Skip to main content
NCBI home page
International Journal of Sexual Health logoLink to International Journal of Sexual Health
. 2021 Aug 27;34(1):130–143. doi: 10.1080/19317611.2021.1944420

No Gender Differences in Enrollment for Replicated, Minimally-Different ‘Pornographic’ Versus ‘Photographic’ Studies

Melanie MacEacheron a,✉, Taylor Kohut b, William A Fisher b
PMCID: PMC10903555  PMID: 38595682

Abstract

Objective: Since women tend to use pornography less, they may enroll less in studies concerning it and/or those who do may be gender-atypical. Methods: One study plus replication, assessed proportion of participants reporting being women, responding to each of two, minimally different (one including the word “pornographic”) study advertisement versions, and their pornography use frequency, Erotophilia-Erotophobia, and Openness to Experience. Results: Proportion responding to each version did not differ. In one sample only, women responding to one version differed in Openness to Experience. Conclusions: Advertising to North American convenience samples using the word “pornographic”, may produce neither self-selection out by, nor over-sampling of gender-atypical, women.

Keywords: Pornography, gender, self-selection into studies, minimal difference procedure, survey research

While some women use pornography (Wright et al., 2013), it is well documented that men tend to use pornography more than women (Bridges & Morokoff, 2011; Carroll et al., 2008; Daneback et al., 2009; Hald, 2006; Kinsey et al., 1953; Maddox et al., 2011; O’Reilly et al., 2007; Petersen & Hyde, 2010). This result has been seen cross-culturally, at least in the past (Abelson et al., 1971; see also Archer, 2019, for a review including cross-national data), as well as very recently in the Czech Republic (Bártová et al., 2020), in Canada (among both “heterosexual cisgender” as well as “sexual and gender minority” adolescent boys, compared with adolescent girls of either such grouping: Bőthe et al., 2020), as well as in US studies. Further, men have been shown to possess more favorable attitudes toward pornography than women (e.g. Carroll et al., 2008; Kohut, 2014; Wilson & Abelson, 1973). Women’s and men’s arousal responses to pornography, however, have often been found to be equivalent (e.g. Fisher, 1983; Fisher & Byrne, 1978a, 1978b; Saunders et al., 1985; but see Mosher & MacIan, 1994). As ethics review boards require that studies involving pornography disclose the nature of the research before would-be participants choose to participate, it seems possible that women selectively avoid studies involving pornography, and/or that women enrolling in such studies differ systematically from women more generally. If either or both cases are true, then conclusions regarding female participants from such studies may be of limited generalizability.

Despite this concern, it appears that only three studies directly examining possible gender or sex self-selection into studies involving pornography have ever been conducted (Kenrick et al., 1980; Saunders et al., 1985; Wolchik et al., 1985): none within the last 30 years. All of these studies were conducted with university students. Kenrick et al. (1980: in Study 2) found self-selection into studies involving viewing pornography by sex, though it is noteworthy that when comparing androgynous1 male and female participants, sex differences in participation disappeared. Saunders et al. (1985) did not find any difference in enrollment sex ratios, for a study concerning pornography and an analogous one concerning personality. These authors, however, did find that 7 female participants out of the 33 who signed up for the non-sexual study did not agree to participate in a second study involving watching pornography, compared to 1 out of 30 male participants who similarly did not. Wolchik et al. (1985) found no self-selection by gender. The period of 1985 to the current year, however, saw substantial change in terms of the availability of pornography (e.g. Bradley, 2018; Cooper, 1998; Fisher & Barak, 2001: see also Blake et al., 2018). It is therefore plausible that attitudes toward pornography, including among women, have changed since these studies involving sex or gender self-selection were conducted. Given these inconsistent findings as well as potential changes in gendered attitudes toward pornography, it remains unclear if gender self-selection effects compromise the generalizability of contemporary research.

In addition to potentially being out of date, two of the above studies possess clear limitations. Kenrick et al. (1980) recruited via direct telephone invitation from a live Research Assistant, seeking participation in either a neutral experiment or one involving pornography. This aspect of the design may have induced some to have chosen the former experiment and not the latter, due to social desirability response bias. Wolchike et al. (1985) asked potential participants who had already signed up for a study involving sexuality whether they would also be willing to participate in a study involving pornography. Thus, in this study, those who were asked to participate in the study concerning pornography may already have self-selected as more open to research concerning sexuality more generally, which may have made them more open to participating in subsequent research involving pornography, compared with potential participants who would not have elected to participate in the initial, sexuality study. Thus, those participating in sexual studies generally may be dissimilar to other studies’ participants.

Even if women do not tend, as a group, to self-select out of studies concerning a particular topic, individual women who do participate may vary systematically from women who do not. The decision to use pornography is often assumed to be a purposive process (Kohut, 2014; Paul & Shim, 2008), regulated by both approach and avoidance motivations. Women differ from men in their motivations for using internet pornography (Paul & Shim, 2008). Moreover, women (as well as, presumably, men) vary in the extent they find sexual stimuli and pornography unpleasant and aversive (Fisher et al., 1988; Kohut, 2014). Finally, Wolchik et al. (1985, p. 104), showed that women who signed up for a study involving sexuality then also signed up, as often as men, for a study involving pornography: These authors noted all their study participants were more likely to “… seek out new and unusual experiences more often than nonvolunteers…”. Taken together, these findings and assumptions indicate it may be that women who enroll in a study involving exposure to pornography are more experienced with pornography, more erotophilic, and more likely to seek out “new and unusual experiences” than women who do not enroll in such a study.

Current study

The aim of this study is to test whether, at a North American university, (1) a greater proportion of women would sign up for a study advertised as including “photographic” images versus an otherwise-identical study advertised as including “pornographic” images, and (2) whether women who enrolled under the latter advertisement would score higher than women who enrolled under the former in three individual difference variables apparently relating to pornography use. Those variables were frequency of pornography use, degree of erotophilia (Sexual Opinions Survey: Fisher et al., 1988), and Openness to Experience (DeYoung et al., 2007). Finally, we replicated this study, which was conducted during the final two weeks of participant pool operation in April, the following September. This replication was conducted, to rule out the possibility that results obtained during the final two weeks of participant pool operation may have minimized women’s choices and obscured the differences we aimed to explore.

