Preliminary Prevalence Estimates of Sexual Exploitation as Measured by the Sexual Experiences Survey-Victimization (SES-V) in a National US Sample

ABSTRACT

The new Sexual Experiences Survey-Victimization (SES-V) was designed to capture a larger range of sexual exploitation and to be applicable to more diverse populations than prior measures. This study represents the first administration of the SES-V in a national sample of adults (N = 347). Participants were recruited from a crowdsourcing platform and selected to reflect the national distribution in terms of age, sex, and race/ethnicity. Rates of sexual exploitation since age 14 were very high in this sample (90%), providing evidence that the SES-V was successful in capturing experiences on the low end of the severity continuum. Using the items corresponding to the FBI definition of rape, 60% of women and 29% of men endorsed rape on the SES-V. Compared to men, women reported higher rates of sexual exploitation overall, and higher rates of every type of sexual exploitation except technology-facilitated. The new SES-V also asks participants to estimate the number of separate instances of four types of sexual exploitation that they have experienced; results provided support for the value of these incident estimates in understanding the scope of sexual exploitation. Finally, this study evaluated new sexual acts and exploitative tactics that were added to the SES-V and found that they demonstrated utility and validity. These findings offer preliminary support for the validity and functionality of the SES-V, although the high prevalence of sexual exploitation on the SES-V is discussed as both a strength and limitation. Future research should evaluate prevalence and demographic differences in a larger national sample.

The prototypical rape victim, as depicted in the news media, entertainment, legislation, and academic scholarship, is a heterosexual, White, cisgender, college woman, who is exploited by a male perpetrator (Davis, 2023; Onwuachi-Willig, 2018; Peterson, 2018). Indeed, college women’s rates of sexual exploitation victimization are alarmingly high, with an estimated one in four to one in five women experiencing nonconsensual sexual contact just during their time in college (Muehlenhard et al., 2017). Given this, it is unsurprising that much of the research on the measurement of sexual exploitation (i.e., nonconsensual sexual acts with and without bodily contact) has focused on college women’s victimization.

Groups Vulnerable to Sexual Exploitation

Focusing exclusively on heterosexual, White, college women, however, ignores other highly vulnerable groups. First, college students are vulnerable by nature of their age—adolescents and young adults experience disproportionately high rates of victimization compared to older adults—not by nature of their status as college students (Muehlenhard et al., 2017); thus, young adults who are not enrolled in college experience similar vulnerability to those that are (Mumford et al., 2020). Further, studies have found that trans and nonbinary individuals experience similar or greater vulnerability to sexual exploitation as compared to cisgender women (Coulter et al., 2017; James et al., 2016; Johnson et al., 2016; Stotzer, 2009). Additionally, bisexual women are more vulnerable to sexual exploitation than heterosexual women, and bisexual and gay men have similar vulnerability to heterosexual women (Chen et al., 2020; Coulter et al., 2017; Ford & Soto-Marquez, 2016; Johnson et al., 2016). Other vulnerability factors for sexual assault victimization include having a physical or cognitive disability (Basile et al., 2016; Bonomi et al., 2018; Holloway et al., 2022), and having a Black, Multiracial, or Indigenous racial identity (Basile et al., 2022). Finally, although cisgender heterosexual men experience sexual exploitation at lower rates than many other gender and sexual identity groups, they do experience sexual exploitation perpetrated by both men and women (and likely by individuals of other genders, as well; Ybarra et al., 2022), and the negative consequences of their exploitation are similar to those experienced by women (Peterson et al., 2011; Smith et al., 2022). All of these studies provide evidence that measures of sexual exploitation need to be appropriate for use in diverse samples beyond heterosexual, White, college women.

The Sexual Experiences Survey-Victimization (SES-V) measure, which is the newest revision of the most widely used measure of sexual victimization, is designed to capture a larger range of sexual exploitation and to better elicit experiences of sexual exploitation among understudied groups than prior measures of sexual victimization (Koss et al., 2024). Consistent with this prediction, Anderson, Peterson, Koss, et al. (2024) demonstrated in a sample of college students that a preliminary version of the SES-V captured more instances of sexual exploitation than the prior version of the SES (Koss et al., 2007), including for example, among men and LGBTQ+ participants.

Important Changes Made in the SES-V

The SES-V revision team chose to refer to the construct measured by the SES-V as sexual exploitation victimization (see also Koss et al., 2024). “Exploitation” refers to using someone unfairly to your own advantage (Cambridge Dictionary, 2022), which seemed to encompass the large range of sexual victimization experiences measured by the SES-V. Further, as described in Koss et al. (2024), the term sexual exploitation avoids some of the murkiness associated with other common phrases such as “sexual assault” or “sexual coercion,” which are inconsistently and confusingly used as both umbrella terms for all sexual victimization and as subcategories of sexual victimization. Also, unlike the term “sexual violence,” the term sexual exploitation does not risk promoting stereotypes of sexual victimization as inherently physically violent.

As described in Koss et al. (2024) and other articles in this special issue, the SES-V includes several important changes from prior versions of the measure, many of which are included to ensure greater relevance across a variety of diverse research populations. The SES-V includes four modules assessing a wide range of sexual exploitation: technology-facilitated, non-contact, verbally pressured, and illegal sexual exploitation. Thus, compared to prior versions of the SES (Koss & Oros, 1982; Koss et al.,1987, 2007), the new SES-V particularly expands the assessment of lower-severity experiences of sexual exploitation. In addition, the SES-V includes new sexual acts and exploitative tactics that were not included in prior versions of the measure.

For example, as described in Peterson et al. (2024), the SES-V expands on the sexual acts assessed in prior versions of the SES by including sexual exploitation that resulted in being made to penetrate another person. That is, in contrast to prior versions of the SES in which respondents were only asked about having their own body (vagina or anus) penetrated, the SES-V asked about experiences in which they were made to penetrate the perpetrator’s vagina/genital opening or anus. The addition of these sexual acts is particularly expected to improve the ability of the SES-V to detect sexual exploitation among heterosexual cisgender men (Anderson et al., 2020; Kern et al., 2022; Littleton et al., 2020).

In addition to new sexual acts, the SES-V also includes new tactics of sexual exploitation that were not included in prior versions of the SES. For example, in the Illegal Sexual Exploitation module of the SES-V, new items measure someone starting sex with “no advance notice” so that there is not an opportunity to provide consent; reproductive coercion, including lying about or tampering with contraception or “stealthing” (removing a condom without permission); and commercial sexual exploitation that involves fraud, coercion, or force to compel another person to engage in sex acts for financial compensation (see Peterson et al., 2024 for further discussion). In the Verbally Pressured Sexual Exploitation module of the SES-V, new items assess coercion through the use of substances. Prior versions of the SES have included incapacitated sexual exploitation (i.e., sex when the respondent was too drunk or high to resist) as an illegal act, and that is retained in the SES-V, but the SES-V also includes uses of substances to coerce sex in cases in which the respondent is not incapacitated but in which they would not have agreed to sex if not pressured to consume substances (see Anderson, Peterson, Canan, et al., 2024 for further discussion).

Another important change in the SES-V is that optional follow-up questions are included after each module to provide greater information about the types of contexts in which different forms of sexual exploitation take place (see Koss et al., 2024). One question in these follow-up modules asks participants to estimate the number of occasions in which they experienced each type of sexual exploitation included in that module. The prior version of the SES asked participants to indicate whether particular acts had occurred 0, 1, 2, or 3 or more times, but a total score on the items could not be interpreted as reflecting a meaningful count of incidents because multiple sexual acts and/or exploitative tactics may have co-occurred during the same incident (Swartout et al., 2015), leading to an overestimation of the number of incidents. A summed score could also lead to an underestimation in some cases, given that the frequency scale for each item was truncated to “3 or more.” In some contexts, it may be important to more accurately estimate the number of separate incidents of sexual exploitation that respondents have experienced; for example, it is possible that prevalence estimates may not differ across certain demographic subgroups, but that fact could potentially disguise an important difference in total number of exploitative incidents experienced by members of the subgroups.

Finally, the SES-V includes a specific incident report at the end of the measure (Koss et al., 2024). Researchers often wish to learn more specific details about respondents’ experiences of sexual exploitation but given that many respondents have multiple experiences of exploitation, asking about every incident might be too onerous. The incident report for the SES-V asks participants follow-up questions—including the sexual acts and exploitative tactics that occurred in that situation, the gender of the perpetrator, and their relationship to the perpetrator—about a single sexual exploitation experience.

Within the incident report, participants are also asked about how they label that sexual exploitation experience, including whether they use the label “rape.” This allows researchers to assess rates of unacknowledged rape among participants who have experienced sexual exploitation that meets legal definitions of rape. A large body of research has demonstrated that many women and men who have experienced something that would legally qualify as rape do not label (or acknowledge) their experience as “rape”; this is often referred to as unacknowledged rape (Anderson et al., 2021; Koss & Oros, 1982; Littleton et al., 2006). For example, in a meta-analysis, researchers estimated that about 60% of female rape victims were unacknowledged, meaning that they did not self-label their experience as “rape” (Wilson & Miller, 2016). One study of male rape victims found that 76% of male rape victims were unacknowledged (Artime et al., 2014); another study directly compared male and female rape acknowledgment and found that men were significantly more likely to be unacknowledged victims than women (80% and 51%, respectively; Reed et al., 2020).

