Psychometric Properties of the Sexual Services Acts Materials for Pay (SSAMP) Index: Identifying the Virtual, In-Person, and Material Sex Trades for Financial Compensation

ABSTRACT

Exchanging sex for financial compensation is thought to be underreported primarily because extant studies (1) use a single item to capture a complex, stigmatized phenomenon, and (2) do not capture the complex types or reasons why people engage in the sex trades. Few survey studies have explored the sex trades among university students. This study evaluated the psychometric properties of a newly developed measure, the Sexual Services Acts and Materials for Pay (SSAMP), among university students. Surveys were administered to undergraduate (N = 544, time 1; N = 362, time 2) and graduate (N = 331, time 1; N = 187, time 2) students two weeks apart at a predominantly White, public, Midwestern university. Findings suggested that our index had (1) strong convergent validity with the single item used in prior studies, (2) identified more cases of sex trading than the single item, (3) good internal consistency reliability (4) moderate to strong test-retest reliability, and (5) strong discriminant validity. Exploratory factor analysis revealed that items loaded above .59 on a single factor. To our knowledge, this study is the first to evaluate a multi-item sex trades measure in the U.S. Future research should continue to advance the SSAMP and adapt this index to provide credible estimates and nuanced understanding of the sex trades across contexts.

Introduction

Most studies on sex trading are limited because they use a single item to assess this complex phenomenon, e.g., “have you ever traded sex for money or drugs?” (Franchino-Olsen et al., 2021; Ulloa et al., 2016). Yet, using multiple, behaviorally specific items that avoid labeling is best practice for detecting highly sensitive and stigmatized behaviors (Cook et al., 2011). A single sex trading item cannot differentiate between virtual, in-person, or material sex trades nor be used to rigorously understand circumstances of the sex trades or establish their behavioral consequences. Furthermore, identifying virtual forms of the sex trades has become increasingly important because of ubiquitous access to the internet and the changing landscape of online sex work (Pew Research Center, 2021; Sanders et al., 2016). Recent work with university students further suggests that in-person and virtual sex trading is more widespread and relevant to a larger segment of the population than previously assumed (Blum et al., 2018; Gerassi, Lowe, et al., 2023). Therefore, developing, testing, and using high quality instruments to understand young people’s sex trading holds crucial implications for policies and interventions.

This study aimed to evaluate a multi-item sex trading index, the Sexual Services Acts and Materials for Pay (SSAMP), using a sample of undergraduate and graduate students at one university. This work is first needed to adequately capture the complex experiences of the sex trades among university students. Further, the SSAMP has the potential to be adapted for future research among minor-aged and adult populations to provide a nuanced understanding and credible estimates of the sex trades across contexts.

Background

The Sex Trades for Financial Compensation

The sex trades (sometimes referred to as sex trading or sex work) can include sex (e.g., oral, anal, vaginal) and sexual acts (e.g., touching) that can occur in-person (e.g., prostitution, exotic dancing, informal arrangements) and virtually (e.g., “webcamming” or videos/photos of stripping or masturbation) or in material form (e.g., underwear, fluids) in exchange for financial compensation (e.g., money, housing, drugs). In the US, studies most often use one item to identify this complex behavior to produce national and local prevalence estimates as well as an understanding of the acts in convenience samples that are expected to have higher rates of sex trading. For instance, the National Longitudinal Study of Adolescent to Adult Health (ADD Health) suggests that approximately 5% of young people in the general population report sex trading (Franchino-Olsen et al., 2021; Ulloa et al., 2016), while rates among high school youth range from 1.4–7.4% of youth (Gerassi et al., 2021; Head et al., 2021; Martin et al., 2020). These estimates rise substantially in populations that experience economic instability. For example, up to 1 in 4 homeless and runaway youth report sex trading (Bigelsen & Vuotto, 2013; Dank et al., 2015; Martin et al., 2010; Walls & Bell, 2011), suggesting substantial disparities among youth (Bigelsen & Vuotto, 2013; Choi et al., 2015; Fernandes-Alcantara, 2018; Greeson et al., 2019).

Albeit methodologically limited, there is growing evidence in the United States (US) and the United Kingdom (UK) suggesting that university students trade sex for financial compensation under diverse circumstances. Recent US studies, using a single item, indicate that 2–4.5% of university students report trading sex (Blum et al., 2018; Gerassi, Lowe, et al., 2023). These studies are important but do not capture the scope of this complex phenomenon. In 2012, UK researchers used a multi-item measure to survey students about their participation in “the exchange of sexual services, performances, or products for material compensation” (Sagar et al., 2016, p. 701). They found that 4.8% of 6000 participants had participated in at least one act; more than half of students who endorsed reported doing so to fund their education and cover basic living expenses (Sagar et al., 2016). Additionally, over one-fifth of participants reported having considered engaging in the sex trades. Another study of 200 university students found that 6% had participated in the sex trades (Roberts et al., 2013) with most reporting difficulty in paying bills and experiencing current debt. These findings are alarming because UK college tuition is generally capped at $12,000 USD, nearly half of the level of tuition in the US (Barr et al., 2019). Gaining a better understanding of students’ experiences using rigorously developed questionnaires has the potential to inform interventions that reduce harm.