Methods

Procedure

The methodology of this research relied on a participant recruitment system employed by the Department of Psychology of a mid-sized, research intensive, Canadian university. At this institution, students enrolled in introductory psychology earn 10% of their grade by either participating in studies as part of a departmental participant pool, or by writing a short paper concerning research. Participant recruitment is managed by a university-run website (“SONA”: Sona Systems Ltd., 1997–2016). After logging in to this system, potential participants can view brief descriptions of the studies that are available, and select which, if any, they wish to participate in. Two samples were collected using the procedure outlined below. The first was collected during the final two weeks the recruitment system was open, in April of 20132. The second sample (a replication) was collected during the first two weeks the recruitment system was open in the next academic year: September 2013. The participant pool was comprised almost entirely of introductory psychology students, so very few individuals would have been able to see the advertisement for both executions of the study. It should be noted, however, that because participants were debriefed after participation, those participating in the first execution might have unblinded those participating in the second as to hypotheses.

In each execution, two minimally different study advertisements were posted on the recruitment system, one at a time, in alternating order. Each advertisement had the same title (“Look at Images”), and was otherwise identical except for one word. Specifically, the advertisement was for either a study involving viewing “pornographic” or “photographic” images. When the study was executed in April of 2013, the “photographic” version was posted first (chosen by coin toss). This advertisement was replaced after 24 h by the “pornographic” version. Study advertisements continued to alternate every 24 h until one advertisement recruited n = 85 total participants, at which point the other advertisement was run continuously until it, too, recruited n = 85 participants.3 In order to counter-balance initial version presentation, the “pornographic” advertisement version was posted first in the September execution. In this execution, n = 85 participants were recruited in the first 24-h period (i.e. all available participation slots under the “pornographic” version of the advertisement were filled). Due to an administrative error, the “pornographic” version was not replaced by the “photographic” version of the study advertisement immediately following this period. To minimize possible confounding of the manipulation – the word “pornographic” versus “photographic” in the advertisement – with a day of the week effect, we posted the “photographic” version of the advertisement exactly one week after the date and time of the first (i.e. “pornographic”) posting. This advertisement was left up until it recruited n = 85 participants, which occurred in approximately 36 h.

Regardless of the source of recruitment, all participants were directed to complete an identical, web-based survey that assessed individual difference variables expected to vary as a function of the recruitment advertisement under which participants had enrolled. The proportion of women enrolling under each of these advertisement versions constituted the dependent variable of primary interest.4 Only participants who stated their gender as “male” or “female” were included in the final samples. A sample size of 170 was chosen for each execution in order to achieve statistical power of 0.80, given a moderate effect size and alpha = 0.05, when performing the planned, statistical tests. These tests were (1) chi-squared testing for difference in proportion of participants who were women, signing up under the two advertisement versions, and (2) testing for mean differences in individual difference variables between those signing up under each advertisement version and by gender (Cohen, 1992). When a participant enrolled in both versions of the study, as determined by comparison of anonymous SONA system identifiers (April Execution n = 10: September Execution n = 5), only the data they provided when taking the study the first time was used in analyses. Such individuals’ double-enrollments in the study constitute one reason each of the two final sample sizes did not reach 85 × 2 = 170. The other reason was some participants did not provide usable data (i.e. did not report gender, required to calculate our primary dependent variable). In the April execution n = 8 did not provide usable data, including one stating a preference not to report gender. In the September execution, n = 5 did not provide usable data. Note that all results presented include data from participants not reporting heterosexual orientation (including not answering the question: n = 9 in April, n = 16 in September): all tests of hypotheses were re-run with such participants’ data omitted and showed no differences.

Materials

The complete advertisement for both executions of the study was as follows:

Look at images

This study asks different people to look at a variety of pornographic (photographic) images. Then, we ask for your personal opinions on a variety of issues. It will take about 30 minutes of your time to complete, and you will receive half of a research credit. You must be at least 18 years old to participate. This study would be conducted by <authors’ names and highest degrees listed>. Interested? Sign up!

In each execution, participants’ gender, sexual orientation, and other demographic information were queried. This was followed by the manipulation check, for the September execution. Next, several individual difference measures were administered: Fisher et al. (1988) 21 item Erotophilia-Erotophobia scale, assessing individual differences in their trait-like tendency to respond to sexual stimuli along a continuum of negative to positive affect; a single scale item measuring pornography use frequency; and DeYoung et al. (2007) 10 item Openness to Experience scale.5 “Pornography” was not defined, due to a desire to not repeat mention of the topic beyond the least necessary to allow for any self-selection out of such studies and to avoid priming the concept. Note that pornography use frequency could range from 1 (“I have never looked at (or watched) pornography”) to 5 (“3 times per week or more”); Erotophilia-Erotophobia scores could range from 0 to 126 (Fisher et al., 1988: Cronbach’s alpha in males = 0.88, and in females = 0.90), and Openness to Experience scores could range from 1 to 5 (DeYoung et al., 2007: Cronbach’s alphas ranged from 0.72 to 0.78 overall, and between 0.72 and 0.77 between two university samples). Higher scores indicated higher pornography use, more erotophilia, or more openness to experience. No images of any kind were actually displayed to participants. Participants received course credit in exchange for participation. This procedure was reviewed and approved by the university’s research ethics board before implementation.

Participants

April execution

A total of N = 153 participants reporting gender (n = 68 reporting “female” gender; n = 85 reporting “male” gender) were recruited in April 2013. All but nine participants (n = 5 females and n = 2 males reporting homosexual orientation, and n = 2 males not reporting sexual orientation) reported heterosexual orientation. Participants ranged in age from 18 to 23 years (M = 18.82, SD = 0.91). 1.96% of participants reported race/ethnicity as African or Black; 27.45% reported as Asian or Pacific Islander; 50.98% reported as European or Caucasian; 3.92% reported as Native Canadian or Amerindian; 6.54% reported as Bi- or Multi-Racial; 4.58% reported “North African/Arab”, “Middle Eastern”, “Arab”, “Israeli”, or “Israeli, Moroccan, Canadian”; and 5.23% stated preference not to report race/ethnicity or reported “Canadian”. Most participants (66.67%) reported relationship status as “Single“; 3.27% reported as “Dating my current partner and others”; 29.41% reported as “Dating my current partner exclusively”; and 0.65% stated preference not to report relationship status. (None reported being married.)

