelings
toward #metoo, arguably one of the most far-reaching and consequential
social media campaigns to date. Insofar that men and women differentially
support the #metoo campaign, we tested whether these gender differences
could be explained by ideologies (hostile sexism, benevolent sexism, feminist
identity, rape myth acceptance, and belief in a just world) and experiences
(victim or perpetrator experiences, pornography consumption) that only
partially align with gender. To establish the cross-cultural generalizability
2J. R. KUNST ET AL.
of these findings, we compared the United States and Norway, which are two
countries differing in gender equality.
Men and women differ in ideologies related to sexual misconduct
Men and women differ substantially in their attitudes toward the social issues
that are at the core of the #metoo campaign. Most centrally, men, compared
to women, demonstrate greater tolerance of sexual harassment, are more
positive toward sexual assault, and are more likely to blame rape victims for
their assault (Anderson, Cooper, & Okamura, 1997; Payne, Lonsway, &
Fitzgerald, 1999; Rotundo, Nguyen, & Sackett, 2001; Russell & Trigg, 2004;
Sakallı-Uğurlu, Yalçın, & Glick, 2007; van der Bruggen & Grubb, 2014).
Importantly, these gender differences are often due to differences in under-
lying ideologies and experiences.
First, gender differences in sexist attitudes toward women might explain
men and women’s differential responses to the #metoo campaign. Classic
and enduring work by Glick and Fiske (Glick & Fiske, 1996;Lee,Fiske,&
Glick, 2010) identifies one form of sexism, hostile sexism, as involving
blatantly misogynic stereotypes and attitudes toward women that construe
women as sexually manipulative and inferior. Hostile sexism thus coheres
with initial definitions of prejudice as antipathy (e.g., Allport, 1954), and
can serve to justify men’s exploitation of women as sexual objects. Indeed,
men are typically higher in hostile sexism than women (Glick et al., 2000),
and hostile sexism is associated with a greater tolerance of sexual harass-
ment, increased moral disengagement from sexual harassment, and even a
higher proclivity to commit sexual assault (Abrams, Viki, Masser, &
Bohner, 2003; Masser, Viki, & Power, 2006; Page, Pina, & Giner-Sorolla,
2016;Russell&Trigg,2004).
Glick and Fiske (1996) further introduced another form of sexism, termed
benevolent sexism, which involves an apparently positive, yet patronizing,
stance toward women (i.e., that women need to be cherished and protected).
Likely due to this positivity, women and men seem to endorse benevolent
sexism to a similar degree (Glick et al., 2000), and in more recent work
benevolent sexism predicted more nuanced attitudes toward sexual miscon-
duct. Specifically, benevolent sexism predicted both opposition to, and sup-
port of, gender equality (Hideg & Ferris, 2016; Sibley & Perry, 2010) and
attitudes surrounding sexual harassment and assault (Abrams et al., 2003;
Masser, Lee, & McKimmie, 2010; Russell & Trigg, 2004). One reason for the
nuanced influence of benevolent sexism is that it is not directed at women,
per se. Often, it elicits the paternalistic protection of traditional women (e.g.,
housewives), but hostility toward nontraditional women (e.g., scientists,
promiscuous women; Bareket, Kahalon, Shnabel, & Glick, 2018; Fowers &
Fowers, 2010; Sakallı-Uğurlu, 2010).
MEDIA PSYCHOLOGY 3
Furthermore, in addition to general sexism, men and women differ in their
support of specific ideologies related to sexual misconduct. Men tend to
accept rape myths more than women (Suarez & Gadalla, 2010). Rape
myths involve various beliefs, including blaming rape victims for their
assault, absolving the perpetrator of blame, and downplaying rape’s serious
adverse consequences for the victim (Payne et al., 1999). Importantly, the
more individuals endorse rape myths, the more they accept rape and inter-
personal violence and show a greater proclivity for sexual aggression and
coercion (Suarez & Gadalla, 2010). Because men endorse rape myths more
than women do, men might be more likely than women to relativize and
downplay the experiences of individuals posting the hashtag #metoo, and to
place the blame on victims instead of perpetrators.
Moreover, men being lower than women in feminist identification might
explain gender differences in attitudes toward #metoo. Politicized social
identifications, such as feminist identification, are particularly strong predic-
tors of collective action (van Zomeren, Postmes, & Spears, 2008). Previous
research has shown that feminist identity, indeed, is associated with greater
engagement in gender-related collective action (Nelson et al., 2008; Yoder,
Tobias, & Snell, 2011) and predicts willingness to confront sexism, sexual
harassment, and other misogynistic issues (Ayres, Friedman, & Leaper, 2009;
Myaskovsky & Wittig, 1997). Individuals who identify as feminists likely also
see the #metoo campaign as beneficial because its goal is to achieve social
change by highlighting that sexual harassment and assault are still common
phenomena that need to be addressed. However, because men are substan-
tially less likely to identify as feminists than women (McCabe, 2005), it
follows that men might be less invested in the #metoo campaign and perceive
it as less beneficial.
Last, we propose that beliefs in a just world (BJW) might explain gender
differences in reactions to #metoo. People high in BJW generally are con-
vinced that the world is a just place where bad things happen to bad people
and good things happen to good people (Dalbert, 1999; Lerner, 1980). Thus,
people high in BJW might defensively slander rape victims and twist their
character so that the victims’fates seem deserved. Indeed, BJW is related to
less positivity toward rape victims, support of more lenient sentences for
rapists and more victim blaming, thus placing less blame on the perpetrator
(Kleinke & Meyer, 1990; Sakallı-Uğurlu et al., 2007; Strömwall, Alfredsson, &
Landström, 2013). Accordingly, because people high in BJW likely believe
that victims who draw attention to their experiences through the #metoo
campaign are in some way blameworthy, they might see the campaign as
being less beneficial and more harmful. However, evidence has been mixed in
terms of gender differences in BJW (e.g., Sakallı-Uğurlu et al., 2007; Whatley,
1993), suggesting that the construct might play less of a role in explaining
gender differences in attitudes toward #metoo.
4J. R. KUNST ET AL.
Men and women differ in experiences related to sexual misconduct
In addition to differing in ideologies relevant to #metoo, men and women
also tend to differ in experiences related to sexual harassment and assault.
Women are 10 times more likely than men to become victims of sexual
violence (NSVRC, 2015). Although many victims of sexual abuse do not
acknowledge that what happened to them was rape, those who do tend to
show less tolerance and more negative attitudes toward rape and tend to
engage in less victim blaming (Anderson et al., 1997; Miller, Amacker, &
King, 2011). Thus, women might be more positive toward #metoo because
they are more often victims of sexual misconduct.
Although women are far more likely to become victims of sexual violence,
perpetrators are almost exclusively men (Sedgwick, 2006). Experience as a
perpetrator has been linked to more positive attitudes toward rape (Anderson
et al., 1997). Thus, men’s greater perpetrator experiences might explain
gender differences in attitudes toward #metoo.
Finally, experiences with pornographic media might influence attitudes
toward sexual misconduct. Pornography has a broad range of themes, and
some content is even designed to be empowering to women (Fritz & Paul,
2017). Yet, exposure to certain pornography might encourage detrimental
attitudes toward women and foster a belief that women ultimately endorse
sexual violence. The consumption of mainstream pornography, and porno-
graphy involving rape or other scenes of violence specifically, is related to
increased sexual aggression, a greater proclivity to make unwanted sexual
advances, greater bystander passivity, and a higher likelihood to commit
sexual assault (Allen, Emmers, Gebhardt, & Giery, 1995; Foubert, Brosi, &
Bannon, 2011; P. J. Wright, Tokunaga, & Kraus, 2016). Men’s tendency to
consume more pornographic material than women (Hald, 2006) might thus
account for gender differences in attitudes toward #metoo.
