Gay and Bisexual men's use of the Internet: Research from the 1990s through 2013

Abstract

In this review, we document the historical and cultural shifts in how gay and bisexual men have used the Internet for sexuality between the 1990s and 2013. Over that time, gay and bisexual men have rapidly taken to using the Internet for sexual purposes: sexual health information seeking, finding sex partners, dating, cybersex, and pornography. Gay and bisexual men have adapted to the ever-evolving technological advances that have been made in connecting users to the Internet—from logging into the World Wide Web via dial-up modem on a desktop computer to geo-social and sexual networking via a handheld device. In kind, researchers too have adapted to the Internet to study gay and bisexual men, though not at the same rapid pace at which technology (and its users) have advanced. Studies have carefully considered the ethics, feasibility, and acceptability of using the Internet to conduct research and interventions with gay and bisexual men. Much of this work has been grounded in models of disease prevention, largely as a result of the ongoing HIV/AIDS epidemic. The urgent need to reduce HIV in this population has been a driving force to develop innovative research and Internet-based intervention methodologies. Moving forward, a more holistic understanding of gay and bisexual men's sexual behavior might be warranted to address continued HIV and STI disparities. The Internet, and specifically mobile technology, is an environment gay and bisexual men are using for sexual purposes. These innovative technologies represent powerful resources for researchers to study and provide rapidly evolving outreach to gay and bisexual men.

Introduction

There is a saying that new technologies, regardless of their intended purposes, are quickly adopted for sex. Telephones gave rise phone sex, the VCR and VHS tapes gave rise to viewing pornography in one's home (versus visiting an adult cinema), hand-held video cameras begot the amateur porn industry, and text messaging begot ‘sexting.’ Thus it is perhaps unsurprising that the Internet has been widely adopted for sexual content. Users engage in cam-to-cam video chatting, streaming/downloading pornographic content, exchanging/broadcasting nude photos and videos, cybersex, dating, shopping for sex toys, escorting, etc.

In this paper, we attempt to document the historical and cultural shifts in how gay and bisexual men have used the Internet for sexuality. Our goal was to build this understanding through the lens of research—exploring shifting paradigms born of methodological advancements. In developing this narrative, we had to make painful decisions about what exactly to cover, how far back to go, and how best to organize this manuscript. Because technological advances brought about shifts in research theory and strategy, we developed a narrative that followed a linear sequence, starting in the 1990s and continuing through the time we finished the manuscript in summer 2013. The four sections include the 1990s, 2000-2004, 2005-2009, and 2010 to present. These designations are one part in the interest of parsimony, and, to some extent, represent real transitions in Internet use. Certainly, this is not to suggest that the year 2000 was somehow entirely different from 1999. Yet, the 1990s (as a decade) were characteristically different from the early 2000s, both technologically and culturally. Similarly, the rapid changes in Internet use throughout the first decade of the 21^st century (including the expansions of high-speed broadband, Wi-Fi, and user-driven content (i.e., ‘web 2.0’)), distinguish the second half of the decade from the first. Meanwhile, the rapid and recent transitions to mobile platforms as a means to access the Internet distinguish Internet use in recent years (and thus more contemporary online scientific inquiry) from those of the previous decade.

To the best of our ability, we have attempted to capture the essence of Internet use by gay and bisexual men and online research among this particular group of users during each of those time periods. Due to publication lags (the delay between when data were collected and when findings ultimately appeared in print), the dates for some of the studies cited in a particular section may appear out of context. For this narrative, we have attempted to characterize research based on when it was conducted, not when it was published. In so doing, we couch our narrative appropriately within the technologies available and cultural milieu of that time. Also, we recognize that there are strengths and limitations to adopting a linear approach. To some extent, readers may notice repetition of themes (e.g., over each time point more and more gay and bisexual men are utilizing the Internet for sexual purposes). In such instances, we attempted to focus our narrative not on the repetition of content, but rather how research methodology, scientific questions, and paradigms have shifted over time. As readers will note, in some instances, there has been little change (e.g., the prioritization of HIV/AIDS prevention and education is present in virtually all research across the last two decades).

Our goal is not to present every article on gay men's use of the Internet from the last three decades, but rather to describe major shifts in innovation, culture, behavior, and research. Much of our attention is given to research from industrialized nations; however, this is not to suggest that research from industrializing nations is less important. The roles that economic and political dissimilarities between industrialized and industrializing nations have played in gay and bisexual men's use of the Internet warrants its own unique investigation. We omitted extensive discussion of the role that the Internet has played in male-for-male escorting. Yet, we recognize that the changes that the Internet has had on escorting have been profound (Parsons, Koken, & Bimbi, 2004; Smith & Grov, 2011; Walby, 2012). So much so, that this topic has received independent focus in an array of sources, including the 2013 issue of the Annual Review of Sex Research (Minichiello, Scott, & Callander, 2013).

Finally, in this manuscript, we refer heavily to gay and bisexual men (meaning men who have adopted an identity as gay or bisexual). At times we may refer to ‘men who have sex with men (MSM),’ which includes gay and bisexual men, but may also include heterosexually-identified MSM and those who have sex with men but do not identify as gay or bisexual. For this manuscript, gay and bisexual men are the targeted population of inquiry, but ‘MSM’ is used in places where we could not determine the sexual identity of all participants in a research study to be gay or bisexual.

The Internet and Sex in the 1990s

The modernization and commercial availability of the Internet in the 1990s revolutionized the ways gay and bisexual men fostered community and connected with sex partners. Though slow speeds over dial-up (i.e., telephone lines) made it impossible for users to stream video pornography or engage in face-to-face video chatting (as is common today), the Internet became a remote yet vibrant venue for gay and bisexual men to engage in political discussion and social support, and to post and respond to personal ads (Benotsch, Kalichman, & Cage, 2002; Friess, 1998). The Internet's evolution into a popular venue for gay and bisexual men was largely shepherded by America Online (AOL). Though in existence throughout the 1980s, AOL gained commercial success in the 1990s largely because of its ease of use and user-friendly graphic user interface. AOL required new users to create profile names that would serve as their identity on the site and to other users. Such names were often independent of a user's real name. For example, John Smith might create a profile name like “BeachFanatic01.” Profile names were often a way for users to convey something about themselves (or something they wanted others to think about them, or something they desired in others) in a manner that fostered anonymity. Thus, AOL developed an infrastructure that was ripe to foster anonymity.