September execution

A total of N = 165 participants reporting gender (n = 122 reporting “female” gender; n = 43 reporting “male” gender) were recruited in September 2013. All but eight participants (n = 5 females and n = 3 males, with one of the females reporting bisexual orientation and the remainder reporting homosexual orientation, plus 1 male stating he preferred not to report orientation) reported heterosexual orientation. Participants ranged in age from 18 to 36 years (M = 18.80, SD = 1.95). 4.24% of participants reported race/ethnicity as African or Black; 32.73% reported as Asian or Pacific Islander; 55.15% reported as European or Caucasian; 0.61% reported as Hispanic or Latino/a; 1.21% reported as Native Canadian or Amerindian; 4.85% reported as Bi- or Multi-Racial; and 1.82% stated preference to not answer, simply did not answer, or reported “Canadian”. Most participants (60.00%) reported relationship status as “Single”; 4.85% reported as “Dating my current partner and others”; 33.94% reported as “Dating my current partner exclusively”; 1.21% reported as “Married”; and no participants stated preference not to report relationship status. Following a colleague’s comment in April that participants might fail to notice the word “pornographic” if it appeared in the study advertisement, a manipulation check was added to the September execution. Immediately after reporting demographic data, participants were asked to type in a brief description of the study advertisement. Of the 82 participants answering the question who had enrolled under the “photographic” version, two used a word derived from the word-stem “porn”, or synonym. Of the 78 participants answering the question who had enrolled under the “pornographic” version, 54 used such a word or synonym.

Results

Proportion female enrolling under each advertisement version

In the April execution, the proportion of participants who were women, enrolling under the “photographic” version of the study (48%), did not differ from the proportion of participants who were women, enrolling under the “pornographic” version (39%): Pearson χ2 (1)=1.32, p=.25. Similarly, the proportion of participants who were women, enrolling under the “photographic” version in the September execution (76%), was also not significantly different from the proportion of participants who were women enrolling under the “pornographic” version (69%): Pearson χ2 (1)=1.07, p=.30. When the samples were combined the proportions of participants who were women enrolling under each of the two advertisement versions, were also not significantly different (63% for those enrolling under the “photographic” advertisement; 55% for those enrolling under the “pornographic” advertisement): Pearson χ2 (1)=2.08, p=.15.

Testing underlying assumptions about gender

Pornography use frequency was lower in women compared with men (see Table 1 for means and SD’s: for April, one-tailed t(150)=12.80, p<.0001, Cohen’s d = 2.12 or very large; for September, one-tailed t(70)=10.74, p<.0001, Cohen’s d = 1.97 or very large). Also lower in women compared with men, was erotophilia (for April, one-tailed t(147)=4.51, p<.0001, Cohen’s d = 0.73 [a moderate effect size]; for September, one-tailed t(89)=5.28, p<.0001, Cohen’s d = 0.90 or large). Although we had no gendered expectations with respect to Openness to Experience, we examined this variable as well. In the September execution only, a gender difference emerged: women scored higher than men in Openness to Experience (two-tailed t(82)= −2.40, p = 0.02, Cohen’s d = 0.41 or small).

Table 1.

Individual Difference Measures in Order of Administration, by Execution for Each Gender.

  Men
     Women
    
April
 September
     April
 September
    
M SD n M SD n t d M SD n M SD n t d
SOS 80.32 18.05 84 82.29 17.25 41 −0.59 – 67.84 15.91 67 64.41 21.97 115 1.18 –
PUF 3.95 1.04 85 3.76 0.94 41 1.06 – 1.99 0.79 67 1.92 0.93 115 0.47 –
OtE 3.56 0.60 85 3.56 0.45 41 1.18 – 3.73 0.53 67 3.76 0.53 113 −0.36 –
Open in a new tab

Notes. d = Cohen’s d (for significant t-tests); SOS: Erotophilia (Fisher et al., 1988); PUF: pornography use frequency; OtE: openness to experience (DeYoung et al., 2007); for all scales greater scores indicate greater such frequency/possession of the trait; all t-tests evaluate differences between replications (April or September) and are two-way; No t-tests were significant.

Differences in pornography use, erotophilia, and openness to experience

Among April participants, number reporting each of “I have never looked at (or watched) pornography” was 18 (12%); “Less than once a month” was 45 (29%); “1–3 times per month” was 29 (19%); “1–2 times per week” was 26 (17%); and “3 times per week or more” was 35 (23%). Among September participants, number reporting each of “I have never looked at (or watched) pornography” was 43 (27%); “Less than once a month” was 54 (34%); “1–3 times per month” was 28 (18%); “1–2 times per week” was 23 (14%); and “3 times per week or more” was 12 (8%).

Across genders, erotophilia was non-significantly albeit marginally greater, among participants who enrolled under the “pornographic” advertisement in the April execution6 (see Table 2 for means and SD’s: one-tailed t(150)=1.46, p = 0.07, Cohen’s d = 0.24 or small). Similarly, in both executions, pornography use frequency was non-significantly albeit marginally greater, among participants who enrolled under the “pornographic” advertisement (for April, one-tailed t(150)=1.39, p=.08, Cohen’s d = 0.22 or small; for September, one-tailed t(154)=1.50, p=.07, Cohen’s d = 0.24 or small).

Table 2.

Individual Difference Measures in Order of Administration, with Genders’ Data Pooled, by Execution for Each Advertisement Version.

  “Pornography”
     “Photography”
    
April
 September
 t d April
 September
 t d
M SD n M SD n     M SD n M SD n    
SOS 76.80 17.80 74 70.17 22.22 78 2.04* 0.33 72.84 18.44 77 68.05 22.33 78 1.46 –
PUF 3.24 1.36 75 2.55 1.24 78 3.26** 0.52 2.94 1.34 77 2.26 1.21 78 3.31** 0.53
OtE 3.81 0.54 75 3.71 0.52 77 1.10 – 3.60 0.51 77 3.70 0.51 77 −1.15 –
Open in a new tab

Notes. d = Cohen’s d (for significant t-tests); SOS: erotophilia (Fisher et al., 1998); PUF: pornography use frequency; OtE: openness to experience (DeYoung et al., 2007); for all scales greater scores indicate greater such frequency/possession of the trait; all t-tests evaluate differences between replications (April or September) and are two-way; **p<.01; *p<.05; †p<.1.