Gaps in Survey Research

There are three primary challenges with extant survey research that predominantly use a single (or, in rare instances, three) item(s) to understanding sex trading. First, young people who engage in in-person fondling or touching but not penetration may not endorse items about trading “sex” (McPhillips et al., 2001). Thus, the sole use of this term may underestimate in-person acts of interest. Second, studies typically do not differentiate between virtual and in-person forms of the sex trades. Significant increases in internet usage, prior to and during COVID-19, have drastically changed the landscape of sex trading (Risman, 2020; Todres & Diaz, 2021). The internet has also facilitated increases in diverse sex trading forms for compensation, such as virtual sex trading (e.g., photos, videos, webcamming) (Jones, 2015, 2020; Nayar, 2017; Upadhyay, 2021; Van Doorn & Velthuis, 2018). Virtual sex trading may initially protect from some in-person harms (e.g., STIs) but is still dangerous. Research understanding the behavioral consequences for people who trade virtual forms for compensation remains largely qualitative but does suggest that they experience in-person and cyber violence, harassment, stalking, doxing (Jones, 2015, 2020; Nayar, 2017; Upadhyay, 2021; Van Doorn & Velthuis, 2018), and complex mental health consequences including substance use (Gerassi, Zimmerman, et al., 2023). Furthermore, Sagar and colleagues among others have identified sexual fluids and materials (e.g., underwear, urine) as another type of the sex trades (Sagar et al., 2016). Third, current work rarely identifies the conditions under which sex trading occur, which are significantly better captured in qualitative work (Gerassi & Nichols, 2017; Hansen & Johansson, 2023; Hoyle et al., 2011; Martin, 2013). People who trade sex experience a wide range of victimization and agency (Gerassi & Nichols, 2017). Further, sex trading occurs via three non-mutually exclusive ways, known as the 3C’s: coercion (e.g. trafficking, victimization), circumstance (i.e. to support basic needs), and/or choice (e.g. as a form of work) (Stoklosa & Ash, 2021; Weber, 2020). Capturing these nuanced experiences in survey research remains a critical gap in the literature and holds important implications for prevention and intervention. Therefore, multi-item measures are needed to (1) expand our understanding of the prevalence and complexity of sex trading, and (2) identify avenues to mitigate potential harms. Long term goals can include using longitudinal research to identify causes of sex trading and identify evidence-based policy and practice solutions. Prior to this, however, developing and testing an index that examines the sexual acts that participants consider or participate in, as well as associated compensations, consequences, harms, and strategies to mitigate harms is a crucial next step to advance scientific understanding.

Study Aims and Hypotheses

The purpose of this study was to test a newly developed and adapted multi-item SSAMP index to identify the sex trades and its associated circumstances. Our hypotheses were as follows:

1. Although we will identify more cases of sex trading with the SSAMP, an expanded multi-item, behaviorally specific sex trading index, compared to prior studies that have used a single item (Blum et al., 2018), there will be strong convergent validity between a sex trading single item and sex trading identified by the SSAMP.

2. Any sex trading reported on the SSAMP will not be statistically significantly associated with an item about food preference or with social desirability concerns (discriminant validity).

3. The index will have high internal consistency and test-retest reliability. We also explored the preliminary factor structure of the index.

4. A substantial portion of students will report each type of reason, and the reasons will have good internal consistency.

Method

This study involved surveying undergraduate and graduate students about their experiences with sex trading. Students could complete a total of two surveys that both included the SSAMP. The SSAMP consisted of items on whether and how they had ever participated or considered participating in trading sexual acts, materials, or services for financial compensation. The first survey also included brief items on mental health, substance use, adverse childhood experiences, social desirability, and demographics. The second survey, which was sent to establish test-retest reliability, was briefer. Our recruitment materials and study’s framing were heavily informed by our student workers, advisory board, and qualitative work developing this index (Gerassi, Zimmerman, et al., 2023). We indicated that our findings would be used to “support, empower, and advocate for students.”

Development of the SSAMP (Participated and Considered)

We first developed and adapted a multi-item sex trading index for university students by drawing from (1) multiple items used in prior studies of sex trading among university students in the UK and US (Blum et al., 2018; Roberts et al., 2007, 2013; Sagar et al., 2015), (2) quantitative research of young people who trade sex in the US (Dank et al., 2017; Middleton et al., 2018); and (3) language adapted from article quotes and online writings of young adults who engage in virtual sex acts (Camming Advice From a Webcam Model, n.d.; Drolet, 2020). Second, we revised the SSAMP based on cognitive interviews with students who were familiar with sex trading (Gerassi, Zimmerman, et al., 2023). This method is used to determine whether (1) respondents comprehend and answer questions accurately, and (2) how items map onto the construct of interest (e.g., sex trading). Participants in that study provided specific examples and language for virtual and in-person acts as well as materials in addition to compensation types, and circumstances that should be reflected in the index. Participants emphasized the need to introduce items with destigmatizing statements that normalize the diverse range of circumstances (including economic needs, wants, exploitation). Effort was made to use brief and concise language in accordance with best methodological practices, while also accounting for the range and nuance of students’ experiences (Schaeffer & Dykema, 2015). Third, in response to student data and research team members who had personal experiences with sex trading, we added another set of items assessing whether students who had NOT participated in sex trading ever considered doing so.

Participation in Sex Trading (Full Sample)

Our final SSAMP index asked all study participants how often they did sexual things to earn money or receive something of value such as a place to stay, clothing, cellphone, gifts, drugs in their lifetime; sold a personal item like socks or bathwater for someone else’s pleasure; sent or posted photos online of all or part of your body for someone else’s sexual pleasure; sexted or had phone sex; provided online sexual companionship; filmed yourself online doing something sexual known as webcamming or camming; danced erotically or stripped at a bar or club; gone on an in-person date or provided companionship, like sugaring, in person; had sexual contact such as sexual touching but not intercourse; had oral, vaginal, or anal intercourse. Response options included never, occasionally, sometimes, often, very often, prefer not to answer. These items were introduced with the statement “sometimes people do things for money or financial compensation because they want to, need to, or for other reasons.”

Participants were asked item-specific questions to indicate the following for each of the 9 items endorsed: (1) type of compensation received, and (2) age at which they began providing the specific act for compensation.

Sex Trading Follow-Up Items (For Those Who Endorsed Any Sex Trading)

All participants who endorsed any of the sex trading items received multiple items to identify circumstances and use of harm reduction strategies in sex trading. Much of the data gleaned from follow-up items will be the subject of a second manuscript focusing on the phenomenology of sex trading among university students. However, including items that identify the conditions under which sex trading occur is a critical component to the refinement of a multi-item index.