In addition, AOL popularized ‘chat rooms’ in the 1990s: a virtual space in which members gather to post and respond to messages in real time through public online discussion forums (Tikkanen & Ross, 2000). These chat rooms provided an often anonymous venue for gay and bisexual men to discuss issues or to ‘cruise’ for partners with minimal concerns of outing, arrest, or violence (Mills, 1998; Shaw, 1997). AOL hosted its own chat rooms (on a range of interest-based topics such as gardening, parenting, sports), and permitted users to create their own public chat rooms.

Despite such innovative utilization of the Internet during the 1990s to meet new partners, data on the percentage of gay and bisexual men who were online during this time are unavailable. We know that, in 1995, 14% of U.S. adults used the Internet; this had risen to 36% by 1998 (Pew Internet and American Life Project, 2012). Gay and bisexual men were certainly frequenting gay-specific Internet sites and resources, however, evidenced by the use of new culture-bound terminology and modes of interaction. Vernacular such as ‘M4M’ (meaning men for men), for example, was quickly adopted such that AOL abounded with user-created chat rooms like ‘SeattleM4M,’ ‘MiamiM4M,’ ‘ChicagoM4M’ (Friess, 1998). These user-created chat rooms could hold up to about 100 users, and members could create additional rooms (e.g., MiamiM4M2) should the demand exist. Rooms existed as long as users remained in them. Once inside a room, users could post messages to the chat board, which were visible to all other members, and could browse profile information of other users in the room. Profile information typically included a description of the user (physical characteristics as well as likes/dislikes), but rarely a photo of the user or personally identifiable information (such as real name). Users could choose to have private conversations with others via ‘instant messaging’ or email, and add other gay and bisexual men to their ‘buddy lists.’ A user's buddy list was his personal/private list of other users (which could include friends, family, and sex partners). AOL would populate a user's buddy list with users who were currently online, thus facilitating instant interactions with other users.

Though AOL was the Internet Service Provider (ISP) that made the greatest headway in the 1990s throughout the United States, those looking for sex online (i.e., ‘cruising’) were not constrained to the chat environment housed by their ISP (i.e., within the AOL environment). Users were welcome to use their Internet browser to visit any number of websites that would cater to their interests; for gay and bisexual men in the 1990s, some examples included gay.com and planetout.com. The infrastructure on these websites was similar to that of AOL. Users would sign into their accounts to take advantage of chat rooms and message services (email or instant messages).

Without the widespread commercial availability of digital cameras or scanners, it was uncommon in the 1990s for users to have digital photos/videos of themselves that could be shared (emailed or posted as part of a profile), much less nude digital photos of themselves, which are now pervasive on sexual networking websites like Adam4Adam.com, GayRomeo.com, or Manhunt.net. Digital cameras, camera phones, and scanners were still many years away from consumer use, and web cameras (i.e., cameras tethered to one's computer) were not widely adopted until later in the 1990s. In contrast to today, it was more socially acceptable to meet someone off the Internet without first exchanging digital photos. The infrastructure of the Internet and limited available technology in the 1990s fostered anonymity by default, providing unique opportunities for gay and bisexual men to make connections online in the privacy of their homes, at all hours of the day (Benotsch et al., 2002). In short, this virtual environment was revolutionary (Bull & McFarlane, 2000; Friess, 1998; Klausner, Wolf, Fischer-Ponce, Zolt, & Katz, 2000).

Prior to the Internet, the main ways in which gay and bisexual men socialized was through gay bars and clubs and community groups/centers, as well as public sex venues: adult bookstores, bathhouses, and cruising parks or bathrooms (Frankis & Flowers, 2009; Humphreys, 1975; Shilts, 1987). Such public atmospheres were unattractive to individuals who may not have been out about their sexual interests in other men, or for those who were still exploring their sexual identities (Weinrich, 1997). There was also the potential for physical harm (i.e., being assaulted, raped, robbed) and police arrest in public sex environments. Gay spaces such as bars, clubs, and bathhouses were also less accessible to those living in non-urban areas. In contrast, the Internet emerged as a space that was available 24 hours a day (unlike bars/clubs), through which users could interact with others without revealing their full identity (Weinrich, 1997).

For gay and bisexual men, the 1990s marked parallel growth in cybersex/cyberfantasy (i.e., sharing erotic material with other users virtually for the purpose of sexual pleasure) as well as meeting partners online for sex offline (Cooper, Delmonico, & Burg, 2000; Griffiths, 2000; Schwartz & Southern, 2000). Researchers, however, were not quick to respond; thus, there is little information available about patterns of use among adults, let alone gay and bisexual men (Binik, 2001). We do not know how much time gay and bisexual men spent in chat rooms, nor do we have a clear estimate of how many partners men met online (or how often they met partners from the Internet). Instead, some of the research questions explored in the later 1990s and into the early 2000s investigated if it was possible for people to become addicted to the Internet (Chaney & Dew, 2003; Dew & Chaney, 2004; Griffiths, 1999), finding evidence that excessive Internet use could have addictive properties for some. Qualitative retrospective accounts from gay men interviewed in 2001 noted that, for some men, excessive time spent cruising for sex partners online could lead to negative outcomes in one's person's life (Grov et al., 2008). Yet, there was little epidemiological data on rates/levels of use, much less addiction.

In 2001, the Journal of Sex Research published a special section on Sexuality and the Internet; however, none of the articles was specifically devoted to men who have sex with men. As the editor of the special section wrote, “When I was first asked to guest edit a special issue of The Journal of Sex Research on the topic of sexuality and the Internet, I thought this would be a relatively easy task… it became apparent that there was not as much new data and theory as I had originally believed” (Binik, 2001, p. 281). Instead of performing studies of gay and bisexual men's use of the Internet, researchers in the 1990s were still grappling with the feasibility of conducting scientific inquiry online (Ross, Tikkanen, & Mansson, 2000). ‘How can we use the Internet to do sex research?’ ‘What are the ethical concerns?’ ‘Can we trust what people tell us online?’ Those few exceptions who analyzed Internet-based behavior in the 1990s often dealt with Internet addiction and, as is still the case today, sexual risk behavior (Bull & McFarlane, 2000; Gauthier & Forsyth, 1999; Griffiths, 1999; Klausner et al., 2000; Shaw, 1997; Toomey & Rothenberg, 2000). Still, these studies did not emerge until the late 1990s, furthering our lack of understanding of what patterns, if any, existed in the early 1990s. These data may have been particularly interesting given that highly effective antiretroviral therapy to treat HIV became available in 1996—it may have proven useful to investigate correlations of Internet use and sexual behavior before and after these treatments were made available.

