Online supplement forCourtice, E. L., & Shaughnessy, K. (2017). The partner context of sexual minority women’s and men’s cybersex experiences: Implications for the traditional sexual script. Sex Roles.Erin Leigh Courtice, University of Ottawa. Email:

Background

Shaughnessy and Byers (2013) proposed a multi-item measure of cybersex experience based on research with heterosexual participants. Single-item measures (for example, “Have you ever had cybersex?”) are criticized for being limited in their reliability, validity, and generalizability (Epstein, 1980) yet they remain prevalent in research on online sexual activity. In contrast, multi-item measures provide participants with multiple contextual examples of an activity as it might occur, and they inquire about instances of that activity (for example, “Have you ever behaved sexually for another person to watch?”). Multi-item measures often outperform single-item measures in terms of validity and reliability (Diamantopoulos, Sarstedt, Fuchs, Wilczynki, & Kaiser, 2012; Gardner, Cummings, Dunham, & Pierce, 1998; GliemGliem, 2003). Shaughnessy and Byers (2013) argued that their multi-item measure provided a more accurate—and thus valid—assessment of cybersex experience than did a single item measure. They had heterosexual participants complete two measures of cybersex experience: a single-item measure (the global measure of cybersex) and a multi-item measure (the lifetime cybersex experience questionnaire). They found that significantly more participants reported concordant (i.e., answered the same to both measures) than discordant responses (i.e., answered differently to the two measures). Of those who reported discordant responses, significantly more endorsed only the multi-item measure than only the single-item measure and these people were more likely to have had cybersex with a primary partner. They also provided evidence of reliability and temporal stability of the multi-item measure.

Previous research has identified differences between heterosexuals’ and sexual minority peoples’ online sexual activity experience (Daneback, Cooper, & Månsson, 2005; Daneback, Ross, & Månsson, 2008; Grov, Breslow, Newcomb, Rosenberger, & Baermeister, 2014). Therefore, sexual minority people may respond differently to the multi-item measure of cybersex than heterosexual people do. As such, we extended Shaughnessy and Byers’ (2013) examination of single- and multi-item measures of cybersex to assess whether or not similar results would be found in a sample of sexual minority people. To do so, we used Shaughnessy and Byers’ (2013) single-item and multi-item measures of cybersex experience. Participants gave a yes/no response to the question of whether or not they had ever had cybersex on the single-item global measure of cybersex. This item included a definition of cybersex, but the partner context was not specified. These responses were compared to those provided on the lifetime cybersex experiences questionnaire (described in the primary article). We then examined whether participants’ reports were concordant or discordant. We also examined whether participants who completed both Study 1 and Study 2 (see procedure section, full manuscript) maintained or changed their responses between the first and second completions of these measures.

Results and Discussion

We compared the percentage of people who endorsed cybersex on the single-item versus the multi-item measures at only Time 1. More participants endorsed the multi-item measure (Multi-Cyber; 98.8%) relative to the single-item measure (Single-Cyber; 87.1%). To determine the rates of concordance and discordance between the two measures, we created three groups. Most participants (85.9%) were in the Concordant Group, followed by the Discordant Multi-Only Group (12.9%) and the Discordant Single-Only Group (1.2%). These findings suggest that there was good validity between the single-item and multi-item measure because the majority of the sample reported concordant results.

To examine the extent to which participants changed their responses to the Single/Multi-Cyber after being primed with the Multi-Cyber question, we compared responses between Time 1 and Time 2 for the 67 participants who completed measures a second time. Of these 67 participants, 29.0% were from the Concordant Group (at Time 1) and 15.6% from the Discordant Multi-Only Group. (None of the participants from the Discordant Single-Only Group completed measures at Time 2.) Of participants who endorsed Single-Cyber at Time 1, 96.8% endorsed it at Time 2. Similarly, 88.1% of those who endorsed Multi-Cyber did so again at Time 2. We then assessed the temporal reliability in providing concordant versus discordant results. First, we computed a kappa statistic for Concordant and Discordant Multi-Only Group membership; there was significant agreement between whether participants were in the Concordant or Discordant Multi-Only Group at Time 1 and Time 2, k = .212, p = .010. Second, we used cross-tabs to examine changes in group membership between Time 1 and Time 2 (see Table 1s). Because of the small cell sizes, we did not compute a chi-square statistic. The majority of participants retained their concordant or discordant responses over time, and the majority of participants at Time 2 remained in the group in which they were at Time 1. However, some participants (12.9%) who were concordant at Time 1 became discordant at Time 2. Together, these findings indicate fairly high temporal reliability for both the global measure of cybersex and the lifetime cybersex experiences questionnaire.

Table 1s

Number (and Percentage) of Participants in Concordant and Discordant Groups at Time 2 Sorted by Group Membership at Time 1

Time 2 group
Time 1 group / Concordant
n(%) / Discordant multi-only
n(%) / Discordant single-only
n(%)
Concordant (n = 62) / 54 (87.1%) / 2 (3.2%) / 6 (9.7%)
Discordant multi-only (n = 5) / 3 (60.0%) / 2 (40.0%) / 0 (0.0%)

Conclusion

Overall, these results are consistent with Shaughnessy and Byers’ (2013) findings. Specifically, we found that significantly more participants reported concordant rather than discordant responses. Additionally, of those who reported discordant responses, more endorsed only the multi-item measure than only the single-item measure. Thus, our findings suggest that the lifetime cybersex experiences questionnaire is a reliable and valid measure of cybersex experience for sexual minority participants.

References

Daneback, K., Cooper, A., & Månsson, S.-A. (2005). Aninternet study of cybersex participants. Archives of Sexual Behavior, 34(3), 321–328. doi: 10.1007/s10508-005-3120-z

Daneback, K., Ross, M. W., & Månsson, S.-A. (2008). Bisexuality and sexually related activities on the internet.Journal of Bisexuality, 8(1-2), 115–129. doi: 10.1080/15299710802142317

Diamantopoulos, A., Sarstedt, M., Fuchs, C., Wilczynski, P., & Kaiser, S. (2012). Guidelines for choosing between multi-item and single-item scales for construct measurement: a predictive validity perspective. Journal of the Academy of Marketing Science, 40(3), 434–449. doi:10.1007/s11747-011-0300-3

Epstein, S. (1980). The stability of behavior: II. Implications for psychological research.American Psychologist, 35(9), 790–806. doi:10.1037/0003-066X.35.9.790

Gardner, D., Cummings, L., Dunham, R., & Pierce, J. (1998). Single-item versus multiple-item measurement scales: An empirical comparison. Educational and Psychological Measurement, 58(6), 898–915. doi: 10.1177/0013164498058006003

Gliem, R. R., & Gliem, J. A. (2003). Calculating, interpreting, and reporting Cronbach’s Alpha reliability coefficient for Likert-type scales. Retrieved from

Grov, C., Breslow, A. S., Newcomb, M. E., Rosenberger, J. G., & Bauermeister, J. A. (2014). Gay and bisexual men’s use of the internet: Research from the 1990s through 2013. The Journal of Sex Research, 51(4), 390–409. doi: 10.1080/00224499.2013.871626

Shaughnessy, K., & Byers, E. S. (2013).Seeing the forest with the trees: Cybersex as a case study of single-item versus multi-item measures of sexual behaviour.Canadian Journal of Behavioural Science, 45(3), 220–229. doi: 10.1037/a0031331