The contexts of comparison
Due to the proliferation of the #metoo hashtag on social media worldwide, an
important aspect of the #metoo campaign is its global dimension. Different
cultural contexts have differing levels of gender equality, which could shape
culturally-contingent reactions to #metoo. Thus, it is important to consider
such cultural variation to establish the generalizability of potential findings. The
#metoo campaign originated in the United States, a relatively gender egalitarian
nation. People in the United States score lower in hostile and benevolent
sexism, compared to those from countries like South Korea and Cuba (Glick
et al., 2000). Yet, the United States is less gender egalitarian than countries such
as Norway. People in Norway score even lower in hostile and benevolent sexism
than in the United States. Although the United States and Norway are quite
MEDIA PSYCHOLOGY 5
similar on cultural dimensions such as power distance, individualism, uncer-
tainty avoidance, long-term orientation, and indulgence, masculinity is much
higher in the United States, which reflects the presence of more traditional
gender roles (Hofstede, Hofstede, & Minkov, 2010). Indeed, the World
Economic Forum (2016)recentlyrankedNorwayasthethirdmostgender-
egalitarian country in the World; the United States was placed at 45 out of 144.
These cross-cultural differences in gender egalitarianism between the United
States and Norway likely have socio-historical roots. For instance, although
women in Norway were granted the right to vote and run for election in 1907, it
was another 13 years before women in the United States obtained these rights in
1920. Most importantly, differences between the countries are also evident in
the current socio-political contexts. Norway has had female prime ministers for
four terms, but a woman has yet to serve as head of state in the United States.
Nearly half (40%) of the lawmaking body in Norway is composed of women,
compared to 19% of the US Congress (Inter-Parliamentary Union, 2017).
Women’s political participation in Norway has impacted policies relevant to
issues of gender inequality, specifically improving childcare (Bratton & Ray,
2002). Country differences in gender inequality are also reflected in terms of
sexual violence. In 2014 (i.e., the latest data available for both countries),
Norway had a prevalence of 21.9 and the United States of 37.0 rape incidents
per 100,000 population (UNODC, 2017).
Importantly, the #metoo campaign made a similar impact in both coun-
tries, exposing sexual harassment and assault in various domains. Several
public figures in both countries suffered professional repercussions due to
allegations of sexual misconduct. For instance, in the United States, comedian
Louis C.K.’sfilm,I Love You, Daddy was pulled from distribution; television
anchor Matt Lauer was fired by the National Broadcasting Company; and
Democrat senator Al Franken resigned from political office. In Norway,
Trond Giske, deputy leader of the Norwegian Labor Party, Kristian
Tonning Riise, leader of the Norwegian Young Conservatives, Ulf Leirstein,
deputy leader of the Progress Party’s parliamentary group, and Davy Wathne,
sports journalist and anchor at the largest Norwegian TV-channel NRK,
resigned from their positions in response to allegations of sexual harassment.
Hence, the United States and Norway provide an interesting comparative
context as the #metoo campaign had a strong impact in both cultures yet they
differ in their degree of gender egalitarianism.
This research
Although social media campaigns have a unique potential to achieve social
change, their inherent value is often perceived differentially by social groups.
Yet, very little knowledge exists about the factors explaining why some
groups show positivity toward specific social media campaigns and perceive
6J. R. KUNST ET AL.
them as beneficial but others perceive them as harmful. Here, we believe that
psychological research has the potential to provide critical insights. Focusing
on the #metoo campaign, we investigated the factors underlying men and
women’s differential attitudes and feelings toward #metoo. A group-based
perspective would frame #metoo as a battle of the sexes and argue that men
and women are distinct types who react differently to #metoo due to funda-
mental group processes (e.g., Geary, 1998). By contrast, dimensional per-
spectives would emphasize that mean gender differences reflect women and
men’s relative positions on underlying dimensions that only partially align
with gender (Reis & Carothers, 2014; Wood & Eagly, 2012). We propose that
ideological differences in sexism, rape myths, feminism, and BJW, along with
differences in experiences with pornography, sexual harassment and assault
might explain gender differences in attitudes toward #metoo. Thus, we
hypothesize that gender differences in attitudes toward #metoo will be
mediated by ideological and experiential differences.
Materials and methods
Supplementary online materials (SOM) for this article can be found at
https://osf.io/5p7fh/?view_only=ca95b9ecfa1f47d39bc33ba93341ea2c.
1
Participants
A power simulation for structural equation modeling using the semTools
package (Jorgensen, Pornprasertmanit, Schoemann, & Rosseel, 2018)inR
suggested that 203 participants would provide a 90% chance to fit a moder-
ately complex model (df = 40) with a hypothesized root mean square of
approximation (RMSEA) = .05 (i.e., a satisfactory fit), and an alternative
RMSEA = .10 (i.e., an unsatisfactory fit). This number of participants also
exceeded the number of participants necessary (N= 172) to have 90% chance
to observe medium (f= .25) gender differences at a .05 significance criterion
using analysis of variance.
Norwegian sample
In total, 206 Norwegian participants (M
age
= 30.78, SD
age
= 12.15; gender:
52.4% women, 46.6% men, 1% other) were recruited through online social
networks for a study on “online social media campaigns”in November 2017.
As financial incentive, all participants could take part in the drawing of a gift
voucher of 200 NOK (ca. $25). Most participants (85.9%) reported that they
knew about #metoo before taking the study. On average, participants
reported having seen 17.90 #metoo hashtags (SD = 37.91) on social media,
after we removed one value (999) that emerged as a multivariate outlier (see
SOM). Of all participants, 85.4% were native Norwegians and 14.6% had at
MEDIA PSYCHOLOGY 7
least one parent with an ethnic minority-group background; 86.9% reported
living in cities and 13.1% in more rural areas. On average, participants had
completed 3.96 years (SD = 2.76) of higher education.
US sample
Via the online panel Amazon Mechanical Turk, 227 participants (M
age
=35.31,
SD
age
= 10.77; gender: 48.2% women, 50.8% men, 1% other) were recruited in
November 2017 for the study, which was described in the same way as it was
to the Norwegian sample. Participants received $1.50 for their participation.
Similar to the Norwegian sample, most participants (74%) reported that they
knew about #metoo before participating in the study. Participants reported
having seen an average of 17.45 #metoo hashtags (SD = 20.06) on social media,
after we removed three values (500, 500, and 400) that emerged as multivariate
outliers (see SOM). Of all participants, 83.7% were White/Caucasian, 5.7%
African American, 5.3% Hispanic, 4.4% Asian, and less than one percentage
indicated to have a Native American or another ethnic background; 79.7%
reported living in cities and 20.3% in more rural areas. On average, partici-
pants had completed 7.69 years (SD = 6.08) of higher education.
Group comparisons showed that the US sample did not differ from the
Norwegian sample in terms of gender, χ
2
(1) = .62, p= .431, place of living, χ
2
(1) = 3.45, p= .063, and number of #metoo hashtags seen, t(246.45) = .12,
p= .903, but the US sample was slightly older, t(404.47) = 4.07, p< .001,
d= .39, more educated, t(322.96) = 8.26, p< .001, d= .79, and had less prior
knowledge about the campaign, χ
2
(1) = −9.35, p= .002.
Procedure
Participants completed an online questionnaire comprising the measures
described in the following sections. At the end, they were debriefed and
received contact details. All measures were forward-back translated from
English into Norwegian by bilingual teams. The research was approved by
the Institutional Review Board of the first author.
Measures
Unless otherwise stated, responses were scored on 7–point Likert scales
ranging from 1 (totally disagree)to7(totally agree). All predictor variables
were presented in random order. Although often ignored in cross-cultural
research, measurement invariance has to be established before groups can be
validly compared (Chen, 2008). In the present study, we used multigroup
structural equation modeling to identify versions of the measures that
showed acceptable measurement invariance. Specifically, for each measure,
we evaluated factor solutions both in terms of configural and metric
8J. R. KUNST ET AL.
invariance. Configural invariance is achieved when the underlying factor
structure is the same in both countries. Metric invariance is achieved when
the factor loadings in addition are identical in both groups. Importantly,
metric invariance must be established to make valid mean score compari-
sons. When describing the measures, we provide brief information about the
measurement invariant versions that were used in analyses; detailed informa-
tion about the underlying tests can be found in the SOM. Two measures that
showed unacceptable measurement invariance or reliability (i.e., a measure of
feminist ideology and a measure of intentions to behaviorally support the
#metoo campaign) were excluded from analyses (see SOM for further
details).