These differences in the (gender-pooled) data may have been due to (non-significantly) greater male participation in the “pornographic” study version. (In other words, we were concerned that [non-significantly] more males enrolling under the “pornographic” version – not the fact of that version being “pornographic”– may have led to greater pornography use frequency for that version.) Thus, we tested whether gender interacted with survey advertisement version enrolled under, to predict each of the three DVs (i.e. pornography use frequency, as well as erotophilia, and Openness to Experience). This was done via six regressions: one each for each such measure for each of the two executions. Each of these six used gender, advertisement version enrolled under, and their interaction, as predictors. These regressions showed main effects only by gender. Note, however, that in one case only, an interaction in one execution, only, between gender and advertisement version enrolled under, was found. That is, women in the September execution who signed up under the “pornographic” advertisement version had higher Openness to Experience than those women in the same execution having enrolled under the other version. Adjusted-R2 in this regression, however, was inadequate for purposes of reporting (Ferguson, 2009). In any case, these post-hoc analyses were underpowered to detect any but moderate or greater effects at α = 0.05 (Cohen, 1992).7

In the April execution, only, Openness to Experience was significantly greater among those having enrolled under the “pornographic” advertisement (one-tailed t(150)=2.30, p=.01, Cohen’s d = 0.38 or small).

Gender and advertisement version enrolled under

As can be seen in Table 3,8 among women in the April execution only, a statistically significant difference in Openness to Experience was observed across advertisement versions. It was in the predicted direction: Women who enrolled under the “pornographic” advertisement scored higher in Openness to Experience (one-tailed t(64)=2.54, p=.008, Cohen’s d =0.64 – a moderate effect size). There was also non-significant (albeit marginal) difference in pornography use frequency among women within the September execution: Women who enrolled under the “pornographic” advertisement reported more pornography use (one-tailed t(110)=1.46, p=.07, Cohen’s d = 0.28 or small). As the latter difference was merely marginal, however, and the product of an underpowered analysis (small effect size) it must be viewed with skepticism. Among women, erotophilia was greater in those enrolled under the “pornographic” advertisement version in April, than in those enrolled under the “photographic” advertisement version in September (one-tailed t(89)= −1.75, p=.04, Cohen’s d=0.40 or small). As this analysis was underpowered (small effect size), however, it must be viewed with skepticism.

Table 3.

Mean Values (± SD) on Individual Difference Measures in Order of Administration, by Gender, Execution, and Advertisement Version.

  Men
 Women
  April
 September
 April
 September
  “Pornography” “Photography” “Pornography” “Photography” “Pornography” “Photography” “Pornography” “Photography”
SOS 81.20a (17.41) 79.35a (18.90) 78.75a (17.66) 87.29a (15.83) 70.33b (16.59) 65.81b,c (15.26) 66.35b,c (23.11) 62.69c (20.95)
PUF 4.04d (1.11) 3.85d (0.98) 3.67d (1.13) 3.88d (0.60) 2.03e (0.62) 1.95e (0.91) 2.06e (0.94) 1.80e (0.91)
OtE 3.74g,h,i (0.57) 3.61f,g (0.62) 3.51f (0.39) 3.62f,g,h (0.53) 3.91i (0.47) 3.59f,g (0.53) 3.81h,i (0.55) 3.72g,h,i (0.51)
Open in a new tab

Notes. SOS: erotophilia (Fisher et al., 1998), PUF: pornography use frequency, OtE: openness to experience (DeYoung et al., 2007) for all scales greater scores indicate greater such frequency/possession of the trait. In April men’s n’s 44–45 for ‘Pornography’ and 40 for “Photography”: women's 30 for “Pornography”, and 37 for “Photography”. In September men’s n = 24 for “Pornography” and 17 for “Photography”: women’s 53–54 for “Pornography” and 60–61 for “Photography”. Where there is not at least one subscript in common between values within a measure, they differ at the p<.10 or marginally significant level (see text for exact p-, t-, and Cohen’s d-values associated with significantly different values). All t-tests two-tailed in this table, for consistency within the table: see text for all significant one-tailed t-test results where predictions concerning individual difference measures (all of which were directional) were made.

Within males, pornography use frequency did not differ between advertisement version enrolled under, in either execution. The same was true for erotophilia (except marginally) in the September execution only, where it was greater in those having enrolled under the “photographic” advertisement version (Table 3 for means and SD’s: one-tailed t(37)= −1.62, p=.06, Cohen’s d = 0.51 – a moderate effect size). Finally, within males, Openness to Experience did not differ between advertisement version enrolled under. Thus, males enrolling under the “photographic” study advertisement version did not differ from males enrolling under the “pornographic” version, on any of the assessed variables.

Discussion

The proportion of participants who were women, enrolling under the “photographic” advertisement version, was not significantly greater or lesser than the proportion of participants who were women enrolling under the “pornographic” advertisement version of the study. This held true in both executions. Crucially, women enrolling under the “pornography” advertisement did not differ on average in their erotophilia or pornography use from women enrolling under the “photography” advertisement, in either study execution. These findings suggest that study advertisements that describe participants’ exposure to pornography may not over-sample women who are higher in porn use frequency or erotophilia. One implication, is that researchers using recruitment materials containing the word “pornographic” or similar, might have some confidence women are not self-selecting out of their studies, and that women participants in their studies are not atypical for their gender.

Women having enrolled under the “pornographic” advertisement were higher in Openness to Experience, albeit only in one of two executions. This adequately-powered result, seen in only one execution, must be regarded with caution: Not only was it not replicated within the same university’s participant pool, but given the number of comparisons we ran in order to test the hypotheses (20) at alpha = 0.05, it must be regarded as potentially the product of Type I error.