The Present Study

In a first attempt at validating the SES-V, Anderson, Peterson, Koss, et al. (2024) provided some preliminary evidence of construct validity in a sample of college students and illustrated the process by which we finalized the wording of the made-to-penetrate items of the SES-V. The current study represents the second step in the process of evaluating the functionality of the new SES-V. Although the SES has been historically used mostly with college students, the SES-V was designed with diverse populations in mind. In this descriptive study, we administered the published version of the measure to a U.S. community sample selected to reflect the national demographic distribution in terms of age, sex, and race/ethnicity. The purposes of the study were to:

1. Provide a preliminary estimate of prevalence and estimated number of instances of sexual exploitation reported on the SES-V;

2. Compare estimated prevalence rates as a function of gender, sexual identity, race, and disability status;

3. Evaluate whether some of the new items added to the SES-V demonstrate utility (i.e., capture unique victims of sexual exploitation who did not endorse other similar items) and validity (i.e., whether endorsement of added items is correlated with endorsement of other items on the SES-V); and

4. Provide descriptive data based on the new incident report that is included as part of the SES-V.

Method

Participants

For this study, adults aged 18 and older who live in the U.S. were recruited from Prolific Academic, a crowdsourced panel created by academics specifically for the purposes of recruiting research participants. In a comparison of three online crowdsourcing research recruitment platforms, Prolific Academic participants performed as well or better than participants on the other platforms in terms of naivete, honesty, attention, and data quality (Peer et al., 2017). Based on our budget for this project, we requested a representative sample of 375 participants, meaning that Prolific Academic aimed to assemble a sample that was nationally representative in terms of age, sex, and race/ethnicity (see Table 1 for a comparison of our sample and the U.S. population in terms of these variables). Prolific sent an invitation to eligible participants with a brief description of the study topic, the amount of payment offered, and the estimated time of completion. The study was described to participants as a study of “sexual experiences” and did not mention that the research focused on sexual exploitation/victimization.

Table 1. Gender, age, and race of the sample (N = 347) compared to adult U.S. population data.

Demographic characteristic	Sample n (%)	Population %
Age categories^a
18 to 24	41 (11.8)	12.1
25 to 44	131 (37.8)	33.7
45 to 54	52 (15.0)	15.8
55 to 64*	82 (23.6)	16.8
65 to 74	37 (10.7)	12.8
75 to 84*	4 (1.2)	6.3
85 and over*	0 (0.0)	2.5
Sex assigned at birth^b
Female	181 (52.2)	50.8
Male	166 (47.8)	49.1
Race^c
Asian/Pacific Islander	20 (5.8)	6.2
Black/African American	38 (11.0)	12.3
Native American/Indigenous	5 (1.4)	0.7
White*	271 (78.1)	61.7
Bi- or multiracial	9 (2.6)	1.7
Missing	4 (1.2)
Hispanic/Latino/a/x ethnicity*^c	18 (5.2)	17.2

^aIn this study, participants were asked to identify their age in an open-ended item; ages were then converted to categories for comparison with Census Bureau (2020) data. For Census data, percentages were calculated based on total number of individuals over age 18.

^bPopulation data based on entire U.S. population from Census Bureau (2020).

^cU.S. adult population race/ethnicity data based on The Kids Count Data Center (2022) estimates for adults over age 18. Racial/ethnic categories are not exactly comparable because the racial categories in the population data exclude those with Hispanic/Latino/a/x ethnicity, whereas in our sample, we measured race and ethnicity separately, so racial categories can include those with Hispanic/Latino/a/x ethnicity.

*Significant difference between sample and population (p < .05).

Measures

Participants completed the published version of the SES-V (Koss et al., 2024), including follow-up questions after each module, the incident report, and the demographics measure as recommended by Koss and colleagues. Specifically, the SES-V includes four separate modules: Technology-Facilitated Sexual Exploitation, Non-Contact Sexual Exploitation, Verbally Pressured Sexual Exploitation, and Illegal Sexual Exploitation (labeled as the “Contact” module in the participant version so as not to bias results). The measure consists of 37¹ items. The items on the Verbally Pressured and Illegal modules are compound items, such that each exploitative tactic is assessed in relation to six different nonconsensual sexual acts that could be obtained using that tactic: (1) non-penetrative sexual contact (both performing and receiving), (2) oral sex (both performing and receiving), (3) vaginal/genital opening penetration of the respondent’s body, (4) being made to penetrate the perpetrator’s vaginal/genital opening, (5) anal penetration of the respondent’s body, and (6) being made to penetrate the perpetrator’s anus. Participants are asked to indicate how often they have experienced each type of sexual exploitation, with 11 response options ranging from 0 to 10 or more.

Following each module, participants are asked to estimate the total number of incidents they have experienced of the type of sexual exploitation described in that module, noting that a single incident could include more than one sexually exploitative tactic and/or nonconsensual sexual act. For example, at the end of the Verbally Pressured module, they were asked, “On how many separate occasions since age 14 have you experienced one of these instances of sexual pressure? (An occasion could include more than one sexual act or type of pressure.).” Follow-ups to the modules also included questions about the sex of the perpetrator(s) and the respondents’ relationship to the perpetrator(s) for all endorsed items on that module.

At the end of the entire SES-V, participants complete an incident report about a single instance of sexual exploitation that they had reported on the SES-V. In this study, they were asked to choose the single incident that they “remember best” and to answer questions about that incident, including which exploitative tactic(s) and sexual act(s) occurred during that incident, the sex of the perpetrator, their relationship to the perpetrator, and how they labeled the experience (including whether they labeled it as “rape”). As described in Koss et al. (2024), researchers might choose different prompts for the incident report (e.g., most recent, most distressing, etc.) depending on their research questions. In the published version of the measure (Koss et al., 2024), participants are prompted to choose the “most serious” incident. For this study, we were interested in which incidents were most salient to participants, and thus we asked about the “best remembered” incident.

Finally, participants completed the demographics measure (see Koss et al., 2024), which includes questions about sex assigned at birth, gender identity, sexual identity, age, race/ethnicity, perceived economic status (Adler et al., 2000), and disability/functional impairments (Centers for Disease Control, 2018). Throughout the questionnaire, attention check questions as well as several blank questions (response options presented in the absence of an actual question) were included to identify careless or random responding.

Procedure

Eligible participants were contacted by Prolific Academic and invited to participate. Participants first read an informed consent statement and then progressed to the SES-V. The median survey completion time (provided by Prolific Academic) was 18.46 minutes. Participants were paid $4.03 for their participation. The methods of this study were approved by the institutional review board at the second author’s institution.

Results

Data Cleaning

A total of 440 individuals started the survey, and 379 (86%) completed it (including attention checks and the demographics presented at the end of the questionnaire). Of the 379, a total of 32 failed one or more of our attention or consistency checks and were eliminated, leaving a final sample of 347. Missing data on individual SES-V items were treated as non-endorsement, which we concluded was the most conservative approach to handle missing data. It was difficult to calculate missing data rates. Some cis men skipped all items asking about vaginal/genital opening penetration. Additionally, the SES-V was administered through Qualtrics, and the response options for each item (0 to 10 or more) were presented in drop-down menus. Some participants seemed to simply skip the item if they had not had that particular experience (rather than choosing 0 in the drop-down menu). In other words, they selected a number only for the acts in which they had one or more experience. These participants then went on to complete follow-up modules fully and consistently (e.g., they said they had experienced a particular item on the technology-facilitated module three times, skipped the other items in the module, and then reported three total incidents of cyber sexual victimization in the follow-up to the module). Given this, treating missing data as non-endorsement (for those that completed the demographics measure and passed attention checks) was the most conservative approach. That is, treating missing data as non-endorsement might lead to an underestimate of the prevalence of sexual exploitation (e.g., if some individuals skipped items because they didn’t feel comfortable disclosing victimization). In contrast, eliminating those with missing data (especially given that some participants skipped items when they had not had the experience) could result in an overestimate of the prevalence. Demographic characteristics for the final sample are presented in Table 2.

Table 2. Demographics of the sample (N = 347).

Demographic characteristic	M (SD)	Range
Age	45.42 (15.52)	18–83
Subjective economic status (on a scale from 1 to 10)	5.94 (1.69)	1–10
Demographic characteristic	n (%)
Sex assigned at birth
Female	181 (52.2)
Male	166 (47.8)
Intersex	0 (0)
Current gender identity
Woman	176 (50.7)
Man	160 (46.1)
Trans man	2 (0.6)
Trans woman	1 (0.3)
Non-binary	6 (1.7)
Genderqueer	1 (0.3)
Missing	1 (0.3)
Sexual Identity
Heterosexual	268 (77.2)
Bi or pansexual	41 (11.8)
Gay or lesbian	22 (6.3)
Queer	4 (1.2)
Asexual	4 (1.2)
Questioning	4 (1.2)
Another identity	4 (1.2)
Race
Asian/Pacific Islander	20 (5.8)
Black/African American	38 (11.0)
Native American/Indigenous	5 (1.4)
White	271 (78.1)
Bi- or multiracial	9 (2.6)
Missing	4 (1.2)
Hispanic/Latino/a/x ethnicity	18 (5.2)
Middle Eastern/North African ethnicity	5 (1.4)
Disability/Functional Impairment (participants could check all that apply)
Hearing difficulty	18 (5.2)
Vision difficulty	5 (1.4)
Concentration/memory/decision-making difficulty	30 (8.6)
Walking/climbing stairs difficulty	14 (4.0)
Dressing/bathing difficulty	4 (1.2)
Errands difficulty	21 (6.1)

Prevalence Estimates, Number of Incidents, and Demographic Correlates

Most participants (89.6%) endorsed at least one act of sexual exploitation on the SES-V. Endorsement rates by item and for each module are presented in Table 3. Notably, there was substantial overlap in endorsement across modules, such that endorsing an experience on any one module was associated with a higher likelihood of endorsing an experience on any other module (ps < .001; φs = .25–.53). For example, among those that had experienced noncontact exploitation, 77.7% had experienced illegal exploitation compared to only 25.0% of those that had never experienced noncontact exploitation, Χ² (df = 1; n = 347) = 82.81, p < .001, φ = .49.

Table 3. Endorsement rates for each module and individual item and total incidents estimated for each module on the Sexual Experience Survey-Victimization (SES-V) as a function of gender.