Predictor variables
Ambivalent sexism inventory. Participants completed the 22-items Ambivalent
Sexism Inventory (Glick & Fiske, 1996). Eleven items measured hostile sexism
(e.g., “Women seek to gain power by getting control over men”) and a further 11
measured benevolent sexism (e.g., “A good woman should be set on a pedestal
by her man”). Translations of the scale into Norwegian were obtained from
Gundersen and Kunst (2018). One item of the benevolent sexism dimension and
four items of the hostile sexism dimension had to be omitted from the factor
solution to achieve metric measurement invariance. The resulting 10-item
benevolent sexism scale (Norway: α= .80; United States: α= .91) and the 7-
item hostile sexism scale (Norway: α= .91; United States: α=.94)showed
satisfactory reliability in both countries.
Feminist identity. A feminist identity scale was adopted from Szymanski
(2004). Participants indicated their agreement with four items such as “I
consider myself a feminist.”The scale showed satisfactory metric measure-
ment invariance and reliability in both countries. (Norway: α= .91; United
States: α= .93).
Rape myths. The 20-item short-form version of the Illinois Rape Myth
Acceptance Scale (Payne et al., 1999) was used to measure the extent to
which participants endorsed various rape myths. Participants indicated their
agreement with statements such as, “Although most women wouldn’t admit
it, they generally find being physically forced into sex a real ‘turn-on.’” After
deleting three items, acceptable metric measurement invariance and reliabil-
ity was achieved (Norway: α= .86; United States: α= .93).
BJW. The six-item scale developed by Dalbert (1999)was used to measure
participants belief in a just world. Participants indicated their agreement with
statements such as, “I think the world generally is just”and “I think people
mostly get what they deserve.”The scale showed metric measurement
MEDIA PSYCHOLOGY 9
invariance and acceptable to satisfactory reliability (Norway: α= .75; United
States: α= .93).
Pornography use. Three items were adopted from Foubert et al. (2011)t
o
measure exposure to different types of pornography. On binary scales (0 =
no,1=yes) participants were asked whether they had seen scenes during the
last 12 months in videos, movies, magazines, books, or online that contained
(a) “graphic sex acts (including penetration)”(i.e., mainstream pornography);
(b) “sadomasochistic portrayals of bondage, whipping, and spanking but
without an explicit lack of consent”(i.e., sadomasochistic pornography);
and (c) “sexually explicit rape depictions in which force is used with explicit
lack of consent”(i.e., rape pornography). Rather than capturing all experi-
ences with pornography, these items focus on the specific types of porno-
graphy that have been linked to detrimental attitudes toward sexual
harassment and assault (e.g., P. J. Wright et al., 2016).
Experiences with sexual harassment and assault. Extending previous
research, we used a broad measure of experiences with sexual misconduct.
On separate binary scales (0 = no,1=yes), participants were asked
whether (a) they personally or any of their (b) acquaintances, (c) friends,
(d), family members, or (e) romantic partners had experienced sexual
harassment. The same questions were asked concerning sexual assault.
We kept participants’personal experiences separate but mean scored
questions concerning acquaintances, friends, family members, and roman-
tic partners to create a peer harassment experiences scale (Norway:
α= .74; United States: α= .85) and a peer assault experiences scale
(Norway: α= .70; United States: α= .78).
Perpetrator experiences with harassment and assault. On binary scales (0 =
no,1=yes), we also asked participants whether they thought that any of
their past behavior could be regarded as sexual harassment or rape using
the following items: “In retrospect, do you think that any of your past
behavior could be perceived as sexual harassment?”and “In retrospect, do
you think that any of your past behavior could be perceived as sexual
assault?”
Outcome variables
As we could not be certain that every participant knew about #metoo, the
following short description was presented after participants had completed
the variables presented previously and right before they completed the out-
come variables presented in the following sections:
10 J. R. KUNST ET AL.
On October 15, the American actress Alyssa Milano posted a tweet, in which she
encouraged women who had experienced sexual harassment or assault to go public
and post “me too”on social media platforms. The goal was to draw attention to the
extent of such behavior. Within 24 hours, 4.7 million people had shared the
hashtag #metoo on Facebook, and the campaign continued to grow to a global
phenomenon the following weeks.
Feelings toward #metoo. On a sliding-response scale (0 = extremely negative
to 100 = extremely positive), participants rated how they felt toward the
#metoo campaign.
Perceived benefit of the #metoo campaign. Participants were asked to indi-
cate their agreement with four statements about the perceived benefit of
#metoo, created based on a review of public responses to the campaign: “The
campaign sheds light on an important challenge that society faces;”“The
campaign is important because it gives victims of sexual assault a voice;”
“The campaign gives a good and precise picture of how wide-spread sexual
assaults actually are;”and “The campaign is important because it makes it
easier for victims of sexual assault to out themselves.”Note that these
statements did not refer to the gender of the victims, which avoided implying
that the #metoo campaign dealt only with female victims. One additional
item was not used due to a translation error (i.e., “The campaign focuses on a
lack of culture that has been overlooked for long times”). The resulting four-
item scale showed satisfactory metric measurement invariance and acceptable
to satisfactory reliability (Norway: α= .79; United States: α= .88).
Perceived harm of the #metoo campaign. Participants completed items also
created based on public responses to the campaign: “The campaign legiti-
mizes false accusations;”“The campaign does more harm than good;”“The
campaign wrongfully labels a lot of people as sexual assaulters;”and “The
campaign creates an exaggerated vigilantism/witch hunt.”After deleting one
additional item (i.e., “The campaign overdramatizes the actual number of
sexual assaults”), the scale showed metric measurement invariance and
acceptable to satisfactory reliability (Norway: α= .77; United States: α= .80).
Analyses
Analyses were conducted in R version 3.4.1 and Mplus 7.2. Due to the low
number of participants reporting other as gender, gender analyses were
conducted with participants identifying as men and women only.
MEDIA PSYCHOLOGY 11
Results
Correlations between the outcome and ideological variables are displayed in
Table 1, and correlations with demographic and experiential/behavioral
variables in Table 2. A series of 2 (country: Norway, United States) × 2
(gender: women, men) analyses of covariance were conducted to test for
country and gender differences in measures using continuous scales. For
categorical measures, corresponding logistic regressions were conducted. In
all analyses, we controlled for demographic variables that differed between
countries (i.e., age and education). In terms of outcome variables dealing
with feelings and attitudes toward #metoo, we also controlled for prior
Table 1. Zero-order correlations between individual ideologies and attitudes toward #metoo.
1. 2. 3. 4. 5. 6. 7. 8.
1. Feminist identity - −.65*** −.25*** −.48*** −.22** .45*** .46*** −.48***
2. Hostile sexism −.53*** - .35*** .67*** .21** −.51*** −.50*** .63***
3. Benevolent sexism −.23*** .29*** - .17* .35*** −.10 −.14* .16*
4. Rape myths −.27*** .66*** .30*** - .22** −.51*** −.51*** .52***
5. BJW −.24*** .22*** .39*** .15* - −.07 −.06 .18*
6. Positive feelings .50*** −.48*** −.05 −.38*** −.05 - .76*** −.71***
7. Perceived benefit .41*** −.48*** .00 −.48*** .02 .70*** - −.63***
8. Perceived harm −.47*** .69*** .21** .63*** .18** −.58*** −.62*** -
Note. Right-hand side estimates are for the Norwegian sample and left-hand side estimates for the US
sample. BJW = belief in a just world. *p< .05, **p< .01, ***p< .001.
Table 2. Correlations between demographic variables and experiences and attitudes toward
#metoo.