Circumstances

This section began with “For the rest of the survey, we refer to any of the items below as providing sexual contact, content, or services for compensation. Sometimes people do so because they need to, want to, or for other reasons.” Participants were asked about 11 different specific circumstances that were grouped as empowerment (e.g., felt personally empowered), exploitation (e.g., pressured by another person), or financial need (e.g., to pay living expenses) that contributed to their experience. Response options included: to avoid debt or pay off debt, because I was pressured or asked in a way where I couldn’t say no, because I was physically forced by another person, to please someone like a friend or dating partner, because I felt empowered to, to have spending money, to pay tuition of school-related expenses, to pay for my living expenses, to pay for my family’s living expenses, to reclaim my body after violence, to seek pleasure or try out fantasies, and an open-ended option. The options “to reclaim my body after violence” and “to please someone like a friend or dating partner” did not fit within the three broad categories and will be reported on separately. Students who indicated that they were (1) pressured or asked in a way they couldn’t say no or (2) physically forced were asked about their relationship to the people involved (e.g., friend, intimate partner, person in a position of authority, etc.)

Considered Participating in Sex Trading (Those Who Did NOT Disclose Sex Trading)

Participants who did not endorse any sex trading items were asked if they had ever considered any of the original sex trading items on a 5-point scale (never to very often). Students who endorsed at least “occasionally” or more were provided with the same summary statement as original items about referring to the items as sexual content, contact, or services for financial compensation. They were then asked, “when you’ve thought about providing sexual content, contact, or services, for financial compensation, to what extent have the following circumstances been a part of your thinking?” Response options included to avoid or pay off debt; because I was pressured or asked in a way I couldn’t say no; because I was physically forced by another person; because I felt personally empowered to; to have spending money; to pay tuition or school-related expenses; to pay for my living expenses; to pay for my family’s living expenses; to please someone like a friend or dating partner; to reclaim my body after violence; to seek pleasure or to try out fantasies; some other reason (open-ended).

Participants were asked about the reasons that they decided not to provide sexual content, contact or services for financial compensation (check all that apply). Response options included: I don’t want to put the time or effort in; I have another job or a way to make money; I was not offered enough money; I was worried about being stigmatized; I was worried about being scammed or blackmailed; I felt dirty or ashamed; I didn’t want others such as family, friends, or intimate partners to know; I didn’t want current or potential employers or academic institutions to know; I was worried about the impact it would have on my self-esteem; I was worried about whether it was legal; I didn’t know how to start or get involved; something else (open-ended option).

Items to Establish Psychometric Properties

Convergent Validity

To establish convergent validity, we included an existing item from Blum et al. (2018), “Have you ever received compensation, financial or otherwise, in exchange for performing a sex act (including oral sex, penetrative sex, manual stimulation, or performance of sex acts for another person to view)?” (yes/no). Due to a survey formatting error, this item did not appear for the first 381 undergraduate and 38 graduate students who completed the initial survey. Among the 362 undergraduates who completed the follow-up survey, 92 did not receive this item; however, all graduate students who completed the two-week retest survey received the item.

Discriminant Validity

We used two items to establish discriminant validity. First, we asked students about a construct that we expected to be completely unrelated to sex trading; in this case, we asked about food preferences with the following response options: Italian, Chinese, Japanese, Mexican, American, Thai, Indian, or other [please specify]. Second, we used a brief validated version of the Marlow Crowne Social Desirability Scale (Reynolds, 1982) to control for concerns about reporting sensitive behaviors (α = .74).

Non-Sex Trading Measures

Demographics

Students were asked about year at university (undergraduates) or program type (graduate), full-time status, whether they considered themselves to be first generation college students, race, ethnicity, gender, sexuality, disability status, relationship status, how often they had difficulty paying for basic necessities, Pell grant status, and whether they had ever experienced homelessness, number of hours spent for working for pay, and adverse childhood experiences (Felitti et al., 1998). Graduate student surveys asked students to create a unique code only identifiable to them of their first and last initials and the last four digits of the cell phone.

Re-Test Survey

All students who completed the original survey were sent a follow-up survey approximately two weeks later. The re-test survey included the item to (1) assess convergent validity (Blum et al., 2018); (2) all sex trading items and their follow-up items (as described above); and (3) all items assessing whether students had ever considered (but not participated in) sex trading. Like the original survey, only students who endorsed sex trading received relevant follow-up sex items and only those who did not endorse any sex trading received items assessing whether they had ever considered sex trading. Students who reported having considered sex trading also received follow-up items.

Data Collection

Undergraduate Students

Undergraduate students were sampled through the SONA Systems,® a cloud-based research and participant management solution for universities. Students who are enrolled in psychology courses earn course credit for participating in research studies via SONA. Therefore, the subject pool includes predominantly first- and second-year students from psychology and non-psychology majors (e.g., Engineering, Business, etc.) who are enrolled in introductory courses. Therefore, the subject pool is typically reflective of first-year students across campus. Since students receive academic credit for participating in research, no monetary compensation was provided. In the Spring 2023, when data collection took place, ~1,000 students were eligible to earn up to 15 extra credits by completing research studies. Of those, 548 students began the initial survey and were prompted to complete the second survey through SONA Systems two weeks later. Of those, 362 completed the re-test survey. Because undergraduate students received academic credit, no monetary compensation was provided.

Graduate Students

Graduate students were invited to complete the anonymous survey via e-mail. We obtained the university mailing list for social and miscellaneous e-mails and selected 3000 random e-mail addresses. Graduate students who completed the initial survey were asked to provide a unique code involving initials and last four digits of phone number known only to them. Participants of the original survey were sent a follow-up invitation two weeks later, which required them to input their code in order for the research team to match their data. Both surveys ended by routing students to another survey to provide their information for compensation (e.g., first name or initials, preferred e-mail, and choice of Amazon, Target, or Walmart). We provided a $20 gift card for completing the first survey and $10 for the second.

In total, 413 participants accessed the original graduate survey. Of these, 66 completed between 1–5% of the survey. Eight participants completed between 6% and 12% of the survey. Of these, two participants responded “no” to the Blum sex trading question and the other 6 did not answer. Only one participant answered any of the sex trading questions. This participant answered only the first question about posting photos online and responded “occasionally.”