Given the often anonymous nature of the Internet, there were many challenges to conducting studies in online environments, some of which continue today, including multiple submissions by a single user and lack of experimental control (Reips, 2000). It was also often challenging to confirm participants' identities and experiences (Gauthier & Forsyth, 1999; Ross et al., 2000). A common ethical concern noted in the 1990s was that investigators may unknowingly conduct research with minors who misrepresented their age in order to be eligible for a study (Binik, Mah, & Kiesler, 1999). Researchers also raised questions about the ethics of analyzing data on participants who might not have known they were being studied (King, 1996). These included studying behavior in chat rooms without disclosing one's role as a researcher, and studying content that might have been posted to the Internet by users, but never with the intention of it being analyzed scientifically.

Thus, some of the questions pursued by researchers in the 1990s investigated the feasibility and validity of conducting research online (Buchanan & Smith, 1999; Couper, 2000; McGraw, Tew, & Williams, 2000). ‘Are online samples similar to “real life” samples, or is there some systematic bias?’ (Joinson, 1999; Ross et al., 2000; Tikkanen & Ross, 2000). Studies from this time concluded that the Internet was a highly effective medium through which to conduct research with gay and bisexual men, particularly because many of these men may be ‘hidden.’ There were (and continue to be) concerns around bias and the validity of generalizability of online samples. For example, it is difficult to determine a response rate to an online survey, and Internet-based research automatically excluded individuals who did not have access to the Internet. In a Swedish study, researchers compared 716 written questionnaires and 678 Internet questionnaires gathered via gay chat rooms (Ross et al., 2000; Tikkanen & Ross, 2000). They noted that Internet chat rooms may attract younger men, men who identify themselves as bisexual, and men who live outside of major cities. To respond to these considerations, researchers in the mid-to-late-1990s began to publish guidelines for ensuring scientific rigor in online studies (Attila & Robert, 1996.) These included maintaining proper net etiquette (‘netiquette’) with participants, obtaining informed consent, not ‘bombarding’ listservs with frequent postings, and obtaining permission from scale developers to implement their questionnaires online, amongst other recommendations (Attila & Robert, 1996; Michalak & Szabo, 1998).

Similarly, researchers were concerned as to whether online sexual behaviors matched offline sexual behaviors, and what impact new use of the Internet to find sex partners may have had on the sex men had offline. ‘Does cyberfantasy translate into reality?’ ‘Is meeting sexual partners on the Internet associated with the transmission of HIV and SITs?’ (Bull & McFarlane, 2000). Researchers at the time noted that gay and bisexual men engaged in cyberfantasy as well as used the Internet to meet sex partners in real life (Gauthier & Forsyth, 1999; Klausner et al., 2000; Shaw, 1997; Tikkanen & Ross, 2000). In 1999, the San Francisco Department of Public Health was able to trace a recent outbreak of Syphilis among MSM using chat room screen names, and subsequently mapping social-sexual networks of men (Klausner et al., 2000). This landmark study highlighted the feasibility of using the Internet to map sexual networks as well as engage hard-to-reach individuals into treatment and care.

Lastly, though studies today suggest there is a complicated relationship between the Internet and sexual risk behavior, results from the 1990s suggest that use of the Internet to meet sex partners was a known risk factor for HIV and STI transmission. From an epidemiological standpoint, because the Internet afforded its users greater access to sex partners, the sheer opportunity for HIV or STI to be transmitted increased (Bull & McFarlane, 2000). Though the concept of barebacking (engaging in intentional unprotected anal sex) was not new in the 1990s, researchers have suggested that, because the Internet allowed users to locate others with specialized interests, it may have subsequently contributed to the growth in, and social acceptability of, engaging in bareback sex (Gauthier & Forsyth, 1999). We note, however, that the Internet is perhaps not the only factor, as new treatment options (e.g., HAART in 1996) coupled with overall HIV prevention fatigue have also played significant roles in attitudes toward barebacking (Carballo-Diéguez & Bauermeister, 2004; Halkitis & Parsons, 2003; Halkitis, Parsons, & Wilton, 2003).

The Internet and sex, 2000 – 2004

The early 2000s brought forth a series of technological advances, including the transition from dial-up to broadband, an uptake in wireless connectivity (i.e., Wi-Fi), and a larger number of ISPs offering monthly subscriptions to gain access to the World Wide Web, rather than hourly charges. These advances fueled a broader adoption of Internet use among gay and bisexual men and introduced opportunities for the creation of sites catering to different communities without being subjected to oversight and regulation inherent in working with large-scale online communities (e.g., CompuServe, America Online). Alongside this liberalization of regulation from service providers came expanded user-driven content (e.g., chat rooms and private instant messaging as previously discussed) as well a new features such as message boards (e.g., craigslist.org), and the expanse of male-for-male websites with personal profiles that included users' (often revealing) images (a byproduct of expanding consumer adoption of digital scanners and digital cameras).

As discussed, earlier research suggested that meeting sex partners on the Internet might be related to unprotected anal sex. In response to growing use of the Internet and studies suggesting the Internet may be a viable medium through which to conduct research, researchers began to consider developing intervention studies online (Bull, Lloyd, Rietmeijer, & McFarlane, 2004; Wang & Ross, 2002). This introduced new methodological questions around using the Internet for both recruitment and retention amongst broader samples of gay and bisexual men (Fernández et al., 2004; Parsons et al., 2004). Related, studies began to explore whether online data collection would be a viable alternative to face-to-face interviewing (Davis, Bolding, Hart, Sherr, & Elford, 2004; Elford, Bolding, Davis, Sherr, & Hart, 2004a, 2004b; Rhodes, DiClemente, Cecil, Hergenrather, & Yee, 2002). This ongoing research quickly helped researchers recognize that the Internet's accessibility, affordability, anonymity and acceptability could help ongoing efforts to understand gay and bisexual men's sexuality (Mustanski, 2001; Pequegnat et al., 2007).