Positive Feelings
Toward #metoo
Perceived Benefit of
#metoo
Perceived Harm of
#metoo
Norway
United
States Norway
United
States Norway
United
States
1. Age −.06 .00 −.05 .03 −.09 −.07
2. Gender
a
−.26*** −.18** −.27*** −.20** .33*** .25***
3. Education .08 .04 .01 .00 −.18* .12
4. Living place
b
−.07 −.12 −.02 −.01 .01 .06
5. Prior knowledge about #metoo .06 .15* .08 .20** −.09 −.12
6. #metoo hashtags seen −.09 .29*** −.01 .25*** .11 −.12
7. Peer harassment experiences .08 .25*** .12 .29*** .09 −.32***
8. Personal harassment experiences .04 .17* .02 .18** −.12 −.15*
9. Peer assault experiences .10 .19** .11 .26*** −.13 −.18**
10. Personal assault experiences −.09 .14* −.22** .12 .02 −.21**
11. Harassment perpetrator −.06 .02 −.06 −.00 .08 .04
12. Assault perpetrator .02 −.00 −.07 −.03 −.03 .15*
13. Mainstream pornograpy use −.07 .06 −.11 .09 .18* −.02
14. Sadomasochistic pornography use −.05 .01 −.08 −.00 .12 .03
15. Rape pornography use −.12 −.01 −.08 −.08 .12 −.01
Note.
a
coded as 0 = women, 1 = men.
b
coded as 0 = city, 1 = rural/district. *p< .05, **p< .01, ***p< .001.
12 J. R. KUNST ET AL.
knowledge about the campaign. We focused on main effects for all variables
except for mainstream pornography and peer assault experiences, which were
the only variables for which the interaction between country and gender was
significant (full statistical details can be found in the SOM).
Country differences
Means and standard errors for the country differences can be found in the
SOM. In terms of ideologies, Norwegian participants showed higher feminist
identification, lower benevolent sexism, and lower BJW, than US partici-
pants. No differences were observed in terms of hostile sexism or rape myths.
In terms of experiential and behavioral variables, US participants consumed
more sadomasochistic and rape pornography. No additional differences were
observed for this set of variables. In terms of outcome variables, Norwegian
participants perceived the #metoo campaign as more harmful than US
participants. No differences were observed in terms of positive feelings
toward, or perceived benefit of, the #metoo campaign. This pattern of results
remained the same when analyzing only data from participants who knew
about #metoo a priori (see SOM).
Gender differences
As displayed in Figure 1, women, compared to men, displayed higher
feminist identification, F(1, 402) = 36.88, p< .001, η
2
= .09; lower hostile
sexism, F(1, 402) = 39.49, p< .001, η
2
= .08; lower benevolent sexism, F(1,
402) = 6.88, p= .009, η
2
= .02; lower rape myth acceptance, F(1, 402) = 49.40,
p< .001, η
2
= .10; and lower BJW, F(1, 402) = 4.18, p= .042, η
2
= .01. In
terms of experiential and behavioral variables, more men than women
reported being a perpetrator of sexual harassment (26.5% vs. 7.8%),
OR = 2.07, 95% CI [1.55, 2.86], z= 4.77, p< .001, and recently consuming
mainstream pornography (71.1% vs. 47.9%), OR = 1.83, 95% CI [1.46, 2.30],
z= 5.25, p< .001. By contrast, men reported fewer peer harassment experi-
ences than women (see Figure 1), F(1, 401) = 6.46, p= .011. Moreover, fewer
men than women reported personal harassment experiences (20.9% vs.
57.6%), OR = .44, 95% CI [.35, .55], z=−7.19, p< .001, and personal assault
experiences (10.0% vs. 23.5%), OR = .59, 95% CI [.43, .78], z=−3.53,
p< .001. No gender differences were observed for experiences as perpetrator
of assault (4.7%
Men
vs. 2.8%
Women
), OR = 3.40, 95% CI [.70, 24.43], z= 1.43,
p= .154; peer assault experiences (M
Men
=.26, SE
Men
=.02 vs. M
Women
=.28,
SE
Women
=.02), F(1, 400) = 1.94, p= .165; sadomasochistic pornography
consumption (28.9%
Men
vs. 21.7%
Women
), OR = 1.22, 95% CI [.95, 1.57],
z= 1.52, p= .128; or rape pornography consumption (22.3%
Men
vs. 20.3%
Women
), OR = 1.00, 95% CI [.77, 1.31], z= .03, p= .976.
In terms of outcome variables, women, compared to men, showed more
positive feelings toward #metoo, F(1, 391) = 15.69, p< .001, η
2
= .04; saw it as
MEDIA PSYCHOLOGY 13
more beneficial, F(1, 401) = 14.49, p< .001, η
2
= .04; and saw it as less
harmful, F(1, 401) = 30.81, p< .001, η
2
= .07, see Figure 1. Finally, the
interaction between country and gender was significant for mainstream
pornography consumption, OR = .67, 95% CI [.53, .84], z=−3.45,
p< .001, and peer assault experiences, F(1, 400) = 5.96, p= .015, η
2
= .01.
Specifically, men were more likely than women to consume mainstream
pornography, especially in Norway, and women reported more peer assault
experiences than men in the United States, but not in Norway (see SOM).
The pattern of results remained the same when analyzing only data from
participants who knew about #metoo a priori (see SOM).
Figure 1. Significant gender differences and response distributions are presented. Points repre-
sent estimated marginal means; error bars represent 95% confidence intervals.
14 J. R. KUNST ET AL.
Test of mediation model
We hypothesized that gender differences in reactions to #metoo would be
mediated by ideological and experiential factors that only partially align with
gender. To investigate this hypothesis, we first tested an unconstrained, fully-
saturated multigroup path model in which the associations of gender with
the outcome variables were expected to be mediated. As potential mediators,
we included the ideological factors (i.e., hostile and benevolent sexism, rape
myth acceptance, feminist identity, and BJW) and experiential/behavioral
factors (i.e., peer harassment experiences, personal harassment experiences,
personal assault experiences, and mainstream pornography consumption)
that were related to at least one outcome variable in zero-order terms and
demonstrated gender differences (see Tables 1 and 2). Given the presence of
categorical mediators, robust weighted least squares was used as the
estimator.
Next, we tested whether our model could be simplified by constraining
paths to be equal for both countries. Wald’s test showed that all parameters
could be constrained without deteriorating model fit, except for the associa-
tions between gender and personal assault experiences, between gender and
mainstream pornography consumption, and between benevolent sexism and
the perceived benefit of #metoo (see SOM). Correlations were allowed to
vary across countries. Constraining the parameters resulted in a model with
satisfactory fit, χ
2
(36) = 39.62, p= .312, CFI = .997, RMSEA = .022, 90% CI
[.000, .055]. In this model (see Figure 2), gender (0 = female,1=male)
predicted higher hostile sexism, benevolent sexism, rape myth acceptance,
and BJW, but less feminist identification, personal harassment experiences,
and peer harassment experiences in both countries. Moreover, gender pre-
dicted fewer personal assault experiences in the United States and more
mainstream pornography consumption, especially in Norway. Note that
even though paths are constrained to be equal, the standardized coefficients
can vary slightly between countries due to different standard deviations.
Of the potential ideological mediators, hostile sexism and rape myth
acceptance predicted higher perceived harm, lower perceived benefit, and
lower positive feelings toward #metoo while feminist identity had the oppo-
site relationship with each measure in both countries. In addition, benevolent
sexism predicted higher positive feelings toward #metoo. In terms of experi-
ential mediators, personal assault experiences predicted less perceived harm
of #metoo. Whereas personal harassment experiences predicted less per-
ceived benefit of #metoo, peer harassment experiences predicted higher
perceived benefit. Mainstream pornography consumption did not predict
any outcome variable.