	Rates of Endorsement
SES-V Module/Item	Total (N = 347)	Women (n = 176)	Men (n = 160)
	n (%)	n (%)	n (%)
Any endorsement on the SES-V*	311 (89.6)	172 (97.7)	129 (80.6)
	n (%)	n (%)	n (%)
Any endorsement on Non-Contact Sexual Exploitation module*	251 (72.3)	153 (86.9)	93 (58.1)
1. Someone made sexual comments about my body or sex life after I asked them to stop.	193 (55.6)	133 (75.6)	55 (34.4)
2. Someone kept staring at me in a sexual way after I asked them to stop.	173 (49.9)	122 (69.3)	48 (30.0)
3. Someone watched me while I was undressing, was nude, or was having sex without first asking if it was okay.	95 (27.4)	52 (29.5)	39 (24.4)
4. Someone exposed the private areas of their body such as the genital area or penis without first asking if it was okay	171 (49.3)	110 (62.5)	57 (35.6)
5. Someone masturbated in front of me without first asking if it was okay.	102 (29.4)	64 (36.4)	35 (21.9)
	M (SD)	M (SD)	M (SD)
Total summed score on the Non-Contact Exploitation module*	9.40 (11.21)	13.41 (11.50)	5.05 (8.74)
Participants’ estimated total number of incidents of non-contact exploitation*	9.41 (21.70)	14.89 (28.45)	3.80 (7.90)
	n (%)	n (%)	n (%)
Any endorsement on Technology-Facilitated Sexual Exploitation module	232 (66.9)	123 (69.9)	101 (63.1)
6. Someone sent me sexual materials electronically (by e-mail, text, or social media) such as pictures, jokes, or stories without first asking if it was okay.	197 (56.8)	103 (58.5)	90 (56.3)
7. Someone posted sexual comments about me online without first asking if it was okay.	57 (16.4)	33 (18.8)	23 (14.4)
8. Someone continued to electronically send me sexual materials after I asked them to stop.	78 (22.5)	43 (24.4)	33 (20.6)
9. Someone threatened to send nude or seminude photos or intimate videos of me to others or to post them online without first asking if it was okay.	52 (15.0)	34 (19.3)	17 (10.6)
10. Someone sent nude or seminude photos or intimate videos of me to others or posted them online without first asking if it was okay.	40 (11.5)	23 (13.1)	16 (10.0)
11. Someone tried to get me to talk about sex online when I did not want to.	150 (43.2)	90 (51.1)	57 (35.6)
12. Someone asked me online for sexual information about myself when I did not want to tell.	136 (39.2)	79 (44.9)	52 (32.5)
13. Someone asked me to do something sexual online when I did not want to.	130 (37.5)	79 (44.9)	47 (29.4)
14. Someone threatened to post online or send others an image or video of an unwanted sexual experience that happened to me.	25 (7.2)	15 (8.5)	10 (6.3)
15. Someone posted online or sent an image or video of an unwanted sexual experience that happened to me.	8 (2.3)	2 (1.1)	5 (3.1)
	M (SD)	M (SD)	M (SD)
Total summed score on the Technology-Facilitated Exploitation module*	11.71 (17.04)	13.76 (8.07)	9.72 (15.92)
Participants’ estimated total number of incidents of technology-facilitated exploitation	10.61 (21.46)	12.48 (22.34)	8.78 (20.72)
	n (%)	n (%)	n (%)
Any endorsement on Illegal Sexual Exploitation module*	219 (63.1)	138 (78.4)	74 (46.3)
16. Someone bought me drinks or gave me strong drinks to make me impaired and unable to give permission or stop them so they could:
a. Have non-penetrative sexual contact	98 (28.2)	67 (38.1)	28 (17.5)
b. Have oral sex	65 (18.7)	37 (21.0)	23 (14.4)
c. Penetrate my vagina/genital opening	58 (16.7)	54 (30.7)	1 (0.6)
d. Make me penetrate their vagina or genital opening	20 (5.8)	5 (2.8)	14 (8.8)
e. Penetrate my anus	25 (7.2)	15 (8.5)	7 (4.4)
f. Make me penetrate their anus	8 (2.3)	0 (0)	7 (4.4)
17. Someone took advantage of me after I drank alcohol by choice and became impaired and unable to give permission or stop them so they could:
a. Have non-penetrative sexual contact	78 (22.5)	52 (29.5)	23 (14.4)
b. Have oral sex	63 (18.2)	41 (23.3)	19 (11.9)
c. Penetrate my vagina/genital opening	59 (17.0)	57 (32.4)	0 (0)
d. Make me penetrate their vagina or genital opening	17 (4.9)	2 (1.1)	14 (8.8)
e. Penetrate my anus	17 (4.9)	10 (5.7)	5 (3.1)
f. Make me penetrate their anus	7 (2.0)	0 (0)	6 (3.8)
18. Someone put drugs into my drink or misled me about the contents of pills, cigarettes, or powders to make me impaired and unable to give permission or stop them so they could:
a. Have non-penetrative sexual contact	25 (7.2)	20 (11.4)	5 (3.1)
b. Have oral sex	17 (4.9)	11 (6.3)	6 (3.8)
c. Penetrate my vagina/genital opening	14 (4.0)	14 (8.0)	0 (0)
d. Make me penetrate their vagina or genital opening	3 (0.9)	1 (0.6)	2 (1.3)
e. Penetrate my anus	2 (0.6)	1 (0.6)	1 (0.6)
f. Make me penetrate their anus	2 (0.6)	0 (0)	2 (1.3)
19. Someone took advantage of me when I took drugs by choice and became impaired and unable to give permission or stop them so they could:
a. Have non-penetrative sexual contact	26 (7.5)	13 (7.4)	12 (7.5)
b. Have oral sex	19 (5.5)	7 (4.0)	11 (6.9)
c. Penetrate my vagina/genital opening	15 (4.3)	14 (8.0)	0 (0)
d. Make me penetrate their vagina or genital opening	8 (2.3)	0 (0)	8 (5.0)
e. Penetrate my anus	7 (2.0)	4 (2.3)	2 (1.3)
f. Make me penetrate their anus	2 (0.6)	0 (0)	2 (1.3)
20. Someone gave no advance notice so I could not agree or object, including times when I was sleeping or semi-awake to:
a. Have non-penetrative sexual contact	91 (26.2)	60 (34.1)	27 (16.9)
b. Have oral sex	14 (13.8)	20 (11.4)	25 (15.6)
c. Penetrate my vagina/genital opening	56 (16.1)	54 (30.7)	1 (0.6)
d. Make me penetrate their vagina or genital opening	19 (5.5)	1 (0.6)	16 (10.0)
e. Penetrate my anus	14 (4.0)	8 (4.5)	4 (2.5)
f. Make me penetrate their anus	5 (1.4)	0 (0)	5 (3.1)
21. Someone threatened to physically harm me, someone close to me, or my pets to:
a. Have non-penetrative sexual contact	27 (7.8)	22 (12.5)	3 (1.9)
b. Have oral sex	17 (4.9)	10 (5.7)	4 (2.5)
c. Penetrate my vagina/genital opening	23 (6.6)	20 (11.4)	2 (1.3)
d. Make me penetrate their vagina or genital opening	6 (1.7)	2 (1.1)	3 (1.9)
e. Penetrate my anus	6 (1.7)	1 (0.6)	3 (1.9)
f. Make me penetrate their anus	2 (0.6)	0 (0)	1 (0.6)
22. Someone used any level of physical force such as holding me down with their body weight, pinning my arms, or other types of physical force to:
a. Have non-penetrative sexual contact	60 (17.3)	51 (29.0)	6 (3.8)
b. Have oral sex	32 (9.2)	24 (13.6)	4 (2.5)
c. Penetrate my vagina/genital opening	55 (15.9)	52 (29.5)	1 (0.6)
d. Make me penetrate their vagina or genital opening	7 (2.0)	2 (1.1)	4 (2.5)
e. Penetrate my anus	15 (4.3)	7 (4.0)	5 (3.1)
f. Make me penetrate their anus	2 (0.6)	0 (0)	1 (0.6)
23. Someone withheld information about their sexual infection status, lied about using birth control, or tampered with or removed a condom after agreeing to use one to:
a. Have non-penetrative sexual contact	17 (4.9)	9 (5.1)	7 (4.4)
b. Have oral sex	13 (3.7)	4 (2.3)	7 (4.4)
c. Penetrate my vagina/genital opening	31 (8.9)	29 (16.5)	0 (0)
d. Make me penetrate their vagina or genital opening	23 (6.6)	1 (0.6)	21 (13.1)
e. Penetrate my anus	5 (1.4)	2 (1.1)	2 (1.3)
f. Make me penetrate their anus	3 (0.9)	0 (0)	2 (1.3)
24. People in positions of authority could be bosses/supervisors, health care providers, counselors, teachers, coaches, religious leaders, youth group leaders, or criminal/legal personnel. Someone misused their position of authority over me to:
a. Have non-penetrative sexual contact	43 (12.4)	33 (18.8)	7 (4.4)
b. Have oral sex	11 (3.2)	9 (5.1)	1 (0.6)
c. Penetrate my vagina/genital opening	10 (2.9)	9 (5.1)	1 (0.6)
d. Make me penetrate their vagina or genital opening	3 (0.9)	1 (0.6)	1 (0.6)
e. Penetrate my anus	2 (0.6)	0 (0)	1 (0.6)
f. Make me penetrate their anus	1 (0.3)	0 (0)	1 (0.6)
^†25. Someone made me engage in sexual acts for money by using threats to limit my ability to meet life needs such as housing, misleading me into doing a job that involved sexual acts, or physically forcing me to:
a. Have non-penetrative sexual contact
b. Have oral sex
c. Penetrate my vagina/genital opening
d. Make me penetrate their vagina or genital opening
e. Penetrate my anus
f. Make me penetrate their anus
26. Someone used any of these behaviors such as abusing their authority, threatening or using bodily harm, or taking advantage of my incapacitation from drugs or alcohol, to have any of the following sex acts, but the attempted act did not happen. They tried to:
a. Have non-penetrative sexual contact	47 (13.5)	35 (19.9)	11 (6.9)
b. Have oral sex	23 (6.6)	14 (8.0)	9 (5.6)
c. Penetrate my vagina/genital opening	28 (8.1)	24 (13.6)	3 (1.9)
d. Make me penetrate their vagina or genital opening	7 (2.0)	1 (0.6)	6 (3.8)
e. Penetrate my anus	5 (1.4)	3 (1.7)	2 (1.3)
f. Make me penetrate their anus	3 (0.9)	0 (0)	3 (1.9)
	M (SD)	M (SD)	M (SD)
Total summed score on the Illegal Exploitation module*	12.97 (25.65)	15.66 (25.84)	8.88 (22.17)
Participants’ estimated total number of incidents of illegal acts of exploitation*	6.31 (15.28)	8.28 (17.36)	3.92 (12.17)
	n (%)	n (%)	n (%)
Any endorsement on Verbally Pressured Sexual Exploitation module*	187 (53.9)	120 (68.2)	60 (32.3)
27. Someone pressured me by saying positive things such as complimenting me, making promises about the future, saying “I love you” or “just do this for me” to:
a. Have non-penetrative sexual contact	103 (29.7)	71 (40.3)	30 (18.8)
b. Have oral sex	86 (24.8)	56 (31.8)	26 (16.3)
c. Penetrate my vagina/genital opening	71 (20.5)	62 (35.2)	5 (3.1)
d. Make me penetrate their vagina or genital opening	28 (8.1)	5 (2.8)	22 (13.8)
e. Penetrate my anus	28 (8.1)	22 (12.5)	4 (2.5)
f. Make me penetrate their anus	5 (1.4)	0 (0)	5 (3.1)
28. Someone pressured me by trying to wear me down by doing things such as asking over and over, pleading, sighing, or rolling their eyes to:
a. Have non-penetrative sexual contact	80 (23.1)	55 (31.3)	21 (13.1)
b. Have oral sex	66 (19.0)	47 (26.7)	16 (10.0)
c. Penetrate my vagina/genital opening	67 (19.3)	63 (35.8)	2 (1.3)
d. Make me penetrate their vagina or genital opening	25 (7.2)	5 (2.8)	19 (11.9)
e. Penetrate my anus	26 (7.5)	19 (10.8)	6 (3.8)
f. Make me penetrate their anus	5 (1.4)	0 (0)	5 (3.1)
29. This question does NOT include situations when you were too intoxicated to give permission for sex. Someone pressured me to drink more alcohol than I wanted by doing things such as buying me drinks or making very strong drinks to engage in sex acts with me when I wouldn’t have otherwise to:
a. Have non-penetrative sexual contact	52 (15.0)	36 (20.5)	14 (8.8)
b. Have oral sex	43 (12.4)	29 (16.5)	14 (8.8)
c. Penetrate my vagina/genital opening	30 (8.6)	28 (15.9)	1 (0.6)
d. Make me penetrate their vagina or genital opening	14 (4.0)	4 (2.3)	10 (6.3)
e. Penetrate my anus	10 (2.9)	5 (2.8)	5 (3.1)
f. Make me penetrate their anus	3 (0.9)	0 (0)	3 (1.9)
30. This does NOT include situations when you were too high to give permission for sex. Someone pressured me to use more drugs than I wanted by doing things such as buying or offering me drugs to engage in sex acts with me when I wouldn’t have otherwise to:
a. Have non-penetrative sexual contact	12 (3.5)	10 (5.7)	2 (1.3)
b. Have oral sex	12 (3.5)	7 (4.0)	5 (3.1)
c. Penetrate my vagina/genital opening	10 (2.8)	10 (5.7)	0 (0)
d. Make me penetrate their vagina or genital opening	4 (1.2)	1 (0.6)	3 (1.9)
e. Penetrate my anus	3 (0.9)	2 (1.1)	1 (0.6)
f. Make me penetrate their anus	2 (0.6)	0 (0)	2 (1.3)
31. Someone pressured me by criticizing, saying things such as, “don’t be so uptight,” “no one else would want you,” “you led me on,” or by questioning my sexuality to:
a. Have non-penetrative sexual contact	57 (16.4)	41 (23.3)	14 (8.8)
b. Have oral sex	35 (10.1)	23 (13.1)	10 (6.3)
c. Penetrate my vagina/genital opening	39 (11.2)	37 (21.0)	1 (0.6)
d. Make me penetrate their vagina or genital opening	15 (4.3)	6 (3.4)	8 (5.0)
e. Penetrate my anus	11 (3.2)	7 (4.0)	2 (1.3)
f. Make me penetrate their anus	4 (1.2)	0 (0)	4 (2.5)
32. Someone pressured me by threatening the relationship in some way such as threatening to break up with me, to have sex with someone else, or questioning my commitment to the relationship to:
a. Have non-penetrative sexual contact	43 (12.4)	26 (14.8)	15 (9.4)
b. Have oral sex	38 (11.0)	24 (13.6)	13 (8.1)
c. Penetrate my vagina/genital opening	40 (11.5)	37 (21.0)	2 (1.3)
d. Make me penetrate their vagina or genital opening	20 (5.8)	1 (0.6)	18 (11.3)
e. Penetrate my anus	6 (1.7)	6 (3.4)	0 (0)
f. Make me penetrate their anus	6 (1.7)	1 (0.6)	4 (2.5)
33. Someone pressured me by threatening my social reputation, such as saying they would spread rumors about me, reveal private information, out me, or exclude me from social situations to:
a. Have non-penetrative sexual contact	20 (5.8)	14 (8.0)	5 (3.1)
b. Have oral sex	12 (3.5)	8 (4.5)	4 (2.5)
c. Penetrate my vagina/genital opening	12 (3.5)	12 (6.8)	0 (0)
d. Make me penetrate their vagina or genital opening	7 (2.0)	1 (0.6)	6 (3.8)
e. Penetrate my anus	0 (0)	0 (0)	0 (0)
f. Make me penetrate their anus	1 (0.3)	0 (0)	1 (0.6)
34. Someone pressured me by saying or doing something that reminded me about how they had been angry, threatening, aggressive, or violent in the past to:
a. Have non-penetrative sexual contact	13 (3.7)	10 (5.7)	3 (1.9)
b. Have oral sex	12 (3.5)	7 (4.0)	4 (2.5)
c. Penetrate my vagina/genital opening	10 (2.9)	9 (5.1)	1 (0.6)
d. Make me penetrate their vagina or genital opening	10 (2.9)	2 (1.1)	7 (4.4)
e. Penetrate my anus	4 (1.2)	2 (1.1)	2 (1.3)
f. Make me penetrate their anus	2 (0.6)	0 (0)	2 (1.3)
35. Someone pressured me by getting mad and by saying and doing things such as yelling, swearing, slamming things, or stomping around (but not using physical force) to:
a. Have non-penetrative sexual contact	18 (5.2)	11 (6.3)	5 (3.1)
b. Have oral sex	20 (5.8)	11 (6.3)	7 (4.4)
c. Penetrate my vagina/genital opening	17 (4.9)	17 (9.7)	0 (0)
d. Make me penetrate their vagina or genital opening	14 (4.0)	2 (1.1)	12 (7.5)
e. Penetrate my anus	4 (1.2)	3 (1.7)	0 (0)
f. Make me penetrate their anus	2 (0.6)	1 (0.6)	1 (0.6)
36. Someone pressured me by threatening my access to something important like my children, money, housing, food, healthcare, medications, or legal documents to:
a. Have non-penetrative sexual contact	14 (4.0)	8 (4.5)	4 (2.5)
b. Have oral sex	14 (4.0)	9 (5.1)	4 (2.5)
c. Penetrate my vagina/genital opening	12 (3.5)	11 (6.3)	0 (0)
d. Make me penetrate their vagina or genital opening	2 (1.2)	0 (0)	4 (2.5)
e. Penetrate my anus	2 (0.9)	2 (1.1)	0 (0)
f. Make me penetrate their anus	2 (0.6)	0 (0)	1 (0.6)
37. Someone pressured me by threatening to harm or kill themselves to:
a. Have non-penetrative sexual contact	18 (5.2)	9 (5.1)	8 (5.0)
b. Have oral sex	13 (3.7)	8 (4.5)	4 (2.5)
c. Penetrate my vagina/genital opening	8 (2.3)	8 (4.5)	0 (0)
d. Make me penetrate their vagina or genital opening	7 (2.0)	0 (0)	7 (4.4)
e. Penetrate my anus	7 (2.0)	5 (2.8)	0 (0)
f. Make me penetrate their anus	3 (0.9)	0 (0)	2 (1.3)
	M (SD)	M(SD)	M(SD)
Total summed score on the Verbally Pressured Exploitation module*	17.46 (41.41)	24.86 (49.54)	9.55 (29.63)
Participants’ estimated total number of incidents of verbally pressured sexual exploitation*	9.34 (41.83)	15.54 (57.00)	2.74 (10.40)