Simultaneously, the ubiquity of Internet in the 21^st century also led to an increase in the number of individuals who met romantic and/or sexual partners online (Bull, McFarlane, & Rietmeijer, 2001a; Chiasson et al., 2007). This may have resulted from the fact that there are substantially fewer gay and bisexual men compared to heterosexuals in the general population, which decreased the likelihood of encountering another gay or bisexual man in offline venues. In a large survey of British MSM under age 30, Bolding et al. (2007) found a significant increase in the number of young men who have sex with men (YMSM) who met their first sexual partner on the Internet between 1993 and 2002 (2.6%–61.0%). Simultaneously, these researchers found a decrease over time in the proportion of men who met their first partners at an offline gay venue (34.2%–16.9%), school (23.7%–1.3%), or a public sex environment, print ad, or telephone chat line (10.5%–1.3%). Thus, it became clear that the Internet had quickly established itself as the modal venue through which to meet sex partners.

In the early 2000s, many of methodological advances in using the Internet for research were driven by public health researchers' needs to understand and address the HIV/STI outbreaks observed among gay and bisexual men who had met sexual partners in chat rooms (Centers for Disease Control and Prevention, 2003; Klausner et al., 2000). Researchers began to document epidemiological evidence linking Internet use to HIV/STI infections across cities in the United States (Benotsch et al., 2002; Halkitis et al., 2003; Rietmeijer, Bull, McFarlane, Patnaik, & Douglas, 2003; Tashima, 2003), Canada (Jayaraman, Read, & Singh, 2003), Europe (Elford, 2000), and Asia (Lau, Kim, Lau, & Tsui, 2003; Wang & Ross, 2002). Identifying the Internet as a ‘modern bathhouse,’ researchers placed increasing attention on gay and bisexual men's online interactions and their relationships to sexual risk behaviors associated with HIV/STI infection (Chen, Gibson, Weide, & McFarland, 2003).

Through their meta-analysis of 22 per-reviewed studies published between 2000 and 2005, Liau, Millet, and Marks (2006) documented the increased growth of the Internet as a medium through which to meet partners, as well as the association between Internet use and HIV/STI risk behaviors. A major obstacle their study overcame was to account for the large variability in data collection modalities and research findings. This meta-analysis found that partner-seeking behaviors online were most prevalent in studies for which MSM had been recruited from chat rooms (84.7%, 95% CI 81.4%-88.0%), as compared to studies for which MSM were recruited and surveyed offline (40.2%, 95% CI 35.2%-45.2%). Only 8 studies tested whether online partner-seeking behaviors were associated with greater sexual risk behaviors. The pooled analyses suggested that MSM who reported looking for partners online were more likely to have engaged in unprotected anal intercourse (OR = 1.68; 95% CI 1.18-2.40) than those who had not, with unprotected anal intercourse being more likely among gay and bisexual men recruited online (OR = 1.57, 95% 1.25-1.98) than those recruited offline. Based on these findings, Liau et al. (2006) expressed concern that different sampling methodologies could be confounding the associations between Internet use and sexual risk behaviors due to selection bias (e.g., men engaging in more high risk sex may also be more likely to be recruited online), and underscored the importance of considering what processes could explain why risk behaviors were more likely to be expressed in online environments.

Across samples, researchers acknowledged that online partner seeking expedited men's ability to have a sexual encounter (Benotsch et al., 2002; Bolding, Davis, Sherr, Hart, & Elford, 2004; Bull, McFarlane, & King, 2001b; Rietmeijer et al., 2003; Tikkanen & Ross, 2000; Uy, Parsons, Bimbi, Koken, & Halkitis, 2004). In addition to this affordance, researchers posited additional explanations for increased risk, including that men seeking partners online were also more likely than gay and bisexual men who met partners face-to-face to use illicit substances (Mettey, Crosby, DiClemente, & Holtgrave, 2003b), have casual sex partners (Kim, 2001; Taylor et al., 2004), be sexually compulsive (Chaney & Dew, 2003; Cooper, 2000), have a history of STIs (Elford, 2000), or have had unprotected sex with one or more partners of an unknown or discordant HIV status (Harterink, Hoek, Hospers, & Veenstra, 2002; Liau et al., 2006). Alongside these explanations, an emerging body of literature also began to document gay and bisexual men's decisions to engage in unprotected sex in HIV-risk contexts (barebacking) when meeting partners online (Carballo-Diéguez & Bauermeister, 2004; Grov, 2004; Halkitis & Parsons, 2003). As this collection of evidence emerged, researchers and practitioners responded with a series of HIV/STI intervention approaches targeting men who sought partners online.

Studies noted that gay and bisexual men sought out sexual health information online when they needed it (Bolding et al., 2004; Rietmeijer et al., 2003), and reported willingness to participate in HIV/STI prevention through chat rooms, email, and websites (Bull et al., 2001b). We identified three online intervention studies for gay and bisexual men between 2000 and 2004. In collaboration with a local AIDS service organization, Rhodes (2004) conducted an exploratory study where sexual health educators participated in local chat rooms and provided users with strategies to reduce their sexual risk, enact social support, and access testing and risk reduction supplies. Klausner, Levine and Kent (2004) proposed a series of intervention activities (e.g., website and chat room outreach, banner ads, message boards, and an online syphilis testing program) for gay and bisexual men living in San Francisco. Bull et al (2004) proposed an Internet-based randomized control trial that directed intervention participants to a tailored website focused on HIV/STI awareness and prevention for gay and bisexual men. These interventions were identified as highly innovative at the time, yet methodological barriers diluted their ability to test whether behavior change occurred post-intervention. Although feasibility and acceptability were documented as high in all three interventions, Rhodes' intervention did not ascertain the effectiveness of online conversations with chat room sex educators. Similarly, Klausner et al. (2004) found some success across intervention conditions based on process evaluation, yet they did not randomize participants to the different conditions, nor did they capture impact evaluation data. Bull et al.'s intervention, on the other hand, randomized men across the conditions, yet their intervention effects were limited given challenges in retaining participants (a 15.2% follow-up) at a 3-month post-intervention follow-up.