Studies’ samples’ (dis)similarity to other convenience samples

In order to broadly estimate how similar or dissimilar this study’s samples were in terms of the three, individual difference measures assessed, it would be helpful to show apparent absence of gross differences between (1) typical scores in each found elsewhere and (2) our participants’ scores. Men’s erotophilia scores in this study were higher than women’s in each execution. This has been previously reported in essentially every study of erotophobia/erotophilia known to the authors. Means on erotophilia of university convenience samples for studies that did not explicitly involve exposure to pornography in recruitment materials (from Canada, the U.S., Israel, Hong Kong, and India: recruited for “sex research” in Canadian and US samples, only), ranged from 59.17 to 78.22 for males and from 47.89 to 71.81 for females. In these studies, moreover, there was a significant difference between each male and (paired) female score (Fisher et al., 1988). In this study, the range of women’s erotophilia mean scores, broken down by execution and study advertisement version (i.e. 62.69–70.33), were within the just-noted female range. All of those of men, however, were higher (i.e. 78.75–87.29) than the upper end of the just-noted male range. It should be noted, however, that it is not necessarily true that sampled men were atypically high in erotophilia. That is so, as this higher, average score among men may reflect, for instance, an average increase in erotophilia in that gender, over the past 30 years.

Porn use frequency has also been shown to be greater in men, across a number of studies. This was the case in this study. There is no previously published range of typical values for men versus women, however, on the porn use frequency measure used herein. Thus, although consistently greater such values for men versus women in this study might be viewed as encouraging as regards our samples’ representativeness on this variable, whether our samples were indeed representative cannot be known. Means of North American university convenience samples in Openness to Experience have previously been reported, though not by gender or sex: Reported mean across two samples (N = 480 and N = 90) was M = 3.52, SD = 0.64 (DeYoung et al., 2007). In this study, note that the range of such means, broken down by execution and advertisement version enrolled under, were 3.59 − 3.91 for women and 3.50 − 3.73 for men.

Men and women (pooled) enrolling under the “pornographic” advertisement were not significantly higher in porn use frequency or erotophilia, than were men and women (pooled) enrolling under the “photographic” advertisement. And crucially, women enrolling under the “pornographic” advertisement did not significantly differ in this measure from women enrolling under the “photographic” advertisement, in either study execution. Neither was this the case for men. Taking these results together, this article provides some evidence that studies the advertisements for which include the word “pornographic”, may not tend to over-sample either women or men higher in porn use frequency or erotophilia, compared to those of the same gender in their population. Given our hypotheses concerned women only, however, we limit our conclusions to that gender.

Self-selection by women out of surveys concerning sexuality unrelated items of “personal concern”

Women’s self-selection out of surveys might occur outside research on pornography or other sexuality-related topics, where items of average lesser “personal concern” to women are queried. Thus, the survey literature citing quantitatively assessed sex/gender difference or similarity in responding, where the instrument included items of “personal concern” not regarding sexuality, and in which self-selection was also quantitatively assessed, was reviewed. These surveys ranged in focus from particular study-types’ participants’ personality (Mahoney & Kyle, 1976), learning (Daugherty & Russo, 2011), judgments relating to risk (Tyler & Lomax Cook, 1984), health monitoring (Gershenson Hodgson & Cutler, 2003), and psychological distress (Sweeting et al., 2010, p. 1825), to politics (Hustinx et al., 2012), and water usage (Mankad & Gardner, 2016) – some items the interest in which may differ on average by sex/gender (e.g. politics: see Hayes & Bean’s 1993 international review). Among these surveys regarding a range of “personal concern” items unrelated to sexuality but potentially differing in average interest by sex/gender, differences by sex/gender were sometimes found in responding, whether or not enrollment self-selection on this basis was detected.

Possible month of year effect

The study’s initial execution occurred during the end of the winter semester in the final two weeks in which the participant pool was open. Given that students had to complete study participation so soon, choice of study was somewhat constrained, potentially affecting choice to enroll in this study. There was lesser such constraint at the beginning of the next term, when the second execution of the study was performed. Given that we analyzed the data for each execution separately, this would not appear to generate an issue with respect to our conclusions.

Strengths and limitations

It is conceivable that pornography researchers might successfully argue to their Internal Review Boards that mention of “pornography” in recruitment materials might cause different data to be gathered, and that the integrity of their studies would thus be adversely affected by inclusion of the word. In that case, recruitment materials might not be made to contain such a word. Then, self-selection out of studies based on recruitment material stating there would be pornographic content, would not exist as an issue. In the country in which this study occurred however (and in many others), there are ethical requirements of informed and voluntary consent (Government of Canada, TCPS 2, 2019), rendering these issues live. Informed consent, which must occur prior to any participation (per Introduction to Chapter 3) requires disclosure of “… all information necessary for making an informed decision to participate …” (from Article 3.2). “Voluntariness” is said to be important because “… it respects human dignity and means that individuals have chosen to participate in research according to their own values, preferences and wishes …” (from Article 3.1). To the extent values are related to pornography use, it may therefore be difficult to argue participants need not be warned of pornography, as part of an IRB-approved study. In any case, where past such research’s recruitment materials and/or recruitment materials in countries which do mandate such warnings contain(ed) them, it may now be argued their data are not incomparable on this ground to that collected without such warnings. A reviewer noted another work-around to the potential problems of self-selection out by gender/atypicality of women recruited for such studies: the use instead of the phrase “sexually explicit” or similar instead of “pornography” in recruitment materials. Though using “sexually explicit” in recruitment materials was debated among the authors, “pornographic” was chosen. This was because the former’s literal meaning – explaining of sex – might indicate materials produced for, for example, teaching of sexuality/sexual health (which is done within the department offering the study) rather than erotica/pornography.

Limited demand characteristics are the primary strength of the current design. This is necessarily the case in minimal difference studies. It is also a function, here, of conducting the study over the internet (anonymously). That the study has limited demand characteristics, however, means a small, rather than a moderate, effect size, only, should perhaps have been anticipated.

The fact only 54 out of 82 September “pornographic” enrollment participants when asked what the study concerned, cited “porn” or similar, might signal that some participants were attempting to be discrete about a controversial topic, and were therefore nonspecific when answering the given question. It could instead, however, signal ineffectiveness of the manipulation for up to 34% of participants. Even if this were the case, it would suggest that participants do not self-select out of studies based on any recruitment materials. This may be because they do not (thoroughly) read them. It may especially be the case for students 10% of a course grade for whom essentially requires such participation, and/or who may justifiably believe their university would not make available harmful studies. To the extent that occurred in our study, we do not see how it would not generally occur with other studies requiring prior, written consent. As such, this possibility, though a significant limitation may be a limitation of all such designs. For this reason, we hold that this issue does not necessarily negatively impact any conclusion that men and women are equally likely to enroll in the two study types, though whether they would remain in such a study once pornographic content was encountered remains an open question. Note that though analyzing data only from participants who had not failed the manipulation check was considered, this was not done in part because our analyses – already perhaps underpowered – would have been rendered more underpowered. Additionally, relatively non-conscientious participants might have been selectively excluded, and there exists evidence of an average sex difference on that trait (e.g. cross-culturally, and particularly in North America: Schmitt et al., 2008).