We did not report separate results for individuals with genders other than men or women because small group sizes for the other genders could lead to unstable estimates and inappropriate conclusions about relative vulnerability. Because of this, summed rates for women and men do not necessarily correspond to rates for the full sample, which included gender categories not reported here.

*p < .05 for comparison of differences between men and women on the total rates of endorsements for each scoring category, module and number of incidents (comparisons not conducted for individual items).

^†This item was added to the SES-V after data were collected for this study, and thus, it was not included in this administration.

The question about estimated number of incidents was added to the end of each module to help address the issue that total summed scores on past versions of the SES could not be interpreted as reflecting a meaningful count of incidents because some acts or tactics may have co-occurred in the same incident (Swartout et al., 2015; see Table 3 for continuous summed scores and participant estimates for each module). The addition of the estimated number of incidents at the end of each module did appear to help in terms of the problem of overcounting—specifically for the Verbally Pressured and Illegal Modules. When we created a total score (i.e., by summing the number of times participants reported experiencing each item from 0 to 10 or more) on the Noncontact Sexual Exploitation module, the estimated mean number of incidents for all participants was 9.40 (SD = 11.21), and when asked to estimate the total number of incidents they had ever experienced in the follow-up module, participants’ estimates were very similar (M = 9.41; SD = 21.70). For the Technology-Facilitated module, the mean summed score was 11.71 (SD = 17.04), and the mean estimated number of incidents reported by participants was 10.61 (SD = 21.46); thus, estimates were fairly similar. However, for the Verbally Pressured Sexual Exploitation module, the mean total summed score was 17.46 (SD = 41.41), and the mean estimated number of incidents reported by participants was 9.34 (SD = 41.83). For the Illegal Sexual Exploitation module, the mean total score was 12.97 (SD = 25.65), and the mean estimated number of incidents reported by participants was 6.31 (SD = 15.28). Thus, for these latter two modules, the summed score seemed to substantially overestimate the number of incidents experienced.