Among the remaining participant responses, those who completed 100% of the survey were compared with those who completed less than 100% of the survey across demographic variables. We found no statistically significant differences in demographics.

Data Analysis

Validity Analyses

Cross-tabulations with McNemar’s tests were used to assess convergent validity between cases of sex trading identified by a prior single item and this current expanded index. McNemar’s test is appropriate for paired nominal data (Fagerland et al., 2013). We also examined how many cases are identified by only one item to examine the hypothesis that the current index will identify more cases of sex trading. Cross-tabulations with Pearson’s chi square tests will be used to assess discriminant validity between sex trading cases identified by a prior single item and this current expanded index. Phi (φ) was calculated to assess the strength of relations for all cross-tabulations.

Reliability Analyses

Cronbach’s alpha was used to assess internal consistency reliability of the index items and the motivations items. Alpha values from .7 to .9 have been considered acceptable to good (Tavakol & Dennick, 2011). Cohen’s weighted kappa was used to assess test-retest reliability for categorical data measuring the number of respondents classified as engaging in sex trading at time 1 and time 2 (two weeks later). Values ≤ 0 indicate no agreement, 0.01–0.20 indicate none to slight, 0.21–0.40 indicate fair, 0.41–0.60 indicate moderate, 0.61–0.80 indicate substantial, and 0.81–1.00 indicate almost perfect agreement (Cohen, 1960).

Exploratory Factor Analysis

Principal Axis Factoring with varimax rotation was used to examine the factor structure of the index in a combined sample of undergraduate and graduate students (to maximize power). Eigenvalues of at least one, the points included in the steep slope of the scree plot, and the variance accounted for were used to determine the number of factors that should be extracted. We report item communalities, factor loadings, and residual correlations for the selected factor structure.

Prevalence and Reliability of Sex Trading Circumstances

We computed the prevalence, internal consistency, and test-retest reliability of the three major categories of sex trading circumstances: empowerment, exploitation, and financial need as well as the item about reclaiming one’s body after violence/abuse.

Prevalence and Reliability of Considering Sex Trading

We computed the prevalence, internal consistency, and test-retest reliability of the nine items about ever considering sex trading.

Results

Sample Characteristics

Table 1 summarizes sample demographics for undergraduates and graduates at time 1.

Table 1. Participant Demographics and Characteristics.

Characteristic	Undergraduate (N=544)	%	Graduate (N=331)	%
Year in School
First year	435	80.0%
Second year	77	14.2%
Third year	24	4.4%
Fourth year and beyond	8	1.5%
Master’s			158	47.7%
PhD			162	48.9%
Another graduate program			11	3.3%
Full-Time Student	539	99.1%	289	87.3%
First-Generation Student
Yes	100	18.4%	93	28.1%
No	432	79.4%	232	70.1%
Unsure	12	2.2%	6	1.8%
	(N = 544)		(N =320)
Race
White	362	66.5%	199	62.2%
Black/African-American	22	4.0%	5	1.6%
American Indian/Alaskan	8	1.5%	6	1.9%
Asian	102	18.8%	86	26.9%
Multiracial or another race	48	8.8%	24	7.5%
Hispanic/Latino/a/x	64	11.8%	21	6.3%
Gender
Woman	320	58.8%	182	56.9%
Man	216	39.7%	115	35.9%
Nonbinary/Gender Queer	7	1.3%	23	7.2%
Sexual Orientation
Heterosexual	456	84.8%	206	62.2%
Bisexual	45	8.3%	49	14.8%
Gay/Lesbian	6	1.1%	11	3.3%
Pansexual, Queer, Other	34	6.3%	51	15.4%
Relationship Status
Single	370	68.1%	118	37.0%
Married/Partnered	173	31.9%	201	63.0%
Have a Disability	27	5.0%	31	9.7%
Pell Grant Recipient	116	21.3%	66	19.9%
Experienced Homelessness	10	1.8%	10	3.1%
Difficulty Paying for Basic Needs
Never	237	43.6%	172	52.0%
Sometimes	104	19.1%	116	35.0%
Often/Always	21	3.9%	32	9.7%

Note: Any demographic characteristic with less than five participants was not included in this table. *only 320 participants completed race, ethnicity, gender, sexual orientation, relationship status, disability status, Pell grant recipient status, homelessness, and difficulty paying for basic needs in the graduate sample because these items were asked after the sex trading items and some participants had dropped out of the survey at that point.

Undergraduate Sample Baseline

Among the 548 undergraduate students who began the survey, two stopped after providing their age and two others stopped before completing any of the screening items. The final analytic sample (N = 544) was predominantly (80%) first-year students with a mean age of 18.96 (SD = 1.28). Most were full-time students and nearly 1 in 5 considered themselves first generation students. Most (66.4%) were White and 11.7% identified as Hispanic or Latinx. Most (58.7%) were women but 40% were men and 1.3% were non-binary, pangender, or gender expansive. About 84% were straight and most (67.9%) were single but nearly one-third (31.7%) reported being in a monogamous partnership. About 5% were disabled. Most (65.7%) had never had difficulty paying for basic necessities since coming to college but more than 1 in 5 students (21.3%) had received a Pell grant, indicating significant family financial need, and 1.8% had experienced homelessness.

Undergraduate Sample Retest

Approximately 360 undergraduate students (66% of the baseline sample) completed the follow-up survey two weeks later. Those who participated in the follow-up did not differ from those who did not on age, F(1, 544) = 0.22, p = .64, race, χ² (5,542) = 9.44, p = .09, gender, χ² (2,543) = 4.19, p = .12, sexual orientation, χ² (3,541) = 1.69, p = .64, relationship status, χ² (1,543) = 2.87, p = .09, Pell grant status, χ² (1,543) = 1.58, p = .21, first generation status, χ² (1,545) = 0.61, p = .74, difficulty paying for basic needs, F (1, 542) = 0.12, p = .73, or social desirability, F (1, 535) = 0.31, p = .58.