Findings suggested that these men who meet their sex partners online may be appropriate candidates for targeted HIV and STI prevention and education, and that the Internet might be an effective medium through which to deliver interventions. Yet, there were many methodological and technological barriers that needed to be addressed in order to deliver interventions including how to recruit and retain participants during intervention trials, to evaluate the efficacy and effectiveness of online interventions, and to design interventions that paralleled the rapid web-design and programming changes emerging during this time. As well, more information was needed on the exact mechanisms through which the Internet might have been related to increased HIV and STI transmission risks (e.g., determining if this was a within-individual effect or between-individual effect). These limitations would become challenges to tackle during the design of online interventions as well as formative research in the following years to come (Chiasson et al., 2006; Pequegnat et al., 2007).

The Internet and Sex, 2005-2009

As the second half of the first decade of the 21^st century progressed, the Internet continued to integrate itself into the daily lives of young and adult gay and bisexual men, so much so that the “shift” to digital communities was implicated as leading to the decline of physical gay communities (i.e., one no longer needs to go to a gay space to interact with other gay men) (Rosser, West, & Weinmeyer, 2008). Young gay and bisexual men in particular reported high levels of daily Internet use (Mustanski, Lyons, & Garcia, 2011a), and gay and bisexual men reported using the Internet for multiple purposes related to sexual health, including making connections with the LGBT community, seeking sexual health information, and meeting partners for dating or having sex (Bauermeister, Leslie-Santana, Johns, Pingel, & Eisenberg, 2011; Bolding et al., 2007; Dehaan, Kuper, Magee, Bigelow, & Mustanski, 2012; Garofalo, Herrick, Mustanski, & Donenberg, 2007; Kubicek, Carpineto, McDavitt, Weiss, & Kipke, 2011; Magee, Bigelow, Dehaan, & Mustanski, 2012; Mustanski et al., 2011a; Pingel, Bauermeister, Johns, Eisenberg, & Leslie-Santana, 2013; Wilkerson, Smolenski, Horvath, Danilenko, & Simon Rosser, 2010). Accordingly, the later 2000s saw a rise in research on gay and bisexual men's use of the Internet to seek sexual health information, and both quantitative and mixed methods research indicated that young and adult gay and bisexual men frequently used the Internet for this purpose (Eisenberg, Bauermeister, Pingel, Johns, & Leslie-Santana, 2011; Kubicek et al., 2011; Magee et al., 2012; Wilkerson et al., 2010). In fact, evidence suggested that gay and bisexual men began to specifically seek information on sexual health online in order to fill in gaps in information they may not have received from offline sources (i.e., school-based sex education, medical providers) (Dehaan et al., 2012). As many participants in this qualitative study described being cautious of trusting information found online, they described several strategies for increasing confidence in online information, such as making comparisons with other websites or offline sources.

The later years of the 2000s saw further inquiry into the role of meeting partners online in transmission of HIV/STIs, and advances in online technology which facilitated more rigorous research methodology (i.e., prospective behavioral diary studies, branching of survey items based on previous responses) shed further light on this association. Though several studies from the early 2000s found evidence of a link between meeting sexual partners online and sexual risk behavior (Benotsch et al., 2002; Berry, Raymond, Kellogg, & McFarland, 2008; Elford, Bolding, & Sherr, 2001; Garofalo et al., 2007; Horvath, Bowen, & Williams, 2006; Kakietek, Sullivan, & Heffelfinger, 2011; Kim, Kent, McFarland, & Klausner, 2001; Liau et al., 2006; McFarlane, Bull, & Rietmeijer, 2000), multiple studies from the later 2000s found no relationship between these variables (Bolding, Davis, Hart, Sherr, & Elford, 2005; Chiasson et al., 2007; Grov, Hirshfield, Remien, Humberstone, & Chiasson, 2013; Jenness et al., 2010; Mettey, Crosby, DiClemente, & Holtgrave, 2003a; Mustanski, Newcomb, & Clerkin, 2011b). Moreover, some studies have actually found evidence for less risk taking with partners met online (Horvath, Rosser, & Remafedi, 2008b; Mustanski, 2007a). In order to account for these inconsistent findings, several researchers in the late 2000s suggested that there may not be a causal relationship between online sex partner seeking and sexual risk. Instead, young and adult gay and bisexual men who have a history of engaging in sexual risk behavior may use the Internet as an efficient means to find partners with whom to engage in unprotected acts (Bauermeister et al., 2011; Downing, 2012; Garofalo et al., 2007; Horvath et al., 2008b; Mustanski, 2007a; Ogilvie et al., 2008). This represented a dramatic shift in our understanding of the role of the Internet in the sexual behaviors of gay and bisexual men.

As previously mentioned, methodological limitations in the literature may further explain the inconsistencies of these findings. Up until the later 2000s, many studies of the association between meeting partners online and sexual risk had been cross-sectional, and few studies had analyzed accounts of multiple sexual encounters within persons. Using notable methodological innovation, Mustanski (2007a) compared retrospective and prospective daily diary approaches to investigating meeting partners online and sexual risk for gay and bisexual men. In his study, the associations between these two variables were in direct contrast when comparing retrospective and prospective accounts of sexual behavior. In retrospective accounts, history of meeting sex partners online was associated with various indices of sexual risk behavior, but prospective daily diary data indicated that meeting partners online was actually associated with a lower likelihood of unprotected anal intercourse within persons. In another study, YMSM who met their partners offline only (compared to those who met partners online only or both online and offline) reported the lowest number of total sex partners but the highest percentage of partners with whom they engaged in unprotected sex (Horvath et al., 2008b).

Although seeking partners online may not be directly associated with sexual risk-taking behaviors, the ease and anonymity of online communication facilitated a more efficient communication of needs and desires related to sexual behavior, and this may have increased risk behavior in some groups while decreasing risk for others. For example, researchers in the later 2000s continued to describe the ease with which self-identified barebackers expressed their desire to have unprotected sex and seek like-minded partners (Berg, 2008; Carballo-Diéguez et al., 2009; Grov et al., 2007). To the extent that the Internet facilitates unprotected sex in this high-risk population, it is an important venue for engaging high-risk gay and bisexual men in sexual risk reduction interventions. Conversely, research during the later 2000s illustrated that the anonymity of the Internet actually facilitated conversations about condom use and HIV status that may have otherwise been uncomfortable, awkward, or distressing in face-to-face encounters (Horvath, Nygaard, & Simon Rosser, 2010b; Horvath, Oakes, & Rosser, 2008a; Ross, Rosser, McCurdy, & Feldman, 2007). As such, meeting sexual partners online may be associated with less risky sexual behavior for certain groups of gay and bisexual men.