The question about estimated number of incidents also clearly revealed that repeat victimization was quite common. Among those that had experienced noncontact victimization, 89.3% had experienced more than one noncontact incident based on their estimate. Among those that had experienced cyber victimization, 87.3% had experienced more than one cyber incident. Among those that had experienced an illegal act, 81.3% had experienced more than one illegal incident, and among those that had experienced verbal pressure, 77.4% had experienced more than one verbally pressured incident.

In addition to scoring dichotomously and continuously (summed score) by module, Koss et al. (2024) proposed some other potential scoring models for the SES-V, which might be helpful for certain research questions and which require future research to examine their reliability and validity. These included (1) a model that scores the SES-V dichotomously by each type of sexual act that the respondent experienced without permission, (2) an ordinal scoring model, and (3) a scoring approach that allows for calculation of experiences that fit with the Federal Bureau of Investigations (2013) definition of rape. Sexual exploitation prevalence rates using these alternative scoring approaches are included in Table 4. Looking just at items that correspond to the FBI definition of rape, for example, 60% of women and 29% of men endorsed rape on the SES-V. Notably, as mentioned in Koss et al. (2024), there are substantial problems with ordinal scoring, and the SES-V revision team does not recommend its use. As revealed by these results (see Table 4), ordinal scores substantially alter the number of participants coded in the less severe categories—for example, using ordinal scoring, 7.2% of participants fall into the verbally pressured sexual exploitation group, but using module scoring, 53.9% of participants have actually experienced verbal pressure—illustrating one potential limitation of this approach.

Table 4. Preliminary estimated prevalence on the SES-V using the other potential scoring models proposed in Koss et al. (2024).

Category	Items included	Rates of Endorsement
		Total (N = 347)	Women (n = 176)	Men (n = 160)
		n (%)	n (%)	n (%)
Dichotomous Scoring by Sexual Act
Penetrative*	Response >0 on any items 16 through 25 and 27 through 36, options b, c, d, e, f	217 (62.5)	136 (77.3)	75 (46.9)
Attempted penetration*	Response >0 on item 26, options b, c, d, e, and f	40 (11.5)	29 (16.5)	10 (6.3)
Non-penetrative contact*	Response >0 on items 16–36 option a	200 (57.6)	129 (73.3)	66 (41.3)
Cybersex	Response >0 on any items 6 through 15	232 (66.9)	123 (69.9)	101 (63.1)
Non-contact *	Response >0 on any items 1 through 5	251 (72.3)	153 (86.9)	93 (58.1)
No non-permissive sexual acts disclosed*	No response > 0	36 (10.4)	4 (2.3)	31 (19.4)
Rape Prevalence using the FBI Definition
Rape (encompassing both completed and attempted)*	Any response > 0 on items 16 through 22, options b, c, or e and/or any response > 0 on item 26, options b, c, or e	159 (45.8)	106 (60.2)	47 (29.4)
Completed rape*	Any response > 0 on items 16 through 22, options b, c, or e	157 (45.2)	105 (59.7)	46 (28.7)
Attempted rape	Any response > 0 on item 26, options b, c, or e and no report of completed rape	2 (0.6)	1 (0.6)	1 (0.6)
Ordinal Scoring*
0 = No sexual exploitation disclosed	No response > 0 on any item	36 (10.4)	4 (2.3)	31 (19.4)
1 = Non-contact sexual exploitation	Response >0 on any items 1 through 5 No other items with > 0 times	17 (4.9)	8 (4.5)	9 (5.6)
2 = Technology-facilitated sexual exploitation	Response >0 on any items 6 through 15 No responses > 0 on items 16 through 37	50 (14.4)	12 (6.8)	35 (21.9)
3 = Verbally pressured sexual exploitation	Response >0 on items 27–37. No responses > 0 on items 16–26.	25 (7.2)	14 (8.0)	11 (6.9)
4 = Non-penetrative sexual exploitation	Response >0 on item 26 or any response > 0 on items 16 through 25, option a. No responses > 0 on the penetration items.	37 (10.7)	23 (13.1)	13 (8.1)
5 = Penetrative sexual exploitation	Response >0 on any items 16 through 25, options b, c, d, e, and f. The score on any other items is irrelevant if penetration is reported.	182 (52.4)	115 (65.3)	61 (38.1)

We did not report separate results for individuals with genders other than men or women. Because of this, summed rates for women and men do not necessarily correspond to rates for the full sample, which included gender categories not reported here. Koss et al. (2024) also proposed a dichotomous and continuous module-level scoring model; those findings are presented in Table 3. * p < .05 for comparison of differences between men and women.

Gender Differences

We conducted chi-square analyses to examine whether there were gender differences in prevalence rates on each of the modules and ANOVAs to examine whether there were gender differences in estimated number of incidents for each type of exploitation. We followed the Centers for Disease Control guidelines and only analyzed and reported on demographic categories that included at least 30 participants (Parker et al., 2017); thus, for gender comparisons, we only compared men and women. Notably, this means that, when reporting prevalence rates, the summed rates as a function of gender often do not match the totals for the full sample. See Table 3 for all descriptive statistics.

There was a gender difference in likelihood of endorsing one or more item on any module of the SES-V, χ² (df = 1; n = 336) = 26.27, p < .001, φ = .28, with women being more likely to endorse than men. There was no significant gender difference in likelihood of endorsing one or more items on the Technology-Facilitated Sexual Exploitation module, χ² (df = 1; n = 336) = 1.72, p = .19, φ = .07, and no difference in reported number of incidents of technology-facilitated experiences, F (1, 327) = 2.40, p = .12, η_p² = .007. There was a gender difference in likelihood of endorsing one or more items on the Non-Contact Sexual Exploitation module, χ² (df = 1; n = 336) = 35.46, p < .001, φ = .33, with women having a higher likelihood of endorsing non-contact experiences than men. There was also a significant difference in reported number of non-contact incidents, F (1, 329) = 22.27, p < .001, η_p² =.06, with women reporting a larger number of incidents than men. There was a significant gender difference in likelihood of reporting one or more items on the Verbally Pressured Sexual Exploitation module, χ² (df = 1; n = 336) = 31.72, p < .001, φ = .31, with women having a higher likelihood of endorsing pressure experiences than men. There was also a significant difference in reported number of incidents of pressured experiences, F (1, 317) = 7.34, p = .007, η_p² =.02, with women reporting more incidents than men. Finally, there was a significant gender difference in likelihood of reporting one or more acts on the Illegal Sexual Exploitation module, χ² (df = 1; n = 336) = 37.22, p < .001, φ = .33, with women having a higher likelihood of endorsing illegal experiences than men. There was also a significant difference in reported number of incidents of illegal acts, F (1, 324) = 6.76, p = .01, η_p² =.02, with women reporting more incidents than men.

Other Common Demographic Correlates

Consistent with guidelines, we only analyzed prevalence differences for demographic categories that included at least 30 participants (Parker et al., 2017). Based on a power analysis (Faul et al., 2009), across demographic comparisons of prevalence rates, the size of our subgroups² in the following analyses provided us between 74% and 96% power to detect a medium effect size (w = .3). Comparing LGB+ individuals to heterosexual individuals,³ we found no statistically significant difference in LGB+ individuals’ (93.7%) and heterosexual individuals’ (88.4%) likelihood of endorsing one or more items on the SES-V, χ² (df = 1; 347) = 1.80, p = .18, φ = .07. Looking only at more severe acts, there also was no significant difference in LGB+ individuals’ (70.9%) and heterosexual individuals’ (60.8%) likelihood of endorsing one or more items on the Illegal Sexual Exploitation module, χ² (df = 1; 347) = 2.66, p = .10, φ = .09. There was also no significant difference in LGB+ and heterosexual individuals’ estimates of the number of incidents of illegal acts that they had experienced (M = 9.14; SD = 16.41 and M = 5.47; SD = 14.86, respectively), t (335) = 1.86, p = .06, d = .24. We compared Black and White participants’ endorsement rates (as these were the only two racial/ethnic groups with greater than 30 individuals). There was no difference in overall prevalence on the SES-V between White and Black participants (88.6% versus 92.1%, respectively), χ² (df = 1; 309) = 0.43, p = .51, φ = .04, nor was there a difference in rates of illegal sexual exploitation (61.6% versus 71.1%, respectively), χ² (df = 1; 309) = 1.27, p = .26, φ = .06. There was no significant difference between White and Black participants in estimated number of incidents of illegal acts either (M = 6.90; SD = 16.61 and M = 3.87; SD = 8.45, respectively), t (297) = 1.10, p =.27, d = .19. Endorsement on the SES-V also was not associated with whether participants reported a disability/functional impairment (15.9% of the sample reported one or more disabilities/functional limitations) or not (92.7% versus 89.0%, respectively), χ² (df = 1; 347) = 0.68, p = .41, φ = .04, nor was endorsement of an illegal act of sexual exploitation (74.5% versus 61.0%, respectively), χ² (df = 1; 347) = 3.67, p = .06, φ = .10. However, there was a significant difference in the estimated number of incidents of illegal acts reported by those with and without a disability/functional impairment (M = 11.85; SD = 21.52 and M = 5.25; SD = 13.57, respectively), t (335) = 2.94, p = .003, d = .44, such that those with a disability reported a larger number of incidents of illegal sexual exploitations than those without a disability. Effect sizes for all the non-significant demographic comparisons were near zero to small.