Graduate Sample Baseline

Among the 347 graduate students who consented and began the survey, 8 respondents indicated that they were not graduate students and were screened out, one respondent stopped after providing their age and seven other respondents stopped after providing a few basic demographics but before they reached the sex trading screening items. The final analytic sample (N = 331) were 26.7 (SD = 4.5; range = 18–54) years old on average; about half (47.7%) were Master’s students and half (48.9%) were PhD students. About 87% were full-time students and 28% considered themselves first-generation college students. Only 320 of the 331 participants provided the remaining demographics. Of the 320, most (62.2%) were white and 6.3% were Hispanic or Latinx. Most (56.9%) were women, but more than one-third (35.9%) were men, and 7.2% were non-binary, pangender, or another gender. Nearly two-thirds (62.2%) were straight but 14.8% were bisexual and 15.4% were pansexual, queer, or another sexual identity. More than one-third (37%) were single and 63% were married or partnered. Nearly 10% had a disability. More than half (52%) had never had difficulty paying for basic needs during graduate school. Approximately 20% had been a Pell grant recipient and 3.1% had experienced homelessness in their lifetime.

Graduate Sample Retest

Approximately 187 graduate students (56.5% of the baseline sample) completed the retest survey two weeks after the baseline. Those who participated in the follow-up did not differ from those who did not on age, F(1, 330) = .26, p = .61, race, χ² (5,320) = 5.72, p = .34, Hispanic/Latinx identity, χ² (1,320) = .13, p = .72, gender, χ² (2,320) = .80, p = .67, sexual orientation, χ² (3,320) = .95, p = .81, relationship status, χ² (1,319) = .04, p = .85, Pell grant status, χ² (1,320) = 1.04, p = .31, first generation status, χ² (1,331) = .48, p = .79, or difficulty paying for basic needs, F (1, 319) = 0.02, p = .90, disability, χ² (1,320) = 2.33, p = .13, or social desirability, F (1, 312) = 3.4, p = .07

Prevalence of Sex Trading

Table 2 summarizes sex trading types, items, and prevalence rates.

Table 2. Sex trading, type, items and prevalence rates of undergraduate and graduate students who report participating in the sex trades at baseline and follow-up.

Type	Sex Trading Item	Time 1		Time2
		Undergrad (N = 544)	Grad (N = 331)	Undergrad (N = 360)	Grad (N = 187)
Materials
	Sold sexual items or fluid	4.0% (22)	3.0% (10)	1.4% (5)	2.1% (4)
Virtual	Sent or posted nude photos	3.1% (17)	9.1% (30)	3.6% (13)	10.7% (20)
	Sexted/phone sex	3.5% (19)	6.0% (20)	1.9% (7)	6.4% (12)
	Online sexual companionship	0.6% (3)	1.8% (6)	0.8% (3)	2.1% (4)
	Filmed yourself online	0.7% (4)	2.7% (9)	1.5% (4)	1.1% (2)
	Any virtual item	5.5% (30)	11.2% (37)	6.6% (24)	13.4% (25)
In-Person
	Danced or stripped	0.4% (2)	0.9% (3)	0.6% (2)	1.1% (2)
	In-person date or sugaring	2.0% (11)	8.8% (29)	0.6% (2)	7.0% (13)
	Sexual contact, no intercourse	0.7% (4)	3.3% (11)	0.8% (2)	4.8% (9)
	Oral, vaginal, anal intercourse	0.9% (5)	4.2% (14)	0.8% (3)	3.7% (7)
	Any in-person	3.3% (18)	11.8% (38)	2.5% (9)	10.7% (20)
Any
	Endorsed any item	11.0% (60)	18.1% (60)	7.7% (28)	17.6% (33)

Undergraduate Sample

At baseline, 11% reported engaging in any sex trading, with 5.5% endorsing at least one virtual sex trading item and 3.3% endorsing at least one in-person sex trading item. The most commonly endorsed virtual item was sexting or phone sex (3.5%), while the most commonly endorsed in-person item was sugaring or providing in-person sexual companionship (2%). Among the 360 undergraduates who completed the two-week follow-up survey, the prevalence of any sex trading was 7.8%, with 5.8% reporting any virtual sex trades and 2.5% endorsing any in-person sex trades.

Graduate Sample

Among graduate students (M age = 26.5, range = 18–54), 18.1% reported engaging in any sex trading, with 11.2% endorsing at least one virtual sex trading item and 11.8% endorsing at least one in-person sex trading item. The most commonly endorsed virtual item was sending or posting photos (9.1%), while the most commonly endorsed in-person item was sugaring (8.8%). Among the 187 graduate students who completed the two-week follow-up survey, 17.6% reported any sex trading, with 12.8% reporting any virtual sex trades and 10.7% reporting any in-person sex trades.

Convergent Validity

Table 3 compares estimates from the SSAMP and the single sex trading item used in Blum et al. (2018).

Table 3. Comparison between single item and SAAMP index.

	Any Sex Trading			Virtual Sex Trading			In-Person Sex Trading
Blum Item*	Yes	No	φ	Yes	No	φ	Yes	No	φ
Undergraduates Time 1 (N = 167)
Yes	4	1		2	2		2	3
No	13	148	0.14	9	157	0.34	4	157	0.34
Undergraduates Time 2 (N = 270)
Yes	4	1		3	2		3	2
No	18	246	0.36	14	250	0.30	4	260	0.50
Graduates Time 1 (N = 293)
Yes	14	1		12	3		10	5
No	31	243	0.05	15	259	0.57	20	254	0.43
Graduates Time 2 (N = 187)
Yes	17	1		13	5		15	3
No	15	152	0.67	11	156	0.58	5	162	0.77

Blum item was only given to 167 undergraduate students at Time 1.

“Any Sex Trading” “Virtual Sex Trading” and “In-Person Sex Trading” are taken from SAAMP.