All participants comprising first-year university students, means one would not anticipate the study’s results to apply in other populations. Thus, replicating this design in the geographically proximate, general population – which would seem feasible given it was conducted entirely online – is suggested as future work. Note that people who agree to participate in any sexuality research may logically differ from those who will not, especially where these need not participate in any study at all (cf. students, such as this study’s participants, for whom 10% of a course mark might depend on such participation). This means the results from this study, may be even less similar to those expectable from an otherwise similar, non-sexual study run on the surrounding, general population. Finally, the similarity between the genders in the within study in enrolling in the two study versions (“pornographic” versus “photographic”), might not be seen outside North America. This might not occur, for instance, where relevant gender expectations tend to overlap less. It is suggested the within study be repeated (again online) cross-culturally, in countries currently more greatly resembling in average mores concerning pornography North America of 30 years ago, than North America of the present day. That way, two research ends might be met. First, data would be gathered, useful in its own right, concerning these other countries’ pornography research participants and their (dis)similarity to those in the within study. Second, clues might be found as to whether cultural change may be explanatory of apparent differences between the current result of gender not impacting sign-up for a study including the word “pornographic” in its advertisement, and differences between North American men and women in enrolling in sexuality studies 30 years ago. Given there is evidence of differences in size of average sex differences in other personality variables cross-culturally (see, e.g. Schmitt, 2015) and in wish for sexual variety (Schmitt et al., 2003), these may exist specifically in regards to pornography, and may affect enrollment by sex or gender in studies explicitly including it.

It is possible that online (and therefore anonymous) recruitment prevented self-selection out of the study by gender, or answering of the given individual difference measures based (in part) on how members of one’s gender may be expected to behave, think, or believe. That is, anonymous participation may have reduced or eliminated any concerns participants had about presenting as socially desirable members of their gender. In contrast, previous research has not recruited anonymously. Self-selection out of those studies by women, which did not occur in this study, may be thusly explicable.

This study cannot speak to the influence of gender on participant behavior within research involving pornography. It cannot be determined, for example, whether women who sign up for studies advertised to involve pornography are less likely than men participants to actually perform the study tasks associated with seeing/using pornography. That is, since no pornography (or any image at all) was actually shown to participants, any greater drop-out or nonperformance of tasks associated with viewing those images by women, could not be assessed. It therefore remains possible that women may indeed self-select out of providing analyzable data, if the study actually involves exposure to pornography. Additionally, even if women do not preferentially drop out of studies advertised as involving “porn”, once pornography is actually displayed, some might respond to the three tested individual difference measures differently, after having been exposed to pornography within the study because of priming effects. Note, however, that Saunders et al. (1985) analyzed all personality and demographic variables twice: once using the data of all their participants (who chose to enroll in either their pornography viewing or personality study), and once using only the data of those who agreed, after enrolling for one of these studies, to then also participate in the other. These authors found no difference in these variables between these two groups. Note these authors did find, however, greater erotophilia among the women who initially chose to enroll for the study involving viewing pornography, than among the women who initially chose to enroll for their other study.

Importantly, this study was not sufficiently powered to detect small effect sizes. While this study does not support the notion of moderate to large self-selection effects when pornography exposure is mentioned in a study advertisement, the number of non-significant but marginal (small effect size) “findings” in this study, suggests that future research in this area should sample enough participants to detect small effect sizes, at least for similar designs. Additionally, note the six regressions run post-hoc to test whether gender interacted with survey advertisement version enrolled under, to predict each of the three, individual difference DVs, were also underpowered: they would have been better powered had enough participants been recruited to enable detection of small effects.

Possible future directions

This research might not only be directly replicated on a general population sample: It could also be conceptually replicated, thereby making it relevant to research on willingness to participate in surveys from other disciplines. For example, as noted, women might participate at a lower rate than they otherwise would, all else equal, when their “personal concern” regarding the survey’s topic, in which they may or may not possess personal interest, will be assessed. If this were true, then if our study were re-run in the same population but instead of asking for participants willing to give their views regarding images, we asked for participants willing to express what they saw to be (for instance) normative views regarding images, we should see an even greater proportion of participants who were women enrolling in the “pornographic” condition.

A change in norms as concern women’s pornography use since the last time possible self-selection out of such studies by women was assessed is possible. It may be why we found no such self-selection. Unfortunately, we are not aware of any study of specifically such norm change over the period in question. We therefore suggest a longitudinal study be run separately to any replication of our study: Such a study would simply survey those of the same demographic we tested, as to whether they endorse statements, such as “Women should not use pornography” less than analogous statements about men. (It would be longitudinal, as it would be run more than once, with enough years intervening to allow any further gender norm change to occur.) This would comprise some, albeit very limited, evidence that these norms, specifically as regard pornography, can change over time.

Conclusion

Some evidence was obtained that studies advertising using the word “pornographic” to North American university convenience samples may neither produce self-selection out by women, nor over-sample gender-atypical women in the presumably relevant variables of pornography usage frequency, erotophilia, or Openness to Experience. Thus, researchers have some basis, in similar samples, for confidence that studies including descriptions of pornography in recruitment materials are not invalid on the basis of self-selection out by women nor of over-sampling gender-atypical women. It must be noted, however, that these null findings might be explained by inadequate sample size to detect small effects. Comparing the current findings to older ones suggests that level of willingness among North American young women university students to sign up for studies involving viewing “pornographic” images may be changeable over time. Such willingness should therefore be periodically reexamined. Where recruitment materials of other research concerning pornography contain(ed) the word “pornographic”, it may now be argued their data are not incomparable on this ground to that collected without such word in those materials.

Acknowledgments

The authors would like to acknowledge and thank Research Assistant, Jodie Baer.

Notes

1

Based on Bem Sex-Role Inventory score (Bem, 1974).

2

Manuscript submission delayed to 2020, for reasons unrelated to study.

3

During the span of the April execution, 58 studies comprising a total of 963 time-slots available to be chosen by qualifying students were available on SONA.