New Items

Because the SES-V includes several new items that were not part of the prior version of the SES (Koss et al., 2007), we wanted to examine whether these new items captured unique instances of sexual exploitation, and importantly, whether they identified individuals who had experienced sexual exploitation and who would have been missed by the prior version of the measure. Although we hoped that endorsement on these items would capture unique instances, we also expected that, on average, endorsement on these items would be correlated with endorsement of other sexual exploitation items given that victims often experience multiple victimizations and given that many incidents of sexual exploitation include multiple exploitative tactics and/or nonconsensual acts within a single experience. Thus, a moderate correlation between these new items and other items on the same module would provide some evidence that the new items fit within the constellation of behaviors that we were labeling as sexual exploitation.

Made-To-Penetrate Items

A total of 58 individuals (16.7% of the total sample; 46 men, 9 women)⁴ endorsed one or more made-to-penetrate item on the Illegal Sexual Exploitation module. Those who endorsed only the made-to-penetrate items and did not endorse any other items on the Illegal Sexual Exploitation module included seven men (4.4% of the men in the total sample; six heterosexual men; one bi/pansexual man) and one heterosexual woman. These eight individuals would have been missed if the made-to-penetrate items had not been included. Endorsement of one or more made-to-penetrate item on the Illegal Sexual Exploitation module was weakly to moderately correlated with endorsing one or more other items on the Illegal Sexual Exploitation module (86.2% of those that endorsed the made-to-penetrate item also endorsed another item on the module), φ = .23, p < .001.

A total of 57 individuals (16.4% of the total sample; 39 men, 16 women) endorsed one or more made-to-penetrate item on the Verbally Pressured Sexual Exploitation module, including 12 men (7.5% of men in the total sample; nine heterosexual, one bi/pansexual, one gay, and one questioning man) and two women (1.1% of women in the total sample; one heterosexual and one queer-identified) who did not endorse any other items on the Verbally Pressured Sexual Exploitation module. Endorsement of one or more made-to-penetrate items on the Verbally Pressured Sexual Exploitation module was weakly to moderately correlated with endorsing one or more other items on the Verbally Pressured Sexual Exploitation module (75.4% of those that endorsed the made-to-penetrate item also endorsed another item on the module), φ = .23, p < .001.

Not Providing an Opportunity to Object Item

A total of 125 individuals (36.0% of the sample; 83 women, 37 men) endorsed experiencing one or more sexual acts when they were given no notice or opportunity to consent or refuse (including when they were sleeping). Of those, 23 individuals (13 women, nine men) did not endorse any other items on the Illegal Sexual Exploitation module, meaning that this item uniquely identified 6.6% of the total sample as a victim of illegal sexual exploitation. Endorsement on the new item measuring no opportunity to object was moderately correlated with endorsing another item on the Illegal Sexual Exploitation module (81.6% of those that endorsed the new item also endorsed at least one other item on the module), φ = .46; p < .001.

Reproductive Coercion Item

A total of 63 participants (18.2% of the sample; 35 women, 25 men) endorsed at least one sexual act obtained through reproductive coercion, including 11 individuals (3.2% of the total sample; four women and seven men) who did not endorse any other items on the Illegal Sexual Exploitation module. Endorsement of the reproductive coercion item had a weak to moderate association with endorsing one or more other items on the Illegal Sexual Exploitation module (82.5% of those that endorsed reproductive coercion also endorsed another item on the module), φ = .22, p < .001.

Substance-Related Coercion Items

A total of 73 individuals (21.0% of the total sample; 49 women, 22 men) endorsed one or more of our new substance-related coercion items, including seven individuals (2.0% of the total sample; three women and four men) who did not endorse any other items on the Verbally Pressured Sexual Exploitation module. Endorsement on the substance-related coercion items was moderately correlated with endorsing other items on the Verbally Pressured Sexual Exploitation module (90.4% of those that endorsed the substance-related coercion items also endorsed one or more other Verbally Pressured module items; φ = .40, p < .001). Unsurprisingly, there was also strong overlap between the substance-related coercion items and the substance-related illegal sexual exploitation items, φ = .44, p < .001. Of those that endorsed a substance-related coercion item, 64 (87.7%) also endorsed a substance-related illegal exploitation item.

Incident Report

Based on all the experiences that we asked about on the SES-V, participants were asked to report details about the single experience that they remembered best. A total of 275 individuals (160 women, 106 men) completed at least some items on the final incident report at the end of the SES-V and indicated that they had had an incident of sexual exploitation.⁵ Of those that reported at least one act of sexual exploitation on the SES-V, 36 individuals (11.6% of those with a sexual exploitation experience) skipped all items on the specific incident report even though they went on to complete items on the demographics measure. In other words, most individuals with a sexual exploitation experience did complete the incident report.

Characteristics of the Best Remembered Incident

Participants could endorse more than one act and tactic that occurred during their best-remembered event in recognition of the fact that a single incident of sexual exploitation often involves multiple acts and tactics.⁶ Among women who completed the item (n = 156), the most common exploitative tactic reported in the incident was having their refusal ignored (n = 74; 47.4%), followed by not being given a chance to deny permission (n = 62; 39.7%). Among women completing the item (n = 151), the most common sexual act experienced was being kissed, touched, stroked, or grabbed in a sexual way (n = 96; 63.6%), followed by vaginal/genital opening penetration of their body (n = 87; 57.6%). Women most often reported that the perpetrators were men (n = 150; 96.2% of the 156 victimized persons who completed the item). The most common relationship respondents (n = 157 who completed the item) reported with the perpetrator was “acquaintance or casual friend” (n = 39; 24.8%), followed by “someone you just met” (n = 36; 22.9%).

Among men completing the item (n = 102), the type of exploitation tactic was evenly divided among nonconsensual electronic communication (n = 35; 34.3%), not being given a chance to deny permission (n = 36; 35.3%), and verbal pressure (n = 37; 36.3%). The most common sexual act experienced among men completing the item (n = 95) was non-consensual kissing, touching, or stroking of body parts (n = 50; 52.6%), with the next most common being having oral sex performed on them (n = 31; 32.6%). Over half of men reported that the perpetrators were women (n = 69; 66.3% of the 104 that completed the item); 37 (35.6%) reported a man as perpetrator. The most common relationship with the perpetrator reported by respondents (n = 104 who responded to the item) was “acquaintance or casual friend” (n = 33; 31.7%), followed by “stranger” (n = 21; 20.2%).

Labeling of the Best Remembered Incident

A total of 88 individuals (32% of the 269 who answered the question), including 71 women (44.9%) and 14 men (13.9%), indicated that they would label their best remembered experience as “rape.” More specifically, using a definition that is similar to legal definitions of rape or sexual assault, of the 110 individuals (78 women, 29 men) whose best remembered event included oral, anal, or vaginal sex obtained through an ignored refusal, threat/force, or incapacitation, 67 (62.6%) labeled the incident as rape (not including three individuals that were missing data on the rape label item). There was a significant gender difference, χ² (df = 2; 107) = 12.76, p = .002, φ = .35. Women were more likely to label the incident as rape (71.1%) than men (35.7%).

Discussion

This study represents the first administration of the new SES-V in a national community sample. One of the primary goals in revising the SES was to increase the inclusivity of the instrument. The first version of the SES was primarily designed to assess college women’s experiences of sexual exploitation perpetrated by men; later iterations have been increasingly more inclusive, with the SES-V representing the revision with the greatest attention to ensuring relevance across a variety of research populations.

Estimated Prevalence and Number of Distinct Incidents

The SES-V was designed to better capture lower severity sexual exploitation (e.g., noncontact and technology-facilitated) and to better capture the experiences of under-researched victims (e.g., men and sexual minorities) than prior versions of the SES. Consistent with this goal and consistent with findings from Anderson, Peterson, Koss, et al. (2024) in a college sample, prevalence rates on the SES-V were quite high in this diverse community sample.

As a point of comparison, the Center for Disease Control’s 2016/2017 National Intimate Partner and Sexual Violence Survey (NISVS; Basile et al., 2022) recruited a nationally representative sample (N = 27,571) and found that 26.8% of women and 3.8% of men reported lifetime experiences of attempted or completed rape, defined as, “completed or attempted unwanted vaginal (for women) or oral, or anal penetration through the use of physical force … or threats to physically harm and includes when the victim was too drunk, high, drugged or passed out and unable to consent.” In contrast, even when excluding the made-to-penetrate sexual acts (which are not included in the NISVS definition of rape), we found that 60.2% of women and 29.4% of men reported attempted or completed rape as defined by the FBI since age 14 on the SES-V (see the Rape Prevalence Using the FBI Definition in Table 4). This discrepancy is notable but not entirely surprising. There are a variety of methodological differences between this study and the NISVS that may account for these differences. NISVS data were collected through a random-digit-dial telephone survey conducted by female interviewers. Prior research suggests individuals may be more likely to disclose sexual victimization and other sensitive experiences in an online survey than through phone or face-to-face interviews (McCallum & Peterson, 2012; Parks et al., 2006). Additionally, the NISVS includes fewer items to measure each construct than the SES-V; for example, to measure completed rape on the NISVS, the researchers asked about two tactics (“you were unable to consent to sex or stop it from happening because you were too drunk, high, drugged, or passed out from alcohol or drugs” and “used physical force or threats of physical harm”) combined with two sexual acts (oral sex and vaginal/anal penetration; Kresnow et al., 2022, pp. 19–20). In contrast, the SES-V completed rape scoring category includes seven separate tactics used to obtain three separate sexual acts. Prior research suggests that using a larger number of behaviorally specific questions—and including fewer components within each item—may result in higher rates of reporting for sensitive behaviors (Abbey et al., 2021; Matthews et al., 2020; Strang & Peterson, 2017).