Undergraduate Sample-Baseline

Among the 167 students who completed the Blum sex trading item and the expanded index at Time 1, evidence of convergent validity was observed (McNemar’s p < .001, φ = .41) such that both measures were significantly positively associated, and the strength of the association was moderate. Among the 5 students who reported sex trading according to Blum, 4 (80%) also endorsed sex trading on SSAMP. However, 13 additional students endorsed an item on the SSAMP but said “no” to the Blum item, confirming the first hypothesis that more cases would be identified by the expanded index. There also was evidence of convergent validity between the Blum item and both virtual (McNemar’s test p = .004, φ = .49) and in-person (McNemar’s p < .001; φ = .34;) sex trading.

Undergraduate Sample-Follow-Up

Among the 270 students who completed the Blum sex trading item and the expanded index at Time 2, evidence of convergent validity was observed (McNemar’s test < .001; φ = .36). Among the 5 students who reported sex trading according to Blum, 4 (80%) also endorsed sex trading on the SSAMP. However, 18 additional students endorsed an item on the SSAMP but said “no” to the Blum item, again confirming the first hypothesis.

Graduate Sample-Baseline

Among the 289 students who completed the Blum sex trading item and the expanded index at Time 1, evidence of convergent validity was observed (McNemar’s p < .001; φ = .50) such that both measures were significantly positively associated, and the strength of the association was strong. Among the 15 students who reported sex trading according to Blum, 14 (93.3%) also endorsed sex trading on the SSAMP. However, an additional 31 graduate students were identified as endorsing sex trading on the SSAMP and said “no” to the Blum item.

Graduate Sample Follow-Up

Among the 183 students who completed the Blum sex trading item and the expanded index at Time 2, evidence of convergent validity was observed (McNemar’s p = .001; φ = .66) and the strength of the association was strong. Among the 17 students who reported sex trading according to Blum, 16 (94%) also endorsed sex trading on the SSAMP. However, an additional 15 graduate students were identified as endorsing sex trading on the SSAMP but said “no” to the Blum item.

Discriminant Validity

Undergraduate Sample

Food preference was not significantly associated with any sex trading (χ² (9,544) = 11.43, p = .24) or in-person sex trading (χ² (9,544) = 14.01, p = .12), but it was significantly associated with virtual sex trading (χ² (9,544) = 17.19, p = .046) among undergraduate students. Marlowe-Crowne social desirability scores also did not differ between those with and without sex trading experiences, F(1,535) = 2.98, p = .09, nor did it differ for those with virtual, F(1,535) = .13, p = .72, or in-person, F(1,535) = 0.20, p = .65, sex trading.

Graduate Sample

Food preference also was not significantly associated with any sex trading (χ² (8,331) = 4.51, p = .81) or with virtual (χ² (8,331) = 7.44, p = .49) or in-person (χ² (8,331) = 6.24, p = .62) sex trading among graduate students. Marlowe-Crowne social desirability scores also did not differ between those with and without sex trading experiences, F(1,312) = 1.38, p = .24, nor did it differ for those with virtual, F(1,312) =,63, p = .43, or in-person, F(1,312) = 3.37, p = .07, sex trading.

Internal Consistency and Test-Retest Reliability of Sex Trading

Undergraduate Sample

The SSAMP had good internal consistency at baseline (Cronbach’s α = .88), with the four virtual items having an internal consistency of .73 and the four in-person items having an internal consistency of .79. At two-week follow-up, the internal consistency for the sex trade items was .95, with the four virtual items having an internal consistency of .85 and the four in-person items having an internal consistency of .92. The sex trade items had moderate 2-week test-rest reliability (Cohen’s weighted kappa = .51).

Graduate Sample

The SSAMP had adequate to good internal consistency at baseline (α =.87), with the four virtual items having an internal consistency of .68 and the four in-person items having an internal consistency of .75. At two-week follow-up, internal consistency was .73, with the four virtual items having an internal consistency of .68 and the four in-person items having an internal consistency of .64. Substantial 2-week test-retest reliability (Cohen’s weighted kappa = .71) was observed in the graduate sample.

Preliminary Factor Structure of Sex Trading

Using a combined undergraduate and graduate sample (N = 875), we conducted Principal Axis Factoring with varimax rotation. The Kaiser-Meyer-Olkin test value was .90, suggesting the data were suited for factor analyses and Bartlett’s test of sphericity was significant, X2 (df = 36) = 4259.57, p < .001, suggesting that the correlation matrix contained non-zero values. Based on eigenvalues > 1 and the scree plot, we extracted a single factor that accounted for 54.58% of the variance in sex trading. Despite low (<.6) communalities for items 1,2,3, and 9, each item loaded on the single factor at .59 or higher (see Table 4). All items had at least one residual correlation over .05.

Table 4. Item communalities, factor loadings, and residual correlations for exploratory factor analysis.

			Residual Correlations
Item	Communalities	Factor Loadings	2	3	4	5	6	7	8	9
1.Sold sexual items or fluid	.35	.59	.049	.044	.08	−.09	−.16	−.09	−.16	−.16
2.Sent or posted nude photos	.38	.62	–	−.05	−.06	−.04	−.04	−.13	−.10	−.13
3.Sexted/phone sex	.37	.61		–	−.07	−.08	−.13	−.05	−.14	.02
4.Online sexual companionship	.69	.83			–	−.01	−.03	−.09	−.06	−.11
5.Filmed yourself online	.82	.91				–	.001	−.04	.06	−.07
6.Danced or stripped	.57	.75					–	−.07	.006	−.10
7.In-person date or sugaring	.59	.77						–	.008	.01
8.Sexual contact	.73	.85							–	−.006
9.Intercourse	.41	.64								–

Prevalence and Reliability of Circumstances Reported for Any Sex Trading

Undergraduate

Of the 60 students who reported any sex trading on the SSAMP at baseline, 25% reported financial circumstances, 31.7% reported exploitation, and 48.3% reported empowerment; 18% also endorsed the item “to reclaim my body after violence” and 35% endorsed the item “to please someone like a friend or dating partner.” The reliabilities for financial circumstances, exploitation, and empowerment were .89, .78, and .67, respectively. Of the 28 students who reported sex trading at follow-up, 32.1% reported financial circumstances, 35.7% reported exploitation, and 39.3% reported empowerment, with 32.1% reporting “to please someone like a friend or dating partner” and 21.4% reporting “to reclaim my body after violence.”