In the model, the relationship between gender and perceived harm of
#metoo (Norway: β= .12, p= .163; United States: β= .10, p= .147) and
gender and positive feelings toward #metoo (Norway: β= -.05, p= .544;
MEDIA PSYCHOLOGY 15
United States: β= -.05, p= .539) were nonsignificant, suggesting full media-
tion, whereas the relationship between gender and the perceived benefit of
#metoo was still significant (Norway: β=–.22, p= .017; United States:
β=–.18, p= .015), suggesting partial mediation. We calculated the product
indirect relationships using the delta method for standard errors. This
method was chosen instead of bootstrapping or Monte Carlo simulation as
it is less computationally demanding, but produces very similar results, with
sample sizes akin to those in our research (Bollen & Stine, 1990). All indirect
relationships were significant, except for the relationship between gender and
perceived harm of #metoo as mediated by personal assault experiences,
between gender and the perceived benefit of #metoo as mediated by BJW,
and peer harassment experiences, and the relationship between gender and
positive feelings toward #metoo as mediated by benevolent sexism (see
Table 3).
We conducted an exploratory follow-up estimation of the model with
participants who knew about #metoo before taking part in the study only.
The model also showed close fit to this subset, χ
2
(36) = 41.66, p= .238,
CFI = .993, RMSEA = .030, 90% CI [.000, .065], yet, the relationships among
hostile sexism and the outcome variables were stronger, whereas feminist
identity showed fewer significant relationships, and rape myth acceptance
Perceived Benefit
of #metoo
Positive
Feelings
Toward
#metoo
Perceived Harm
of #metoo
Feminist
Identity
Rape
Myths
Hostile
Sexism
Benevolent
Sexism
Belief in a
Just World
R2= .12***/.08**
R2= .12**/.08**
R2= .12**/.09**
R2= .44***/.57***
R2= .39***/.42***
R2= .32***/.41***
R2= .02/.01
Personal
Assault
Experiences
Peer
Harassment
Experiences
Mainstream
Pornography
Consumption
-.34***/
-.29***
Gender:
Female (0),
Male (1)
Gender:
Female (0),
Male (1)
Personal
Harassment
Experiences
R2= .02/.01
R2= .03/.13*
R2= .20***/.20***
R2= .01/.01
R2= .19**/.03
.28***/.26***
Figure 2. Path model with standardized coefficients are displayed. Bold paths are constrained to
equality across samples. The first values represent estimates for Norway and the second values
for the United States. Only significant paths are displayed. For presentational purposes, correla-
tions between variables are not displayed. *p< .05, **p< .01, ***p< .001.
16 J. R. KUNST ET AL.
showed no significant relationships with the outcome variables (see SOM).
Results suggested that all direct associations were fully mediated.
Discussion
This study cross-culturally investigated underlying factors that explain why
men and women differ in their attitudes and feelings toward #metoo, argu-
ably one of the most pervasive and consequential social media campaigns to
date. Gender differences were primarily explained by underlying differences
in ideologies. This finding emphasizes that such gender differences are often
best conceptualized as men and women’s relative position on underlying
dimensions that only partially align with gender (Reis & Carothers, 2014;
Wood & Eagly, 2012), rather than men and women being fundamentally
different types engaged in a 'battle of the sexes' (also see Geary, 1998).
On average, participants’stance on #metoo was relatively positive, with
mean scores being above the midpoint on the perceived benefit scale and
feeling thermometer, and below the midpoint on the perceived harm scale.
However, consistent with a large body of research showing that men gen-
erally have more positive and tolerant attitudes toward rape and harassment
than women and tend to blame its victims (Anderson et al., 1997; Payne
et al., 1999; Russell & Trigg, 2004; Sakallı-Uğurlu et al., 2007), substantial
gender differences were observed on all outcome variables, with men gen-
erally being less supportive of #metoo. Importantly, these gender differences
were largely accounted for by our proposed mediators.
In terms of ideologies, men’s tendency to score higher than women on
hostile sexism, to endorse rape myths to a greater degree, and to identify less
as feminist explained why men were less positive toward #metoo than
Table 3. Indirect relationships between gender
a
and the outcome variables.
Mediator Outcome Variable
Norway United States
βSE pβSE p
Hostile sexism Perceived harm .16 .03 <.001 .13 .03 <.001
Rape myths Perceived harm .06 .02 .004 .05 .02 .006
Feminist identity Perceived harm .05 .02 .011 .04 .02 .012
Personal assault experiences Perceived harm .05 .04 .226 .10 .05 .058
Hostile sexism Perceived benefit −.08 .03 .007 −.06 .02 .005
Rape myths Perceived benefit −.07 .02 .003 −.06 .02 .004
Feminist identity Perceived benefit −.09 .03 <.001 −.08 .02 <.001
Belief in a just world Perceived benefit .01 .01 .120 .01 .01 .120
Personal harassment experiences Perceived benefit .24 .12 .035 .20 .09 .030
Peer harassment experiences Perceived benefit −.04 .02 .113 −.03 .02 .110
Hostile sexism Positive feelings −.05 .02 .027 −.05 .02 .019
Benevolent sexism Positive feelings .01 .01 .098 .01 .01 .097
Rape myths Positive feelings −.07 .02 .001 −.07 .02 .002
Feminist identity Positive feelings −.12 .03 <.001 −.12 .03 <.001
Note.
a
0 = Female, 1 = Male. Only indirect effects involving significant paths are displayed in Figure 2.
MEDIA PSYCHOLOGY 17
women, perceived it as having less benefit and as causing more harm.
Benevolent sexism was, in zero-order terms, related to more perceived
harm of #metoo in both countries, less perceived benefit of #metoo in
Norway, and was otherwise unrelated to the outcome variables. However,
although men scored higher in benevolent sexism than women, benevolent
sexism was only weakly and positively related to positive feelings to #metoo
in the multivariate path models. This finding likely resembles a suppressor
effect. Possibly, when controlling for hostile sexism and rape myth accep-
tance, the unique variance in benevolent sexism might have primarily cap-
tured the considerate and protective aspects of benevolent sexism concerning
traditional women, which could explain the positive associations (Hideg &
Ferris, 2016; Sibley & Perry, 2010). However, it is important to note that
because the indirect relationship that was mediated by benevolent sexism was
nonsignficant, it did not account for any gender differences.
BJW also played only a minor role in explaining gender differences in
reactions to #metoo, as it was only weakly related to perceiving #metoo as
being more beneficial in our model. This finding contrasts with previous
research, in which BJW predicted more negativity toward rape victims and
more tolerance toward sexual transgressions (Kleinke & Meyer, 1990; Sakallı-
Uğurlu et al., 2007; Strömwall et al., 2013). Individuals who believe in a
universal moral principle that rewards good people and punishes bad people
might see the #metoo campaign as a means to bring just punishment to those
who have committed sexual harassment and assault. However, given the
weak relationship and the lack of mediation, the role of BJW was minor in
this study.
Experiential and behavioral variables also played only minor roles in
explaining gender differences in feelings and attitudes toward #metoo. In
terms of zero-order correlations, sexual harassment and assault experiences
were consistently related to more positive and less negative attitudes toward
#metoo in the United States, but not so in Norway. Because women reported
higher personal and peer harassment experiences across countries and higher
personal assault experiences in the United States, these variables were
included as potential mediators in the path model. However, in this model
that controlled for the ideological variables as predictors, personal harass-
ment experiences, unexpectedly, predicted less perceived benefit of #metoo.
This relationship could suggest that victims of harassment believe that the
campaign has led to an inflationary use of the term sexual harassment,
thereby trivializing their own experiences. Alternatively, victims of harass-
ment might perceive hashtag campaigns as an ineffective means to achieve
social change in a legislative environment that seems to protect harassers
more than their victims. Future research could address these possibilities.