Still, as discussed in Peterson et al. (2024), the SES-V Illegal Sexual Exploitation module may capture some experiences that would not fit with all definitions of a sex crime. Throughout its multiple iterations, the SES has included items that are intended to correspond to legal definitions of sex crimes. Following from this important historical precedent, the goal with the SES-V Illegal Sexual Exploitation module was to identify behaviors that correspond to legal definitions in many jurisdictions internationally, while also acknowledging (1) that legal definitions vary widely and (2) the fact that a behavior corresponds to a legal definition in a particular jurisdiction does not mean that it would necessarily be prosecuted in that jurisdiction. For example, the new SES-V item, “Someone gave no advance notice so I could not agree or object, including times when I was sleeping or semi-awake” when resulting in oral, vaginal, or anal penetration, is included in the calculation of completed rape on the SES-V (and is not reflected in the NISVS operational definition of rape, for example), but even though this item describes a clearly non-consensual sexual behavior, it is not clear that this item would necessarily be perceived by most prosecuting attorneys in the U.S. as fitting legal definitions of a sex crime (see Peterson et al., 2022). Even though prosecutors might not count this item within their definition of rape, this item does seem to fit within the federal definition of rape (“The penetration no matter how slight, of the vagina or anus with any body part or object or oral penetration by a sex organ of another person, without the consent of the victim;” U.S. Department of Justice, 2013). Further, as noted by Peterson et al. (2024), inclusion of this item and other items that fall on the border of legality are, to some degree, aspirational, as we believe that all sexual acts occurring without an opportunity to provide permission should be illegal.

Thus, the breadth of the SES-V can be viewed as a strength, in that it captures a larger range of sexual exploitation than other measures, but also as a limitation, in that it may capture some behaviors that fall on the ambiguous borderline between criminalized and non-criminalized exploitation, and this may make it difficult to draw clear lines between different categories of sexual exploitation, such as between coercion and illegal rape (see also Koss et al., 2024, for a discussion of the overlapping categories of sexual exploitation).

Importantly, one new feature in the SES-V is a question asking participants to estimate the number of separate instances of sexual exploitation they have experienced in relation to each module. This was added to address concerns that a total summed score on the prior version of the SES would likely result in an overestimation of the number of incidents participants had experienced because some incidents could include multiple sexual acts and/or exploitative tactics (Swartout et al., 2015). Indeed, we found that summed scores, except in the case of the Non-Contact Module, resulted in higher estimates of the number of incidents than were obtained through participants’ own estimated number of unique incidents.⁷ This was especially true for the Illegal Sexual Exploitation and Verbally Pressured Sexual Exploitation modules, likely because in those modules, summed total scores included multiple different sexual acts as well as multiple different exploitation tactics, allowing for more opportunity of multiple experiences within a single incident. Given that, we recommend not calculating a total score on the SES-V as a way to estimate the number of sexual exploitation incidents that a respondent has experienced. Relying on the participants’ own estimates from the module follow-ups is likely to produce a more accurate estimate if the researcher’s goal is to identify total number of sexual exploitation experiences. Importantly, the new items estimating number of incidents are provided at a module-level, so it could still be misleading to add the incident numbers across different modules because some incidents might involve both illegal tactics and verbally coercive tactics, for example.

Notably, the fact that the total score on the SES-V (or on any given module) overestimates incident number does not mean that there is no value in asking the number of times that participants have experienced each type of sexual exploitation. In the SES-V, we specifically chose to ask about frequency and to include a larger number of response options (0 to 10 or more) as compared to most prior measures of sexual exploitation because research suggests that individuals are more likely to endorse sexual victimization when provided with a scaled frequency response format as compared to a yes/no response format (Anderson & Cuccolo, 2022). This is consistent with classic work by Schwarz et al. (1999) demonstrating that participants infer that the midpoint of a response scale is the normal or typical response, so especially for stigmatized experiences such as sexual victimization, providing a larger response range may increase participants’ likelihood of disclosing on the measure. Thus, we believe there is value in using the frequency response, even if researchers opt to dichotomize scores for the purposes of analyses.

Further, there might be value in calculating a total summed score on the SES-V or on its modules in some contexts even though that total score is likely to overestimate the number of separate incidents that participants have experienced. For example, it is possible that experiencing a larger total number of nonconsensual sexual acts and/or exploitative tactics may result in worse outcomes regardless of whether the acts/tactics occurred during the same or separate incidents. Kern et al. (2022) found that a larger number of sexual acts and a larger number of exploitative tactics experienced during a single incident of sexual exploitation was associated with more posttraumatic stress disorder symptoms among a sample of male and female sexual exploitation victims. Future research is needed to examine the predictive validity of summed SES-V module scores versus scores based on the participants’ own estimates of the number of sexual exploitation incidents.

Demographic Correlates

Prior research on sexual exploitation has identified some consistent demographic correlates of victimization. Thus, demonstrating a relationship between endorsement on the SES-V and those demographic correlates would provide some evidence of construct validity. Replicating expected patterns found in prior research (Basile et al., 2022; Iverson et al., 2013), compared to men, women reported a higher prevalence of total sexual exploitation and a higher prevalence of exploitation on all but one of the modules. The single exception was the Technology-Facilitated Sexual Exploitation module. It is surprising that we did not find a gender difference in prevalence of technology-facilitated exploitation, as prior research has found higher rates of online sexual harassment among women than men (Buchanan & Mahoney, 2022; Copp et al., 2021). Notably, though, many of the prior studies of technology-facilitated sexual harassment were conducted with adolescent or college samples. Our community sample ranged in age from 18 to 83. Older adults’ experiences of online sexual exploitation have received minimal attention, and correlates of technology-facilitated sexual exploitation in an older community sample may be different than in younger student samples. It is also possible that participants in our Prolific Academic sample spend more time online than the average person, making them more vulnerable to technology-facilitated sexual exploitation. Given that the technology-facilitated content on the SES-V is entirely new (i.e., did not exist on prior versions of the SES), further research on these items in a variety of populations is needed.

Although we mostly found the expected pattern of results in relation to gender, we did not replicate most of the other demographic differences that we expected based on past research. For overall endorsement of sexual exploitation, we did not find differences between sexual minority and sexual majority individuals, between Black and White individuals, or between individuals with and without a disability/functional impairment. All of these non-significant findings are inconsistent with prior research (e.g., Basile et al., 2016, 2022; Chen et al., 2020). Importantly, because the SES-V is specifically designed to better capture low severity sexual exploitation (including online and noncontact), most participants in the sample endorsed at least one item on the SES-V; this may have made it more difficult to detect subgroup differences. Yet, when we compared demographic groups on more severe forms of sexual exploitation (i.e., the Illegal Sexual Exploitation module), we still did not find the demographic differences that we anticipated based on prior research. The only exception is that we did find that those with a disability/functional impairment reported a larger number of incidents of illegal sexual exploitation than those without, providing some evidence of the value of estimates of number of incidents in addition to estimates of prevalence.

It is difficult to know what accounts for the fact that we didn’t find the demographic differences that have been found in other studies, but there are multiple potential explanations: First, the SES-V, including the Illegal Sexual Exploitation module of the SES-V, assesses a broader range of sexual exploitation—including more sexual acts and exploitative tactics—than the measures used in prior studies; perhaps this broader conceptualization of sexual exploitation led to reductions in the demographic differences in prevalence estimates. Second, as noted above, much of the prior research—especially the research comparing prevalence rates as a function of sexual identity—has focused on college students, whereas this study utilized a community sample that was diverse in terms of age; demographic vulnerability factors may change with age. Third, we may have failed to detect differences in this sample simply because our subgroups of LGBQ+ individuals, Black individuals, and individuals with disabilities/functional impairments were fairly small and provided only enough power to detect moderately sized main effects. This means that we lacked power to detect smaller, but perhaps still clinically relevant, group differences, and we lacked power to examine interactions among demographic characteristics (e.g., how gender and sexual identity interact to predict vulnerability). Further research is needed in larger national samples with oversampling to ensure that all racial, sexual, and gender groups are adequately represented. This is particularly important given that many of the changes in the SES-V were specifically designed to increase the relevance of the measure for individuals with diverse gender and sexual identities. In fact, this research is partially underway, with a national study of the SES-V in a LGBTQ+ population already having completed data collection. It is also important to note that our subsample of individuals with disabilities/functional impairments combined individuals with many diverse types of impairments, and we did not assess whether the onset of the disability/functional impairment preceded or followed the experiences of sexual exploitation. More nuanced research is needed to examine how different disabilities or impairments do or do not increase vulnerability to sexual exploitation. Further, to our knowledge, researchers have not specifically examined the psychometric properties of sexual victimization measures in samples of individuals with disabilities; research is needed to evaluate whether existing measures (including the SES-V) adequately capture experiences of sexual exploitation among individuals with a variety of different types of disabilities.

New Items

The new SES-V sexual acts and exploitative tactics that we evaluated in this study demonstrated utility in detecting sexual exploitation that would have been missed by prior measures. For example, as we anticipated based on prior research (Anderson et al., 2020; Kern et al., 2022; Littleton et al., 2020), the inclusion of the new sexual acts measuring instances in which someone was made to penetrate another person’s vagina/genital opening or anus allowed us, in particular, to capture sexual exploitation experiences of heterosexual cis men that might otherwise have been missed (e.g., situations in which men are made to engage in penile-vaginal intercourse with a woman). Although the numbers were comparably small, there were also women who reported being made to penetrate another person through verbal pressure or illegal tactics. The NISVS includes an assessment of experiences of being made to penetrate another person, but that item is administered only to cis men (Basile et al., 2022); our results suggest that it may have relevance for some women as well.

The new exploitative tactics that we added to measure illegal acts and verbal pressure also functioned as intended. That is, they captured unique instances of sexual exploitation that would have been missed using older versions of the SES, but endorsement of the new items was significantly correlated with endorsement of other items on the same module, suggesting that the new items capture something that is consistent with the experiences that we are labeling as sexual exploitation.