Graduate

Of the 60 students who reported any sex trading on the SSAMP at baseline, 41.7% reported financial circumstances, 48.3% reported exploitation, and 50% reported empowerment; the reliabilities for financial circumstances, exploitation, and empowerment were .75, .94, and .80, respectively; 13.3% also endorsed the item “to reclaim my body after violence” and 35% endorsed the item “to please someone like a friend or dating partner.” Of the 33 students who reported sex trading at follow-up, 66.7%, 36.4%, and 60.6% reported financial circumstance, exploitation, and empowerment, respectively; 12.1% also endorsed the item “to reclaim my body after violence” and 33.3% endorsed the item “to please someone like a friend or dating partner.”

Prevalence and Reliability of Considering Sex Trading, Reasons for Considering, and Reasons for Deciding Not To

Table 5 summarizes type, items, and prevalence rates of those who had considered sex trading. Table 6 summarizes the reasons why students who considered engaging in the sex trades chose not to do so.

Table 5. Sex trading, type, items and prevalence rates of undergraduate and graduate students who reported considering sex trades at baseline and follow-up.

Type	Sex Trading Item	Time 1		Time 2
		Undergrad (N = 484)	Grad (N = 271)	Undergrad (N = 332)	Grad (N = 154)
Materials
	Sold sexual items or fluid	12.2% (59)	29.5% (80)	9.9% (33)	28.6% (44)
Virtual
	Sent or posted nude photos	19.4% (94)	32.5% (88)	15.4% (51)	35.5% (55)
	Sexted/phone sex	13.8% (67)	23.6% (64)	12.0% (40)	21.9% (34)
	Online sexual companionship	4.1% (20)	16.2% (44)	3.6% (12)	18.2% (28)
	Filmed yourself online	9.1% (44)	15.1% (41)	9.3% (31)	16.1% (25)
	Any virtual item	26.0% (126)	40.6% (110)	21.1% (70)	41.3% (64)
In-Person
	Danced or stripped	13.8% (67)	11.8% (32)	11.4% (38)	11.7% (18)
	In-person date or sugaring	14.6% (71)	25.1% (68)	12.0% (40)	31.6% (49)
	Sexual contact	11.6% (56)	12.2% (33)	9.0% (30)	12.9% (20)
	Intercourse	10.1% (49)	10.7% (29)	10.2% (34)	11.0% (17)
	Any in-person	29.5% (143)	31.4% (85)	25.3% (84)	39.4% (61)
Any
	Considered any item	38.8% (188)	48.7% (132)	33.4% (111)	51.0% (79)

All questions asked participants to respond to a Likert scale with the following response options: never, occasionally, sometimes, often, very often.

Table 6. Reasons for deciding not to engage in sex trading among undergraduate and graduate students who considered sex trading.

Item	Undergraduate (N = 188)	Graduate (N = 132)
Did not want to put in time or effort	52.1%	43.9%
Had another job or way to make money	58.0%	50.0%
Was not offered enough money	6.4%	6.1%
Worried about stigma	34.6%	54.5%
Worried about being scammed or blackmailed	52.1%	56.8%
Felt dirty or ashamed	58.5%	40.9%
Did not want friends, family, partners to know	63.3%	68.9%
Did not want employers or schools to know	53.2%	64.4%
Worried about impact on self-esteem	39.9%	27.3%
Worried about legality	33.0%	35.6%
Did not know how to start	27.7%	48.5%
Another reason	2.2%	7.6%

Undergraduate Sample

Among the 484 students who reported never engaging in sex trading at baseline, 34.6% (n = 188) had ever considered engaging in sex trading, with 26% (n = 126) reporting they had considered virtual acts and 29.5% (n = 143) reporting they had considered in-person acts. At follow-up, 33.4% (n = 111) considered any sex trading, with 21.1% reporting that they had considered any virtual sex trading and 25.3% reporting that they had considered any in-person sex trading. These items had good internal consistency at baseline (α = .81) and follow-up (α =.76). Among the 188 students who had ever considered engaging in sex trading at baseline, 64.4% reported financial reasons for considering sex trading; 22.9% reported that pressure or force were a factor in why they considered sex trading; and 43.6% reported empowerment reasons for considering. Among the 111 students who reported considering sex trading at follow-up, 44.1% reported financial reasons, 12.6% reported pressure or force reasons, and 18.9% reported empowerment reasons for considering sex trading. The most common reason for deciding not to trade sex was not wanting family, friends, or intimate partners to know (Table 5).

Graduate Sample

Among the 271 students who reported never trading sex at baseline and thus were asked whether they had ever considered trading sex, 48.7% (n = 132) reported that they had considered engaging in at least one sex trading behavior and 51% of the 154 who had never traded sex at two-week follow-up reported that they had ever considered trading sex. These items had good internal consistency at baseline (α = .88) and follow-up (α =.87). Among those who had considered sex trading at baseline, 90.5% reported financial circumstances, 12.4% reported exploitation, and 63.5% reported empowerment. Among the 80 who had considered any sex trading at follow-up, 66.3% reported financial circumstances, 8.8% reported exploitation, and 27.5% reported empowerment circumstances. As with the undergraduate sample, the most common reason for deciding not to trade sex was not wanting family, friends, or intimate partners to know (Table 5).