The cross-cultural scope of this study provided some indication of the
generalizability of our findings. Strikingly, the associations among most of our
18 J. R. KUNST ET AL.
main predictors and outcome variables and their mediating roles in terms of
gender differences were the same across Norway and the United States. This
finding suggests that men and women’s attitudes toward online social media
campaigns targeting misogyny might have similar roots across these two cul-
tures. This consistency also suggests that the content of the campaign did not
change markedly when it spread from the United States to Norway. Norwegians
tend to have a high proficiency in the English language, arguably due to high
consumption of US media (Education First, 2017). Thus, this lack of a language
barrier might have facilitated the spread of identical information about #metoo
and prevented the emergence of a more culturally dependent version of the
campaign in Norway. Yet, when interpreting our findings, one has to keep in
mind that Norway and the United States are relatively similar cultures. Hence, it
remains possible that attitudes and feelings toward #metoo do depend on
cultural differences on dimensions such as perceived gender egalitarianism
and injustice, which were not assessed empirically in this research.
Although the relationships among variables were fairly consistent across
countries, there were some country differences in terms of overall reactions to
#metoo. One might have expected Norwegian participants to be generally more
positive toward #metoo than US participants due to Norwegians displaying
lower sexism, lower endorsement of rape myths, and higher feminist identifica-
tion. Yet, although Norwegians, compared to US participants, showed lower
benevolent sexism, higher feminist identification, and substantially lower BJW,
Norwegians actually reported that #metoo was relatively more harmful. Hence,
Norwegians, who arguably belong to a more gender-egalitarian population,
seemed to acknowledge the potential downsides of the #metoo campaign
more than US participants did, and thus Norwegians possibly demonstrated a
more nuanced view of #metoo. It is, however, important to keep the absolute
scores in mind. Although Norwegians scored close to the midpoint of the
perceived harm scale, they clearly scored above the midpoint in terms of
perceived benefit and positive feelings (see SOM). Thus, although Norwegians
acknowledged the potentially negative aspects of the campaign more than US
participants did, they were still generally supportive. Furthermore, country
comparisons should be treated with caution, given that we used nonrepresen-
tative samples. Participants in Norway were recruited through online social
networks and US participants were recruited through an opt-in online panel.
Future research could thus test for country differences using representative
samples and, optimally, compare countries that differ even more in their gender
egalitarianism than Norway and the United States.
Practical and theoretical implications
Sexual harassment and assault are, unfortunately, still frequent phenomena
in most parts of the world. Given the advance of modern communication
MEDIA PSYCHOLOGY 19
infrastructure and the subsequent ability to cost-effectively and directly reach
out to a large number of individuals online, social media campaigns have a
unique potential to achieve social change globally and at a fast pace.
However, they are seldom uniformly received and can lead to enthusiastic
support among some groups but resistance and negativity among others.
Thus, social media campaigns have the potential to both achieve social
change and to divide and polarize social groups (e.g., men and women).
Yet, very little is known about the psychological factors underlying these
different reactions. Understanding these underlying factors can inform how
social media campaigns might be tailored to specific audiences. For instance,
based on the results from this research, one way to reduce the opposition
toward #metoo that is rooted in hostile sexism (i.e., negative attitudes toward
women) might be to highlight that the campaign raises awareness about
sexual violence experienced by both women and men.
Beyond providing insights into the mechanisms underlying gender differ-
ences in attitudes toward #metoo specifically, our research demonstrates how
applied social psychological research can help understand attitudes toward
social media campaigns more generally. For instance, our results suggest that
politicized social identities, such as feminist identity in this case, that are
well-known for predicting collective action offline (van Zomeren et al., 2008),
might have comparable effects in terms of online collective movements.
Similarly, our results suggest that the extent to which individuals’endorse-
ment of broader ideologies and beliefs either conflict or align with the goals
of an online campaign might determine whether individuals perceive the
campaign as advantageous or harmful. Speaking to the potency of such a
social psychological approach, our model was able to explain between 32%
and 57% of the variance in attitudes and feelings toward #metoo.
Limitations and future directions
It is important to note that, whereas the associations between gender and
perceived harm, and between gender and positive feelings toward #metoo
were fully mediated, the association between gender and the perceived
benefit of #metoo was only partially mediated. On the one hand, this lack
of full mediation could indicate that the perceived benefit of #metoo falls
more firmly along gender group lines. On the other hand, it is possible that
other potential mediators, unaccounted for in our research, were at play. For
instance, although pornography consumption played a minor role in our
models, it is possible that other types of media consumption might have
explained gender differences in our outcome variables. For instance, gender
differences in consumption of violent video games, soap operas, and roman-
tic-themed movies that relate to more tolerance of sexual harassment, more
tolerance of sexual assault, and more victim blaming (Dill, Brown, & Collins,
20 J. R. KUNST ET AL.
2008; Kahlor & Eastin, 2011; Ward, Seabrook, Manago, & Reed, 2016) may
have played a role. Gender-specific use of social media might also play an
important role in reactions to #metoo, given that it is an online social media
campaign. Whereas we are unaware of studies connecting social media
consumption to harassment-related beliefs or sexism, it would be of interest
to test whether the type and frequency of social media use and potentially
their interaction with such ideologies play a role. For instance, the number
and context of me too hashtags might differ across different social media
platforms. Further, it would be of interest to assess a more objective measure
of exposure to me too hashtags, as retrospective measures are prone to biases
and can be less accurate (e.g., when exposed to a certain hashtag very
frequently, people might lose a sense of the exact number of hashtags they
have seen).
We also highlight the limitation that many of the statements measuring
sexism, feminist identity, and rape myth acceptance specifically focused on
female targets. Hence, the predictive power of these ideological measures
likely depends on whether individuals perceive #metoo as primarily giving a
voice to female victims or victims of sexual violence generally. Yet the
consistent correlations between the ideological variables and attitudes toward
#metoo suggest that participants, to a large extent, perceived #metoo as
dedicated to female victims, even though male and gender nonbinary victims
also came forward in the media.
We assessed self-reported feelings toward #metoo and its perceived benefit
and harm, rather than behavioral reactions to it. To investigate the extent to
which such attitudes and feelings align with behavior, future studies might
assess behavioral support of #metoo, for instance, by measuring the extent to
which participants like or comment on different type of content related to
online campaigns and share hash tags themselves.
We also draw attention to the fact that we had to drop items from various
scales to achieve measurement invariance in this study. This makes direct
comparisons of the mean scores observed here with those of future research
challenging. For such comparisons to be meaningful, one would need to
compare means derived from factorial solutions that are invariant across the
current two populations and the new populations investigated in replications.
We are glad to share our data for such a purpose.
Finally, we encourage future work to employ a longitudinal approach.
Investigating online social media phenomena that rapidly and unexpectedly
evolve, peak, drop and reemerge, using longitudinal designs is challenging.
Even still, the cross-sectional design of our study limits the interpretation of
our findings. Although we identified consistent relationships among attitudes
toward #metoo and variables that are known to predict attitudes toward
rape- and gender-related issues over time, our data cannot speak to causality.
Indeed, given that the focus of the #metoo campaign is to challenge existing
MEDIA PSYCHOLOGY 21
misogynic attitudes and behaviors, it would be of interest to test longitudin-
ally if and for whom the campaign causes attitude change. Based on the
results from this study, future research could test whether the #metoo
campaign reduces negative attitudes toward victims (e.g., rape myths) and
women (e.g., hostile sexism) or whether being high in these ideologies a
priori makes people resistant to any possible positive effects of #metoo.
Conclusion
To conclude, this article aimed to investigate whether gender differences in
attitudes and feelings toward #metoo could be explained by underlying
ideologies and experiences. Generally consistent with a dimensional perspec-
tive on gender differences, results showed that men’s more negative stance
toward #metoo could largely be explained by men being higher in hostile
sexism, higher in rape myth acceptance, and lower in feminist identification
compared to women. These findings were consistent across two countries,
Norway and the United States, differing in their degree of gender egalitar-
ianism. Together our research provides important insights into why some
groups see social media campaigns that aim to expose the prevalence of
misogynistic behavior as beneficial, but others see them as harmful. More
generally, it demonstrates how social psychological approaches can improve
understanding of attitudes toward social media campaigns across cultures.
Notes
1. The supplementary materials include outlier analyses, measurement invariance tests,
detailed statistical information for group comparisons and follow-up path analyses.