4

Note that because the proportions, only, of women versus men enrolling by condition was sought, we did not seek overall student proportion of women, which information, in any case, to our best knowledge is unpublished.

5

Additional such measures were employed to test hypotheses beyond the scope of this article. Results concerning these are available upon request from the first author.

6

One-tailed testing was performed when comparing men and women, together, on individual difference measures between advertisement version enrolled under. This was done, in order to make that testing comparable to testing performed comparing women (and, separately, men) enrolling under each advertisement version (same measures).

7

Details available on request from the first author.

8

All differences between individual difference means, broken down by gender, execution, and advertisement version enrolled under, are noted in this table.

Conflict of interest statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

References

  1. Abelson, H., Cohen, R., Heaton, E., & Suder, C. (1971). National survey of public attitudes toward and experience with erotic materials (Technical report (no. O-424-372) of the commission on obscenity and pornography) (Vol. 6, pp. 1–137). Government Printing Office. [Google Scholar]
  2. Archer, J. (2019). The reality and evolutionary significance of human psychological sex differences. Biological Reviews of the Cambridge Philosophical Society, 94(4), 1381–1415. 10.1111/brv.12507 [DOI] [PubMed] [Google Scholar]
  3. Bártová, K., Androvičová, R., Krejčová, L., Weiss, P., & Klapilová, K. (2020). The prevalence of paraphilic interests in the Czech population: Preference, arousal, the use of pornography, fantasy, and behavior. The Journal of Sex Research, 58:86–96. 10.1080/00224499.2019.1707468 [DOI] [PubMed] [Google Scholar]
  4. Bem, S. L. (1974). The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 42(2), 155–162. [PubMed] [Google Scholar]
  5. Blake, K. R., Bastian, B., Denson, T. F., Grosjean, P., & Brooks, R. C. (2018). Income inequality not gender inequality positively covaries with female sexualization on social media. Proceedings of the National Academy of Sciences, 115(35), 8722–8727. 10.1073/pnas.1717959115 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bőthe, B., Vaillancourt-Morel, M. P., Girouard, A., Štulhofer, A., Dion, J., & Bergeron, S. (2020). A large-scale comparison of Canadian sexual/gender minority and heterosexual, cisgender adolescents’ pornography use characteristics. The Journal of Sexual Medicine, 17(6), 1156–1167. 10.1016/j.jsxm.2020.02.009. [DOI] [PubMed] [Google Scholar]
  7. Bradley, G. V. (2018). Prolegomenon on pornography. Harvard Journal of Law and Public Policy, 41(2), 447–498. [Google Scholar]
  8. Bridges, A. J., & Morokoff, P. J. (2011). Sexual media use and relational satisfaction in heterosexual couples. Personal Relationships, 18(4), 562–585. 10.1111/j.1475-6811.2010.01328.x] [DOI] [Google Scholar]
  9. Carroll, J. S., Padilla-Walker, L. M., Nelson, L. J., Olson, C. D., McNamara, B. C., & Madsen, S. D. (2008). Generation XXX: Pornography use and acceptance among emerging adults. Journal of Adolescent Research, 23(1), 6–30. 10.1177/0743558407306348 [DOI] [Google Scholar]
  10. Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155–159. 10.1037/0033-2909.112.1.155. [DOI] [PubMed] [Google Scholar]
  11. Cooper, A. (1998). Sexuality and the Internet: Surfing into the new millennium. CyberPsychology & Behavior, 1(2), 187–193. [Google Scholar]
  12. Daneback, K., Traeen, B., & Månsson, S. (2009). Use of pornography in a random sample of Norwegian heterosexual couples. Archives of Sexual Behavior, 38(5), 746–753. 10.1007/s10508-008-9314-4. [DOI] [PubMed] [Google Scholar]
  13. Daugherty, A. L., & Russo, M. F. (2011). An assessment of the lasting effects of a stand-alone information literacy course: The students’ perspective. The Journal of Academic Librarianship, 37(4), 319–326. 10.1016/j.acalib.2011.04.006. [DOI] [Google Scholar]
  14. DeYoung, C. G., Quilty, L. C., & Peterson, J. B. (2007). Between facets and domains: 10 aspects of the Big Five. Journal of Personality and Social Psychology, 93(5), 880–896. 10.1037/0022-3514.93.5.880. [DOI] [PubMed] [Google Scholar]
  15. Ferguson, C. J. (2009). An effect size primer: A guide for clinicians and researchers. Professional Psychology: Research and Practice, 40(5), 532–538. 10.1037/a0015808. [DOI] [Google Scholar]
  16. Fisher, W. A. (1983). Gender, gender role identification, and response to erotica. In Allgeier E. R. & McCormick N. B. (Eds.), Gender roles and sexual behavior: Changing boundaries (pp. 261–284). Mayfield. [Google Scholar]
  17. Fisher, W. A., & Barak, A. (2001). Internet pornography: A social psychological perspective on Internet sexuality. Journal of Sex Research, 38(4), 312–323. [Google Scholar]
  18. Fisher, W. A., & Byrne, D. (1978a). Sex differences in response to erotica? Love versus lust. Journal of Personality and Social Psychology, 36(2), 117–125. 10.1037/0022-3514.36.2.117. [DOI] [Google Scholar]
  19. Fisher, W. A., & Byrne, D. (1978b). Individual differences in affective, evaluative, and behavioral responses to an erotic film. Journal of Applied Social Psychology, 8(4), 355–365. 10.1111/j.1559-1816.1978.tb00789.x. [DOI] [Google Scholar]
  20. Fisher, W. A., Byrne, D., White, L. A., & Kelley, K. (1988). Erotophobia-erotophilia as a dimension of personality. Journal of Sex Research, 25(1), 123–151. 10.1080/00224498809551448. [DOI] [Google Scholar]
  21. Gershenson Hodgson, L., & Cutler, S. J. (2003). Looking for signs of Alzheimer’s Disease. The International Journal of Aging and Human Development, 56(4), 323–343. 10.2190/E6J1-PUX7-J43C-090B. [DOI] [PubMed] [Google Scholar]