Discussion

Our study suggests that using a multi-item index is critical to understanding the scope and complexity of sex trading among university students and, likely, for other populations as well. We found strong convergent validity between a single item and the SSAMP, and as expected, the SSAMP identified more cases of sex trading, including more cases of virtual and in-person sex trading than the single item. Prior literature using a single item suggested 2 to 4.5% of undergraduate students report sex trading for financial compensation (Blum et al., 2018; Gerassi, Lowe, et al., 2023). However, our study found 11% of undergraduates (predominantly 1^st and 2^nd year students) and 18.1% of graduate students report at least one type of sex trading. This increase is likely in part due to the integration of virtual acts, which account for more than half of reports by undergraduate and graduate students. However, the higher prevalence estimates are likely due to the methodological benefits of using multiple items to understand highly stigmatized phenomena (Cook et al., 2011). It is also important to note that a key framing of the SSAMP was to normalize and better understand the wide range of circumstances under which the sex trades occur. Research in this field occurs within the context of extensive feminist and legal debates as to the legitimacy of adult sex trades (Nichols, 2016). We do not yet know if this framing yields higher endorsements of the sex trades. However, our study does suggest that research using a multi-item index may capture diverse sexual acts, services, and materials that are not yet reflected in most survey studies. Using measures developed with feedback from the communities they purport to assess, such as the SSAMP for university students in similar contexts, is crucial to capturing nuanced data needed for better interventions and policies.

Our study also highlights that the SSAMP can be used to reliably assess sex trading among undergraduate and graduate students, has acceptable to good test-retest reliability, and strong discriminant validity. Of note, and consistent with other work on the sex trades with university students in the UK (Sagar et al., 2015, 2016), and qualitative work in the US (Gerassi, Zimmerman, et al., 2023), the item about selling sexual materials or fluids had good internal consistency reliability with the other items assessed here, making a strong case for this expanded definition of the sex trades.

Using a combined sample of undergraduate and graduate students, exploratory factor analysis revealed preliminary evidence that the SSAMP consists of a single factor with relatively strong loadings across all items. However, it should be noted that the single factor only accounted for slightly more than half of the variance (~55%) in sex trading and every variable had at least one residual correlation, suggesting additional work should be conducted in larger samples with more varied sex trading experiences. The sex trading index items do not address motivations for engaging in various sex trading behaviors. Literature regarding the 3 Cs suggests that there could be different circumstances driving sex trading behavior (Stoklosa & Ash, 2021; Weber, 2020), including both latent characteristics of the individual (e.g., sensation seeking) and factors entirely outside the control of the individual such as coercion or victimization. In the latter case, there is no latent characteristic of a person that could be expected to explain responses to items reflecting victimization experiences (e.g., Koss et al., 2024). Whether factor analysis is an appropriate technique for analyzing sex trading behaviors remains to be seen.

Limitations

There are several important strengths and limitations to consider. A small number of students in each sample (n = 60) reported any sex trading and some items were endorsed by only a few people; thus, preliminary factor analysis results may not hold up in larger samples with more varied sex trading experiences. Comparisons between our undergraduate and graduate samples should be done with caution, as data were obtained using different sampling strategies. On the one hand, the use of the SONA subject pool led to increased data quality (e.g., less missing data), which is critical in assessing the psychometric properties of an index. However, a substantial majority of students were first-year students; thus, findings may not generalize across the range of years in school, and we have little to no understanding of students whose majors do not require an introduction to psychology course, e.g., Humanities. A strength of the graduate sample was the use of students selected randomly from the population of students with registered e-mails. However, the risk for spam participants and decreased data quality is higher. Our team implemented multiple items to identify bots (e.g., open-ended items, Captcha) and examined original and re-test surveys for suspicious responses to mitigate this limitation (Griffin et al., 2022; Storozuk et al., 2020). Out of concern for survey length and participant burden, we chose not to ask students who endorsed one to eight of the nine sex trading items whether they had considered other acts (e.g., we did not ask students who reported engaging in virtual sex trading only whether they had also considered in-person forms). Additionally, two circumstances involved for the “considered” items may have been confusing for participants. The items asked students whether they had considered engaging in the sex trades because they had been “physically forced” and “pressured in a way they couldn’t say no,” to an act that they had only considered doing. Therefore, it is possible that the low endorsement of those circumstances reflects participant confusion rather than a lack of potential coercion/victimization. Due to a survey formatting error, only 31% of the baseline survey participants and 75% of the re-test participants in the undergraduate sample received the Blum convergent validity item. Finally, a strength and limitation of this study is that it took place at a PWI in a Midwestern city in the US. While the study revealed that students within a racially privileged university setting report trading sex, findings should not be generalized to other contexts. Future research should use random sampling to understand sex trading among diverse university settings (e.g., community colleges, private institutions) and diverse geographic contexts.

Conclusion

Our study underscores the need to expand our understanding of the sex trades using nuanced, rigorously developed, muti-item measures in diverse populations. This research is critically needed to inform interventions and policies that are reflective of the scope and range of sex trading experiences among university students and beyond. Future research should adapt and use the SSAMP in diverse student populations (e.g., community colleges, private institutions, public universities with differing regional contexts and demographics). These institutions should work with students from those contexts to revise the index according to the needs of localized student populations. Future research that refines, adapts, and uses the SSAMP to identify sex trading among diverse populations (beyond university contexts) is also crucial to informing localized and nationally representative studies. It is our hope that our study and future iterations of the SSAMP can serve as a critical step to understanding the complexity of this phenomenon and its motivations, benefits and consequences.

Acknowledgments

The authors would like to express their deep appreciation for the participants of this study as well as knowledgeable students who informed this work. Special thanks to Mia Warren, Jason Hill, Jelani Williams, and Jessica Melnick. We are grateful to have had a review process that significantly advanced our thinking on this manuscript. To that end, we thank our reviewers for their careful attention to this manuscript along with Drs. Mary Koss, RaeAnn Anderson, and David DiLillo for their input in our revised version. Finally, we sincerely thank Drs. Christian Grov and Cynthia Graham for their leadership and helpful guidance at differing stages of the review process.

Disclosure Statement

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

Additional information

Funding

Research reported in this publication was supported in part by the National Institute of Arthritis and Musculoskeletal and Skin Diseases, the Office of Research on Women’s Health, Building Interdisciplinary Research Careers in Women’s Health (BIRCWH) program, the Office of The Director, National Institutes of Health and the National Cancer Institute, under Award Number [8K12AR084227] (2020-2025). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The second author’s work was also supported by funding from the 4W Initiative at the University of Wisconsin-Madison.