As noted, research utilizing the Internet for participant recruitment, survey administration and intervention delivery increased rapidly during the first decade of the 21^st century. In particular, the ubiquity of the Internet allowed sexual health researchers to conduct behavioral surveys entirely online, which allowed them to more efficiently verify the quality of the data collected online (Bauermeister et al., 2012; Bowen, Williams, & Horvath, 2004; Konstan, Rosser, Ross, Stanton, & Edwards, 2005). As technology advanced and the Internet became an increasingly viable medium on which to meet partners, the later 2000s saw the rise of longitudinal designs (e.g., online sexual diaries) that were once much more cumbersome and time-intensive with offline administration. Online sexual diary studies involve administering surveys to participants in which sexual encounters are tracked prospectively over a specified period of time, and they often assess sexual behaviors, sexual partner characteristics, and a variety of other situational and contextual variables relevant to the sexual encounter or partner. Researchers in the later 2000s, and continuing to today, explored the quality of data from diaries over different assessment and recall periods. These have included daily diaries for one-month (Grov, Golub, Mustanski, & Parsons, 2010; Mustanski, 2007b), two months (Gillmore et al., 2002; Leigh et al., 2008), or once-weekly diaries during 6-week (Boone, Cook, & Wilson, 2012) and 3-month (Newcomb & Mustanski, 2013) follow-up periods. The frequency of survey administration in the sexual diary methodology once involved significant participant burden when administered via paper and pencils surveys, but the Internet greatly improved the efficiency of this design. Furthermore, despite initial concerns that retention rates were low with online samples (Bull et al., 2001a), online sexual diary studies have reported high completion rates (Mustanski, 2007a; Newcomb & Mustanski, 2013) and have been able to recruit racially diverse samples (Boone et al., 2012; Newcomb & Mustanski, 2013).

In total, research into the later 2000s established that gay and bisexual men could be retained in longitudinal online assessments and interventions (Bowen, Williams, Daniel, & Clayton, 2008; Hightow-Weidman et al., 2012; Mustanski, Garofalo, Monahan, Gratzer, & Andrews, In Press) including those targeting racial minority men, rural men, non-gay/bisexual identified MSM, gay and bisexual men who are not ‘out,’ and recruiting participants from HIV testing clinics to enroll in online interventions (Bowen, Horvath, & Williams, 2007; Bowen et al., 2008; Rhodes et al., 2010; Saxton, Dickson, & Hughes, 2013). As researchers continued to retool HIV/STI prevention and education efforts for online environments, it also became necessary to develop, administer, and test more technologically savvy content. The expansion of high-speed broadband and wireless technology (e.g. Wi-Fi), as well as improved computer processing, allowed for the content of online interventions to include digital media (e.g., streaming video) as well as interactive websites, live chat in chat rooms, and social networking with peer leaders (Amirkhanian et al., 2005; Bowen et al., 2007; Carpenter, Stoner, Mikko, Dhanak, & Parsons, 2010; Chiasson, Shaw, Humberstone, Hirshfield, & Hartel, 2009; Hirshfield et al., 2012; Mustanski et al., In Press; Rhodes et al., 2011).

In addition to these formally-developed HIV prevention interventions, the Internet facilitated the delivery of a variety of other services that may improve the sexual and physical health of MSM, including web-based partner notification services for HIV/STI exposures (Mimiaga et al., 2008; Rietmeijer et al., 2011), health educators available for live chat on social and sex partner websites (McFarlane, Kachur, Klausner, Roland, & Cohen, 2005), development of websites targeting vulnerable groups (e.g., information for newly HIV-diagnosed individuals; Horvath et al., 2010a), HIV/STI testing location finders (CDC, 2013), and purchasing of at-home HIV tests (Sharma, Sullivan, & Khosropour, 2011).

Throughout the 2000s and continuing to today, racial minority MSM have been disproportionally represented in the HIV epidemic, particularly YMSM (Phillips, Wohl, Xavier, Jones, & Hidalgo, 2011; Prejean et al., 2011). As studies demonstrated the potential effectiveness of delivering HIV prevention interventions online, researchers continued to grapple with questions over representativeness—particularly if online samples were systematically excluding racial minority MSM (i.e., higher proportion of White MSM in online samples; Du Bois, Johnson, & Mustanski, 2012; Sanchez, Smith, Denson, Dinenno, & Lansky, 2012b). However, certain online recruitment techniques were more fruitful than others in obtaining samples of racial minority MSM. For example, racial minorities may be more likely to respond to direct marketing techniques (i.e., banner ads) as opposed to web-based respondent-driven sampling, and some research indicates that online recruitment may even be more successful than certain offline venue-based recruitment strategies (e.g., bar/club outreach) at enrolling racial minority MSM (Sanchez, Smith, Denson, Dinenno, & Lansky, 2012a). Additionally, online racial minority recruitment appears to be most effective when banner ads use pictures of other racial minorities (Sullivan et al., 2011). Other research has suggested that online recruitment facilitates obtaining samples of “hidden populations” of MSM that may not be readily accessible through traditional offline venue-based strategies: bisexual men, HIV-positive MSM, drug-using MSM, and MSM who are less connected to the gay community (Fernandez et al., 2007; Parsons, Vial, Starks, & Golub, 2013).

In addition to these advancements in recruitment of racial minority men, data from the 2000s suggested that online data collection could be just as accurate as in-person data collection (Gosling, Vazire, Srivastava, & John, 2004); while emerging data suggests online recruitment might be more effective in reaching higher-risk MSM who would otherwise not be reached via venue-based sampling (Parsons et al., 2013; Sanchez et al., 2012b). As well, online recruitment of MSM has resulted in large national datasets including tens of thousands of MSM generated over very brief periods of time (Rosenberger et al., 2011b). The Internet may be perceived as more anonymous, thus reducing the effect of social desirability with self-report data (Gosling et al., 2004). This new trust in online research grew in part with the advent of mobile technologies, and with MSM's rapid uptake of new ways to connect via the Internet.