  22. Government of Canada. (2019). Tri-council policy statement: Ethical conduct for research involving humans – TCPS 2 (2018). Her Majesty the Queen in Right of Canada. [Google Scholar]
  23. Hayes, B. C., & Bean, C. S. (1993). Gender and local political interest: Some international comparisons. Political Studies, 41(4), 672–682. 10.1111/j.1467-9248.1993.tb01665.x. [DOI] [Google Scholar]
  24. Hald, G. M. (2006). Gender differences in pornography consumption among young heterosexual Danish adults. Archives of Sexual Behavior, 35(5), 577–585. 10.1007/s10508-006-9064-0. [DOI] [PubMed] [Google Scholar]
  25. Hustinx, L., Meijs, L. C. P. M., Handy, F., & Cnaan, R. A. (2012). Monitorial citizens or civic omnivores? Repertoires of civic participation among university students. Youth & Society, 44(1), 95–117. 10.1177/0044118X10396639. [DOI] [Google Scholar]
  26. Kenrick, D. T., Stringfield, D. O., Wagenhals, W. L., Dahl, R. H., & Ransdell, H. J. (1980). Sex differences, androgyny, and approach responses to erotica: A new variation on the old volunteer problem. Journal of Personality and Social Psychology, 38(3), 517–524. [Google Scholar]
  27. Kinsey, A. C., Pomeroy, W. B., Martin, C. E., & Gebhard, P. H. (1953). Sexual behavior in the human female. Indiana University Press. [Google Scholar]
  28. Maddox, A. M., Rhoades, G. K., & Markman, H. J. (2011). Viewing sexually-explicit materials alone or together: Associations with relationship quality. Archives of Sexual Behavior, 40(2), 441–448. 10.1007/s10508-009-9585-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Mahoney, J., & Kyle, D. (1976). Personality characteristics of volunteers for thanatological research. OMEGA-Journal of Death and Dying, 7(1), 51–57. [Google Scholar]
  30. Mankad, A., & Gardner, J. (2016). The role of personal experience in frequency of rainwater tank maintenance and policy implications. Local Environment, 21(3), 330–343. 10.1080/13549839.2014.959907. [DOI] [Google Scholar]
  31. Mosher, D. L., & MacIan, P. (1994). College men and women respond to X-rated videos intended for male or female audiences: Gender and sexual scripts. Journal of Sex Research, 31(2), 99–113. 10.1080/00224499409551736. [DOI] [Google Scholar]
  32. O’Reilly, S., Knox, D., & Zusman, M. E. (2007). College student attitudes toward pornography use. College Student Journal, 41(2), 402–406. [Google Scholar]
  33. Paul, B., & Shim, J. W. (2008). Gender, sexual affect, and motivations for internet pornography use. International Journal of Sexual Health, 20(3), 187–199. 10.1080/19317610802240154. [DOI] [Google Scholar]
  34. Petersen, J. L., & Hyde, J. S. (2010). A meta-analytic review of research on gender differences in sexuality, 1993–2007. Psychological Bulletin, 136(1), 21–38. 10.1037/a0017504. [DOI] [PubMed] [Google Scholar]
  35. Saunders, D. M., Fisher, W. A., Hewitt, E. C., & Clayton, J. P. (1985). A method for empirically assessing volunteer selection effects: Recruitment procedures and responses to erotica. Journal of Personality and Social Psychology, 49(6), 1703–1712. 10.1037/0022-3514.49.6.1703. [DOI] [Google Scholar]
  36. Schmitt, D. P., Alcalay, L., Allik, J., Ault, L., Austers, I., Bennett, K. L., Bianchi, G., Boholst, F., Cunen, M. A. B., Braeckman, J., Brainerd, E. G., Caral, L. G. A., Caron, G., Casullo, M. M., Cunningham, M., Daibo, I., De Backer, C., De Souza, E., Diaz-Loving, R., … Zupanèiè, A. (2003). Universal sex differences in the desire for sexual variety: Tests from 52 nations, 6 continents, and 13 islands. Journal of Personality and Social Psychology, 85(1), 85–104. 10.1037/0022-3514.85.1.85. [DOI] [PubMed] [Google Scholar]
  37. Schmitt, D. P., Realo, A., Voracek, M., & Allik, J. (2008). Why can’t a man be more like a woman? Sex differences in Big Five personality traits across 55 cultures. Journal of Personality and Social Psychology, 94(1), 168–182. 10.1037/0022-3514.94.1.168. [DOI] [PubMed] [Google Scholar]
  38. Schmitt, D. P. (2015). The evolution of culturally-variable sex differences: Men and women are not always different, but when they are… it appears not to result from patriarchy or sex role socialization. The evolution of sexuality (pp. 221–256). Springer. [Google Scholar]
  39. Kohut, T. A. (2014). An empirical investigation of the concept of “Pornography [Doctoral dissertation, University of Western Ontario]. University of Western Ontario Electronic Thesis and Dissertation Repository. [Google Scholar]
  40. Sona Systems Ltd. (1997. –2016). Western department of psychology: Psychology research participation pool. https://uwo-psychology.sona-systems.com/Default.aspx?ReturnUrl=%2f. [accessed 22 January 2016]
  41. Sweeting, H., West, P., Young, R., & Der, G. (2010). Can we explain increases in young people’s psychological distress over time? Social Science & Medicine, 71(10), 1819–1830. 10.1016/j.socscimed.2010.08.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Tyler, T. R., & Lomax Cook, F. (1984). The mass media and judgments of risk: Distinguishing impact on personal and societal level judgments. Attitudes and Social Cognition, 47(4), 693–708. [Google Scholar]
  43. Wilson, W. C., & Abelson, H. I. (1973). Experience with and attitudes toward explicit sexual materials. Journal of Social Issues, 29(3), 19–39. 10.1111/j.1540-4560.1973.tb00086.x. [DOI] [Google Scholar]
  44. Wright, P. J., Bae, S., & Funk, M. (2013). United States women and pornography through four decades: Exposure, attitudes, behaviors, individual differences. Archives of Sexual Behavior, 42(7), 1131–1144. 10.1007/s10508-013-0116-y. [DOI] [PubMed] [Google Scholar]
  45. Wolchik, S. A., Braver, S. L., & Jensen, K. (1985). Volunteer bias in erotica research: Effects of intrusiveness of measure and sexual background. Archives of Sexual Behavior, 14(2), 93–107. [DOI] [PubMed] [Google Scholar]

Articles from International Journal of Sexual Health are provided here courtesy of Taylor & Francis

RESOURCES