Acknowledgments
We thank Kristoffer Asp, Frida Bordenich Ballo, Inger Helene Godø, Anette Haug, Felix
Koppe, Annahita Sayadian, August Schartau, Henriette Ekeland Svenningsen, Lina Tosterud,
Frøydis Johannessen Woldstad, and Aisha Zarar for helping designing the study and collect-
ing the data in Norway.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This work was supported by the Norwegian Research Council [231157/F10]
22 J. R. KUNST ET AL.
ORCID
April Bailey http://orcid.org/0000-0002-9427-8542
References
Abrams, D., Viki, G. T., Masser, B., & Bohner, G. (2003). Perceptions of stranger and
acquaintance rape: The role of benevolent and hostile sexism in victim blame and rape
proclivity. Journal of Personality and Social Psychology,84(1), 111–125. doi:10.1037/0022-
3514.84.1.111
Allen, M., Emmers, T., Gebhardt, L., & Giery, M. A. (1995). Exposure to pornography and
acceptance of rape myths. Journal of Communication,45(1), 5–26. doi:10.1111/j.1460-
2466.1995.tb00711.x
Allport, G. W. (1954). The nature of prejudice. Cambridge, UK: Addison Wesley.
Anderson, K. B., Cooper, H., & Okamura, L. (1997). Individual differences and attitudes
toward rape: A meta-analytic review. Personality and Social Psychology Bulletin,23(3), 295–
315. doi:10.1177/0146167297233008
Astor, M. (2017, December 14). Texas attorney general’s aide resigns after mocking #MeToo
movement. Retrieved from https://www.nytimes.com/2017/12/14/us/andrew-leonie-texas-
attorney-general.html
Ayres, M. M., Friedman, C. K., & Leaper, C. (2009). Individual and situational factors related
to young women’s likelihood of confronting sexism in their everyday lives. Sex Roles,61(7),
449–460. doi:10.1007/s11199-009-9635-3
Bareket, O., Kahalon, R., Shnabel, N., & Glick, P. (2018). The Madonna-Whore dichotomy:
Men who perceive women’s nurturance and sexuality as mutually exclusive endorse
patriarchy and show lower relationship satisfaction. Sex Roles. doi:10.1007/s11199-018-
0895-7
BBC. (2018, January 10). Catherine Deneuve defends men’s‘right to hit on’women. BBC.
Retrieved from http://www.bbc.com/news/world-europe-42630108
Bollen, K. A., & Stine, R. (1990). Direct and indirect effects: Classical and bootstrap estimates
of variability. Sociological Methodology,20, 115–140. doi:10.2307/271084
Bratton, K. A., & Ray, L. P. (2002). Descriptive representation, policy outcomes, and muni-
cipal day-care coverage in Norway. American Journal of Political Science,46(2), 428–437.
doi:10.2307/3088386
CBS. (2017, October 17). More than 12M “me too”Facebook posts, comments, reactions in 24
hours. Retrieved from https://www.cbsnews.com/news/metoo-more-than-12-million-face
book-posts-comments-reactions-24-hours/
Chen, F. F. (2008). What happens if we compare chopsticks with forks? The impact of
making inappropriate comparisons in cross-cultural research. Journal of Personality and
Social Psychology,95(5), 1005–1018. doi:10.1037/a0013193
Dalbert, C. (1999). The world is more just for me than generally: About the personal belief in
a just world scale’s validity. Social Justice Research,12(2), 79–98. doi:10.1023/
a:1022091609047
Dill, K. E., Brown, B. P., & Collins, M. A. (2008). Effects of exposure to sex-stereotyped video
game characters on tolerance of sexual harassment. Journal of Experimental Social
Psychology,44(5), 1402–1408. doi:10.1016/j.jesp.2008.06.002
Education First. (2017). English proficiency index. Retrieved from https://www.ef.no/epi/
MEDIA PSYCHOLOGY 23
Fallon, C. (2018, December 1). A feminist dystopia for the #metoo moment. Huffington Post.
Retrieved from https://www.huffingtonpost.com/entry/the-power-naomi-alderman-book_
us_5a20763de4b03350e0b55dfe
Foubert, J. D., Brosi, M. W., & Bannon, R. S. (2011). Pornography viewing among fraternity
men: Effects on bystander intervention, rape myth acceptance and behavioral intent to
commit sexual assault. Sexual Addiction & Compulsivity,18(4), 212–231. doi:10.1080/
10720162.2011.625552
Fowers, A. F., & Fowers, B. J. (2010). Social dominance and sexual self-schema as moderators
of sexist reactions to female subtypes. Sex Roles,62(7), 468–480. doi:10.1007/s11199-009-
9607-7
Fritz, N., & Paul, B. (2017). From orgasms to spanking: A content analysis of the agentic and
objectifying sexual scripts in feminist, for women, and mainstream pornography. Sex Roles,
77(9), 639–652. doi:10.1007/s11199-017-0759-6
Geary, D. C. (1998). Male, female: The evolution of human sex differences. Washington, DC:
American Psychological Association.
Glick, P., & Fiske, S. T. (1996). The ambivalent sexism inventory: Differentiating hostile and
benevolent sexism. Journal of Personality and Social Psychology,70(3), 491–512.
doi:10.1037/0022-3514.70.3.491
Glick, P., Fiske, S. T., Mladinic, A., Saiz, J. L., Abrams, D., Masser, B., & López, W. L. (2000).
Beyond prejudice as simple antipathy: Hostile and benevolent sexism across cultures.
Journal of Personality and Social Psychology,79(5), 763–775. doi:10.1037/0022-
3514.79.5.763
Gundersen, A, & Kunst, J.R. (2018). Feminist ≠Feminine? Feminist Women are Visually
Masculinized Whereas Feminist Men are Feminized. Sex Roles. doi:10.1007/s11199-018-
0931-7
Hald, G. M. (2006). Gender differences in pornography consumption among young hetero-
sexual danish adults. Archives of Sexual Behavior,35(5), 577–585. doi:10.1007/s10508-006-
9064-0
Hideg, I., & Ferris, D. L. (2016). The compassionate sexist? How benevolent sexism promotes
and undermines gender equality in the workplace. Journal of Personality and Social
Psychology,111(5), 706–727. doi:10.1037/pspi0000072
Hofstede, G., Hofstede, G. J., & Minkov, M. (2010). Cultures and organizations, software of
the mind. New York, NY: McGraw-Hill.
Jorgensen, T. D., Pornprasertmanit, S., Schoemann, A. M., & Rosseel, Y. (2018). semTools:
Useful tools for structural equation modeling. R package version 0.5-0. Retrieved from
https://CRAN.R-project.org/package=semTools
Kahlor, L., & Eastin, M. S. (2011). Television’s role in the culture of violence toward women:
A study of television viewing and the cultivation of rape myth acceptance in the United
States. Journal of Broadcasting & Electronic Media,55(2), 215–231. doi:10.1080/
08838151.2011.566085
Keldsen, T. (2018). Er menn og kvinner splittet om effekten av Me Too-kampanjen? [Are men
and women divided on the effect of the Me Too campaign?]. Retrieved from https://
yougov.no/news/2018/03/08/er-menn-og-kvinner-splittet-om-effekten-av-me-too-/
Kleinke, C. L., & Meyer, C. (1990). Evaluation of rape victim by men and women with high
and low belief in a just world. Psychology of Women Quarterly,14(3), 343–353.
doi:10.1111/j.1471-6402.1990.tb00024.x
Lee, T. L., Fiske, S. T., & Glick, P. (2010). Next gen ambivalent sexism: Converging correlates,
causality in context, and converse causality, an introduction to the special issue. Sex Roles,
62(7), 395–404. doi:10.1007/s11199-010-9747-9
24 J. R. KUNST ET AL.
Lerner, M. J. (1980). The belief in a just world: A fundamental delusion. New York, NY:
Springer.