The Mobile Internet and Sex Today

For gay and bisexual men, the 1990s through the end of the first decade of the 21^st century saw a steady pattern indicating increased access and use of the Internet, particularly for sexual purposes. By and large, the method of interacting with other web users involved using a physical computer (desktop or notebook). Certainly, advances in technology made computing more mobile (i.e., notebook sales increased), but the method of connecting to the Internet (via a computer that often weighted several pounds) did not change. The later 2000s through today have seen a new and dramatic shift in the ways users connect to the Internet.

Although ‘smart’ phones (devices capable of connecting to the Internet) were available in the early and mid-2000s, the introduction of the iPhone in 2007, the subsequent ‘app’ (short for ‘application’) marketplace in 2008, and the iPad in 2010 ushered in the era of mass mobile computing. An ‘app’ is a third-party software application designed for use on mobile devices. Apps can include games, productivity tools (e.g., calendar), e-commerce (e.g., mobile shopping), media consumption (e.g., reading e-books, watching videos), and geo-social-networking (to name a few).

According to the International Data Corporation (IDC), it is projected that tablets will outsell desktop computers in 2013 and notebook computers in 2014, highlighting the web—as it has existed over the past decade—is rapidly changing and adapting to the new mobile market (Businesswire, 2013). In 2013, the US consumer spent an average of 2 hours and 38 minutes per day on smartphones and tablets (Khalaf, 2013). Eighty percent of that time (2 hours and 7 minutes) was spent inside apps and 20% (31 minutes) on the web (Khalaf, 2013). Between December 2011 and December 2012, the average time spent inside mobile apps by a US consumer grew 35%, from 94 minutes to 127 minutes. By comparison, the average time spent on the web declined 2.4%, from 72 minutes to 70 minutes (Khalaf, 2013). Although the number of apps available for download surpassed 100,000 in 2009, over 1.1 million apps have appeared in the United States as of April 12^th, 2013. The Android app market opened in 2009 with just 2,300 apps, and grew significantly; by 2013, it had the same number of available applications as Apple, bridging the connection between Android and iOS users. In May 2013, Apple celebrated its 50 billionth app download.

Although games and productivity apps are popular, mobile devices are also home to apps that connect users to both social (e.g., Facebook) and sexual networks (e.g., Grindr). In January 2013 alone, the 20 top dating apps had a combined 17 million active users and delivered more than 2.1 billion sessions. Examining app use by sexual orientation, those who identified as heterosexual typically opened their dating apps eight times a week and used them for seventy-one seconds at a time compared to users of dating apps for gay men which averaged twenty-two times a week for ninety-six seconds each time (Gordon, 2013). MSM have been shown to have greater access to and use of cell phone technologies compared to heterosexual populations (LGBT Market Research and Development Lab, 2012). Although previous estimates have suggested as many as 6.2 million gay and bisexual men use virtual tools for sexual connections, the exact number remains unknown (Liau et al., 2006). Creating estimates that account for both physical encounters resulting from a connection made via the Internet and those that occur strictly online (e.g., webcam) remains a challenge. Nonetheless, given the considerable number of users, a further understanding of mobile Internet products and their potential influence on MSM sexual behavior and sexual health is warranted.

The introduction of Grindr, a mobile-based GPS application for ‘gay, bi, and curious guys looking for dating or friends’ signaled the introduction of ‘gay sex’ apps in the marketplace (Burrell et al., 2012; Landovitz et al., 2012; Rice et al., 2012). Grindr launched on March 25^th, 2009; while uptake was slow at first (500,000 users in its first year), it acquired more than six million users across 192 countries in the three years that followed (Grindr, 2013). As of 2013, more than one million Grindr users logged onto the app every day and transmitted more than seven million chat messages and two million photos. Users spent approximately 1.5 hours using the app daily, logging in an average of eight times per day (Grindr, 2013). Interestingly, although it is widely understood that Grindr can be used as a place to meet other men for sex, due to standards set forth by Apple, the application itself does not market itself as an app designed to facilitate sex. Added, Grindr strictly forbids any references to sex in profiles, including restrictions on public pictures (e.g. no nudity, no underwear). Due to publication delays between when study conduction and publication of results, little research exists on how gay and bisexual men are using apps like Grindr. One study found 195 MSM aged 18-24 in Los Angeles reported that 76% had sexual encounters with partners met on Grindr (Rice et al., 2012). The study also found that participants reported significantly higher rates of condom use with partners met on Grindr (59.8%) relative to those partners met elsewhere (41.9%), which although not definitive, is certainly a shift from research a decade prior noting the elevated risks gay men experienced when meeting partners off the Internet.

Following the success of Grindr, other apps have emerged on the market, providing gay and bisexual men with more tailored products geared towards specific sub-communities/identities (e.g. Scruff (men attracted to beards/body hair), Growlr (the bear community), Skout (younger men looking for gay social networking)) and specific sexual behavior practices (e.g. Recon, for fetish/leather play). As these other apps continue to expand, this will likely be an arena for future research.

Given the recent popularity of Grindr, its use for research recruitment has been quite minimal. However, a recent study by Burrell and colleagues (2012) utilized Grindr as a means to send out recruitment advertisements for enrollment into a clinical trial to inform the development of rectal microbicides. The demographics and sexual risk behaviors of men recruited through Grindr were compared to those recruited using more traditional methods. On average, men recruited from Grindr were younger, more likely to be White, and reported more sexual partners in the past 2 weeks (Burrell et al., 2012). Similarly, a study by Landovitz et al. (2012) examined sexual risk behaviors and HIV prevention practices among MSM using Grindr. Their findings suggested that study participants had high rates of sexual partnering, and that among men engaging in unprotected anal intercourse, the majority (70.0%) perceived themselves to be at low risk for acquiring HIV.

Interestingly, the most popular social-sexual networking apps used by gay and bisexual men do not have complementary access via the Internet on a computer. One cannot visit grindr.com and expect to have a similar experience as the app. Apps like Grindr are in direct competition with many of the “traditional” websites gay and bisexual men had taken to over the years for sexual networking. According to HitWise, the largest global information company that measures Internet usage patterns, some of the most popular sexual networking websites in the US are geared toward gay and bisexual, including adam4adam.com, manhunt.net, and gay.com. These websites often feature sexually graphic content which would not be permitted in an app. Sites like Manhunt.net and Adam4Adam.com have responded to the growing interest in mobile platforms by developing what can best be described as “PG-rated” apps that only allow content that conforms to app standards (e.g., photos in the app area allowed to show a shirtless torso, but one in underwear would not be permitted). Websites like Manhunt.net and Adam4Adam.com have also created mobile versions of their sites that look and behave like apps (m.manhunt.net and radar.adam4adam.com). These Internet-based websites have created independent mobile-based products that are accessed via a smart device web-browser (as opposed to an app that is downloaded from the app store). In turn, this allows for mobile use while simultaneously circumventing “graphic content” restrictions in place by Apple and Android operating systems. It also provides a bridge between “static” websites and the mobile market.