Incident Report and Rape Acknowledgement

The specific incident report at the end of the SES-V asks participants to report on details of a single incident of sexual exploitation. In this study, we asked about the incident that they remember best. In the current study, this report offered insights into common sexual acts and exploitative tactics experienced by men and women. There were gender similarities in the sexual act, with non-penetrative sexual contact being most common among both men and women. The gender differences were a bit more pronounced in relation to the exploitative tactics: Women most often reported that the exploitation happened because their refusal was explicitly ignored or because they were not provided an opportunity to object. Men most commonly reported one of three exploitative tactics: electronic/online pressure, verbal pressure, or a lack of opportunity to object. This illustrates that, not only do men and women experience sexual exploitation at different rates, but that their experiences are also qualitatively different in many cases. This inclusion of the incident report in the SES-V may allow for a better understanding of how sexual exploitation experiences differ across different subgroups.

The incident report also includes a measure of rape acknowledgment. Of those individuals who completed the incident report based on an experience of oral, vaginal, or anal sex obtained through ignored refusal, threat, force, or incapacitation, 28.9% of women and 64.3% of men were unacknowledged rape victims, meaning that they had experienced something that would fit most definitions of rape but did not label their experiences as “rape.” This is consistent with prior studies that have found that women are more likely to label their experiences as rape than men (Reed et al., 2020). However, rates of unacknowledged rape in this study were lower than has been found in prior studies—slightly so for men (64% in this study versus 76% and 80% in other studies; Artime et al., 2014; Reed et al., 2020) and substantially so for women (29% in this study versus 60% in a prior meta-analysis; Wilson & Miller, 2016). Again, though, our unique sample may help to account for this difference. The majority of studies (25 out of 30) in Wilson and Miller’s (2016) meta-analysis of women’s rape acknowledgment were young adult college samples, and unacknowledged rape rates were higher in college than non-college samples in their meta-analysis. Similarly, the male samples in Artime et al. (2014) and Reed et al. (2020) were relatively young (32 and 22 years old on average, respectively) compared to our sample in this study, and prior research has found that increased age is associated with greater likelihood of using the label rape (Donde et al., 2018).

Limitations and Future Directions

This study was designed to provide preliminary prevalence estimates in a national sample using the new SES-V measure. These are just preliminary estimates, however. This sample is not large enough to allow for conclusions about stable prevalence rates. Further, although we aimed to recruit a sample that was representative in terms of age, sex, and race/ethnicity, there were some demographic categories (namely the oldest age categories and the Hispanic/Latino/a/x ethnic category) that were underrepresented in this sample. Further, we were limited in our ability to conduct demographic comparisons because of our many small demographic subgroups. Further, the individuals who volunteer for the Prolific Academic panels are likely to differ from the general population in a variety of ways (including access to the Internet and a device to complete online questionnaires). Additional research on the SES-V using a larger-scale nationally representative sample is needed.

The SES-V clearly captures a larger range of sexual exploitation than prior versions of the SES (see also Anderson, Peterson, Koss, et al., 2024) and than other measures of sexual exploitation (e.g., the NISVS). As noted, the high estimated prevalence of sexual exploitation detected by the SES-V can be viewed as both a strength and a limitation. One the one hand, the high rates of sexual exploitation found in this study provide evidence that we are fully capturing the lower severity end of the sexual exploitation continuum as intended in the conceptual development of the SES-V (see Koss et al., 2024). Scholars have argued that acceptance of low-level sexual exploitation, such as unwanted sexual comments and cyber sexual harassment, contributes to a culture that tolerates more severe sexual exploitation, such as rape (e.g., Dodge, 2016; Logan, 2015). Given that, the broad range of exploitation included on the SES-V helps to draw attention to this constellation of problematic behaviors. The result of applying a broader conceptualization of sexual exploitation is—predictably—high prevalence rates when scores on all items are considered. On the other hand, sexual exploitation researchers must discuss results based on the SES-V with care. Historically, skeptics and critics of sexual exploitation research in general, and research using the SES specifically, have argued that feminist researchers are intentionally exaggerating rates of sexual exploitation to advance their political policy agendas (see Koss & Cook, 1998; Muehlenhard et al., 1994; Rutherford, 2017, for reviews of and responses to these criticisms). Reporting that nearly 90% of participants have been victims of sexual exploitation on the new SES-V might fuel such critiques. Thus, it will be important for scholars to thoughtfully address the large range and diversity of behaviors that is included within that 90% endorsement rate. Reporting on prevalence rates at the modular level is probably more useful than reporting a total prevalence rate on the SES-V to communicate the range of behaviors assessed and to isolate those that are illegal or potentially illegal.

In future research, we plan to scale the items on the three dimensions of sexual exploitation discussed by Koss et al. (2024): invasiveness, pressure, and social tolerance. We expect that many individuals will not perceive the lower-level behaviors assessed by the SES-V as exploitative (i.e., will rate those items low on each of the three dimensions), whereas we expect that nearly all individuals will agree that rape is exploitative (i.e., will rate those items high on each of the three dimensions). Yet, by including relatively socially acceptable behaviors within the spectrum of sexual exploitation, we hope to start a dialogue in which we challenge the status quo, which is a society where technology is commonly used to sexually victimize people, where some people feel free to invade personal boundaries both verbally and physically through unsolicited sexual comments and touching, and where sex acts are performed without providing an opportunity for discussions of consent.

Although we found expected gender patterns in rates of sexual exploitation, we did not find significant correlations between endorsement on the SES-V and many of the other demographic variables that have been well established as correlates of sexual victimization in prior studies. Additionally, rates of rape acknowledgment were substantially higher in this study than has been found in prior studies. More research is needed to evaluate whether these unexpected findings reflect our unique (and relatively small) sample (which is substantially older than the samples included in most studies of sexual exploitation) or whether these findings reflect some validity issues with the new SES-V. In particular, research is needed to evaluate convergent and divergent validity of the SES-V in different types of samples. Cognitive interviewing also would be helpful in establishing construct validity.

Additionally, further research on the follow-up module questions and specific incident report is warranted. For example, given that many participants in our study reported multiple sexual exploitation incidents, how did they estimate incident numbers and how confident were they in their estimates? Similarly, how do participants interpret the instructions to focus on a single incident of sexual exploitation in the incident report? How do different commonly used prompts, such as “most distressing,” “most upsetting,” or “most serious,” impact reporting on the incident report (e.g., Abbey et al., 2004; Kern et al., 2022; Ullman et al., 2007)? In this study, we chose to ask participants to focus on the “best remembered” incident, thinking that might make it easier for them to answer detailed questions about the experience (especially if it occurred long ago). However, it is possible that this prompt may result in an underrepresentation of events that occurred during intoxication or when asleep; such events may be important and upsetting but difficult to remember. Given this concern, in the final published version of the SES-V (Koss et al., 2024), we revised our thinking and suggested that researchers ask about the “most serious” incident. Of course, the best prompt for the incident report will depend on scholars’ specific research questions. Future research should compare responses on the incident report when different types of prompts are provided.

Finally, it is also worth noting that the SES-V is longer than its predecessor, the SES-SFV (Koss et al., 2007). Thus, another disadvantage to the expanded breadth of the measure is a longer completion time and greater participant burden. In this study, participants took a median of 18 minutes to complete all modules of the SES-V, the follow-up items for each module, the incident report, and the demographics measure. As discussed in Koss et al. (2024), the SES-V development team was aware of this increased burden but felt that it was important to capture a range and diversity of sexual exploitation in this measure. To accommodate researchers’ needs for a shorter version of the measure, however, the SES-V was modularized, meaning that researchers can choose only the module that best fits their research interests, and the follow-items and the incident report are optional.

Conclusions

Although many research questions remain and more research is needed on the psychometrics of the SES-V, this study represents the first administration of the full SES-V with the (nearly final) published wording in a national sample that is demographically representative of the population. This study provides some preliminary evidence that the SES-V functions as expected in a diverse community sample. Specifically, findings from this study suggest that the new items on the SES-V have utility (meaning that they contribute unique information) and validity (meaning that they were correlated with other items), that the questions from the module follow-ups asking participants to estimate the number of incidents of sexual exploitation they have experienced is more meaningful than summed scores for assessing number of incidents, and that the specific incident report included at the end of the SES-V provides novel and valuable information about the common contexts of sexual exploitation. This is the first step in what we hope will be a large body of research applying and evaluating the SES-V in a variety of different populations.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

During preparation of this article, Dr. Anderson’s time was partially funded by a grant from the National Institute on Alcohol Abuse and Alcoholism [NIAAA grant K01AA026643]. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Funding for this project was provided by the University of Arizona Regents’ Professor Funds awarded to Mary Koss.

Notes

¹ One item in the published version of the SES-V (Koss et al., 2024) was not included in this administration. After data were collected for this study, an item was added to the Illegal Sexual Exploitation module (item 25) to assess commercial sexual exploitation.

² Given the unequal subgroup sizes in the demographic comparisons, power was calculated based on the n for the smallest subgroup in each analysis.

³ We combined all non-heterosexual individuals even though there were more than 30 bisexual participants because bisexuality was heavily confounded with gender. Specifically, about 71% of the bisexual participants were women. Unfortunately, though, our group of bisexual women, specifically, fell below our threshold of 30 participants, so we could not examine them separately.

⁴ Total numbers include individuals that identified with genders other than man or woman, but we do not report rates of endorsement separately for gender categories with fewer than 30 participants.

⁵ One individual answered some questions on the incident report despite not reporting an experience of exploitation on the SES-V. That individual was excluded from all analyses of the incident report.

⁶ The exact wording for the acts and tactics within the incident report was slightly revised between the collection of these data and the finalization of the published version in Koss et al. (2024). Additionally, “electronic communication” was included as a tactic in the version of the incident report administered in this study. In the final published version of the SES-V, this was reconceptualized as a sexual act and was moved as a response option to the item asking about the sexual act(s) that occurred during the incident.

⁷ Although on average, a summed score on the SES-V items generally seemed to overestimate the number of incidents that participants had experienced, it is also possible that, in some specific cases, a summed score on the SES-V could underestimate the actual number of incidents. Because the frequency ratings have a ceiling of 10 or more times, some individuals that experienced recurrent sexual exploitation might not be able to report an accurate number of incidents within the items. The summed score for women on the Non-Contact Module was slightly lower than women’s own estimates, suggesting that this phenomenon may have been occurring in that case.