Masser, B., Lee, K., & McKimmie, B. M. (2010). Bad woman, bad victim? Disentangling the
effects of victim stereotypicality, gender stereotypicality and benevolent sexism on acquain-
tance rape victim blame. Sex Roles,62(7), 494–504. doi:10.1007/s11199-009-9648-y
Masser, B., Viki, G. T., & Power, C. (2006). Hostile sexism and rape proclivity amongst men.
Sex Roles,54(7), 565–574. doi:10.1007/s11199-006-9022-2
McCabe, J. (2005). What’s in a label? The relationship between feminist self-identification
and “feminist”attitudes among U.S. women and men. Gender & Society,19(4), 480–505.
doi:10.1177/0891243204273498
Miller, A. K., Amacker, A. M., & King, A. R. (2011). Sexual victimization history and
perceived similarity to a sexual assault victim: A path model of perceiver variables
predicting victim culpability attributions. Sex Roles,64(5), 372–381. doi:10.1007/s11199-
010-9910-3
Myaskovsky, L., & Wittig, M. A. (1997). Predictors of feminist social identity among college
women. Sex Roles,37(11), 861–883. doi:10.1007/bf02936344
Nelson, J. A., Liss, M., Erchull, M. J., Hurt, M. M., Ramsey, L. R., Turner, D. L., & Haines, M.
E. (2008). Identity in action: Predictors of feminist self-identification and collective action.
Sex Roles,58(9–10), 721–728. doi:10.1007/s11199-007-9384-0
NSVRC. (2015). Statistics about sexual violence. Retrieved from https://www.nsvrc.org/sites/
default/files/publications_nsvrc_factsheet_media-packet_statistics-about-sexual-violence_
0.pdf
Ohlheiser, A. (2017). The woman behind ‘Me Too’knew the power of the phrase when she
created it —10 years ago. Retrieved from https://www.washingtonpost.com/news/the-
intersect/wp/2017/10/19/the-woman-behind-me-too-knew-the-power-of-the-phrase-
when-she-created-it-10-years-ago/?utm_term=.23d02a96bdbc
Parliamentary Union. (2017). Proportion of seats held by women in national parliaments (%).
The World Bank. Retrieved from https://data.worldbank.org/indicator/SG.GEN.PARL.ZS?
view=chart&year_high_desc=true
Page, T. E., Pina, A., & Giner-Sorolla, R. (2016). “It was only harmless banter!”The
development and preliminary validation of the moral disengagement in sexual harassment
scale. Aggressive Behavior,42(3), 254–273. doi:10.1002/ab.21621
Payne, D. L., Lonsway, K. A., & Fitzgerald, L. F. (1999). Rape myth acceptance: Exploration of
its structure and its measurement using the Illinois rape myth acceptance scale. Journal of
Research in Personality,33(1), 27–68. doi:10.1006/jrpe.1998.2238
Reis, H. T., & Carothers, B. J. (2014). Black and white or shades of gray:Are gender
differences categorical or dimensional? Current Directions in Psychological Science,23(1),
19–26. doi:10.1177/0963721413504105
Rotundo, M., Nguyen, D.-H., & Sackett, P. R. (2001). A meta-analytic review of gender
differences in perceptions of sexual harassment. Journal of Applied Psychology,86(5), 914–
922. doi:10.1037/0021-9010.86.5.914
Russell, B. L., & Trigg, K. Y. (2004). Tolerance of sexual harassment: An examination of
gender differences, ambivalent sexism, social dominance, and gender roles. Sex Roles,50
(7), 565–573. doi:10.1023/B:SERS.0000023075.32252.fd
Sakallı-Uğurlu, N. (2010). Ambivalent sexism, gender, and major as predictors of Turkish
college students’attitudes toward women and men’s atypical educational choices. Sex
Roles,62(7), 427–437. doi:10.1007/s11199-009-9673-x
Sakallı-Uğurlu, N., Yalçın, Z. S., & Glick, P. (2007). Ambivalent sexism, belief in a just world,
and empathy as predictors of Turkish students’attitudes toward rape victims. Sex Roles,57
(11), 889–895. doi:10.1007/s11199-007-9313-2
MEDIA PSYCHOLOGY 25
Sedgwick, J. L. (2006). Criminal victimization in the United States, 2005 Statistical tables:
National crime victimization survey. Retrieved from Washington, US-DC: https://www.bjs.
gov/content/pub/pdf/cvus05.pdf
Sibley, C. G., & Perry, R. (2010). An opposing process model of benevolent sexism. Sex Roles,
62(7), 438–452. doi:10.1007/s11199-009-9705-6
Strömwall, L. A., Alfredsson, H., & Landström, S. (2013). Blame attributions and rape: Effects
of belief in a just world and relationship level. Legal and Criminological Psychology,18(2),
254–261. doi:10.1111/j.2044-8333.2012.02044.x
Suarez, E., & Gadalla, T. M. (2010). Stop blaming the victim: A meta-analysis on rape myths.
Journal of Interpersonal Violence,25(11), 2010–2035. doi:10.1177/0886260509354503
Szymanski, D. M. (2004). Relations among dimensions of feminism and internalized hetero-
sexism in lesbians and bisexual women. Sex Roles,51(3), 145–159. doi:10.1023/b:
sers.0000037759.33014.55
Tajfel, H. (1982). Social identity and intergroup relations. Cambridge, England: Cambridge
University Press.
Tannen, D. (1990). You just don’t understand. New York, NY: Ballantine Books.
UNDP. (2016). Table 5: Gender inequality index. Retrieved from http://hdr.undp.org/en/
composite/GII
UNODC. (2017). Crime and criminal justice. Retrieved from Vienna, Austria: https://data.
unodc.org/.
Vagianos, A. (2017, October 18). In response to #metoo men are tweeting #HowIWillChange.
Huffington Post. Retrieved from https://www.huffingtonpost.com/entry/in-response-to-
metoo-men-are-tweeting-howiwillchange_us_59e79bd3e4b00905bdae455d
van der Bruggen, M., & Grubb, A. (2014). A review of the literature relating to rape victim
blaming: An analysis of the impact of observer and victim characteristics on attribution of
blame in rape cases. Aggression and Violent Behavior,19(5), 523–531. doi:10.1016/j.
avb.2014.07.008
van Zomeren, M., Postmes, T., & Spears, R. (2008). Toward an integrative social identity
model of collective action: A quantitative research synthesis of three socio-psychological
perspectives. Psychological Bulletin,134(4), 504–535. doi:10.1037/0033-2909.134.4.504
Ward, L. M., Seabrook, R. C., Manago, A., & Reed, L. (2016). Contributions of diverse media
to self-sexualization among undergraduate women and men. Sex Roles,74(1), 12–23.
doi:10.1007/s11199-015-0548-z
Whatley, M. A. (1993). Belief in a just world scale: Unidimensional or multidimensional? The
Journal of Social Psychology,133(4), 547–551. doi:10.1080/00224545.1993.9712180
Wood, W., & Eagly, A. H. (2012). Biosocial construction of sex differences and similarities in
behavior. In M. Zanna & J. M. Olson (Eds.), Advances in experimental social psychology
(Vol. 46, pp. 55–123). San Diego, CA: Academic Press.
World Economic Forum. (2016). Global gender gap report 2016. Retrieved from http://
reports.weforum.org/global-gender-gap-report-2016/rankings/
Wright, P. J., Tokunaga, R. S., & Kraus, A. (2016). A meta-analysis of pornography con-
sumption and actual acts of sexual aggression in general population studies. Journal of
Communication,66(1), 183–205. doi:10.1111/jcom.12201
Wright, T. (2018, February 11). Filmmaker Michael Haneke calls the #MeToo movement a
‘witch hunt’.Vulture. Retrieved from http://www.vulture.com/2018/02/director-michael-
haneke-calls-metoo-movement-a-witch-hunt.html
Yoder, J. D., Tobias, A., & Snell, A. F. (2011). When declaring “I am a feminist”matters:
Labeling is linked to activism. Sex Roles,64(1), 9–18. doi:10.1007/s11199-010-9890-3
26 J. R. KUNST ET AL.