As the capacity for individual men to connect virtually with potential sex partners has increased, so has the opportunity to harness emerging technologies for the purposes of sexual health promotion. For example, on Valentine's Day 2011, the New York City Health Department launched the NYC Condom Finder, a free smartphone application designed to locate the five nearest New York City venues that distribute free condoms. Using global positioning system (GPS) technology on a smartphone or by manually entering an address, a user is provided specific directions to each venue, the hours of operation for each location, the types of safer sex products available, and helpful tips on condom usage (Huffington Post, 2012).

Although mobile apps could also provide a strong platform for providing tailored disease prevention, those currently available for HIV and STI education and prevention have so far failed to become popular. Muessig et al. (Muessig, Pike, Legrand, & Hightow-Weidman, 2013b) searched both the Apple iTunes Store and the Android Google Play Store for any apps with the key words ‘HIV/AIDS,’ ‘STD,’ ‘STI,’ ‘sexual health,’ ‘safe sex,’ or ‘condom.’ A total of 1,937 unique apps were identified, though the authors only conducted further analysis on those that included HIV/STD information. In total, only 55 apps met the inclusion criteria. Among the findings in this study were that these apps were infrequently downloaded (median 100-500 downloads) with only 11 apps exceeding 1,000 downloads. Additionally, a minority of apps (n = 10) were inclusive of information about anal intercourse or featured information delivered specifically for lesbian, gay, bisexual, or transgender (LGBT) persons. The most popularly downloaded app was Sex Facts, which featured a revolving index of sex-related facts that users could share via social network sites (e.g., Facebook and Twitter). Though sexuality was listed in the app's description, the overwhelming majority of facts centered on heterosexual content (Muessig et al., 2013b).

It may be that smart devices are the next platform through which to deliver sexual health information. However, little is known about the feasibility and acceptability of delivering HIV and STI prevention via smart devices. Muessig et al (2013a) conducted surveys and focus groups with 22 black MSM aged 18-30. Participants were queried about what they would like in an HIV-related app. Responses included user-friendly content about test site locators, sexually transmitted diseases, symptom evaluation, drug and alcohol risk, safe sex, sexuality and relationships, gay-friendly health providers, and connection to other gay/HIV-positive men.

Although user-driven content (often called Web 2.0) began to gain popularity in the later 2000s, there has been tremendous growth in recent years, perhaps fueled by the amount of users participating via mobile devices. By definition, a Web 2.0 allows users to interact and collaborate with each other in a social media dialogue as creators of user-generated content. Each of these various user-generated modalities offers different ways for MSM to engage with each other in both sexual and non-sexual ways, with the prime example of the It Gets Better Project™. Launched in 2010 by sex columnist Dan Savage and his partner Terry Miller in response to targeted bullying of young LGBT individuals, they created a YouTube video as a way to support those facing harassment. To date, The It Gets Better Project™ has become a worldwide movement, inspiring more than 50,000 user-created videos, viewed more than 50 million times, with submissions from individuals across all walks of life, including the President of the United States (It Gets Better Project, 2010).

Surprisingly, given the overwhelming dominance of Internet and mobile technologies in men's daily lives, the current literature examining the Internet and sexual practices is limited, and primarily focused on YMSM. Several recent studies have documented that men routinely indicate the Internet as a source for meeting sexual partners. In congruence with other studies, Grov et al. (2010) found that HIV-positive and HIV-negative unknown status men reported a sizable portion of their recent partners from the Internet (55% and 29%, respectively) (Grov, Golub, & Parsons, 2010). Similarly, in a study examining differences among venue preference for meeting sexual partners (i.e., Internet, bathhouse, bar/club), men who preferred to meet partners online had the greatest number of recent sex partners and men who used the Internet as a primary meeting place were less likely than men who used other places to identify as gay (Grov et al., 2013). In addition to examining differences among venues, researchers have also begun to examine differences across types of websites. With the increasing acceptability of online dating sites (e.g., Match.com), researchers have also begun to document how gay and bisexual men's online partner seeking behaviors may vary based on the type of partner being pursued. For example, Bauermeister et al. (2011) explored the association between sexual behaviors and online partner-seeking behaviors for casual and romantic partners in a sample of 431 YMSM. Most participants reported using the Internet for romantic and casual partner-seeking, with half of the sample reporting at least three hours per week spent online seeking a romantic partner (M = 6.19, SD = 8.16) and at least two hours seeking a casual partner (M = 4.26, SD = 7.01). Further, over half of the sample reported being ‘a little confident’ that they would find their ideal romantic partner in a dating or social networking site, but ‘not confident at all’ about finding their ideal romantic partner in a hookup site.

Another area of research among gay and bisexual men that has increasingly garnered attention is the use of the Internet and mobile devices for the consumption of pornography. Up until recently, little research has investigated gay and bisexual men's pornography consumption, let along the mediums through which they accessed pornography. In the largest study to date, Stein et al. (2012) examined pornography viewing among a sample of 2,552 MSM in New York. Almost all participants reported viewing gay pornography (99.0 %), with most men equally indicating having watched depictions of protected anal intercourse and unprotected anal intercourse (95.0% and 94.0% respectively). The median time spent viewing gay pornography was 60 minutes per week, with the most common media used to view gay pornography being the Internet (96 %). In another study of 1,391 MSM living in the US, 98.5 % of participants reported exposure to sexually explicit media (SEM), during the last 90 days (Rosser et al., 2013). Confirming the dominance of Internet-mediated SEM, most participants (97.8 %) reported accessing SEM on a computer, followed by video/DVD (45.4 %), then by Internet through a phone or mobile device (42.0 %) (Rosser et al., 2013). In this study, 41% indicated they preferred bareback porn, 17% preferred porn that depicted condom use, and 42% reported no preference. There appears to be an association between watching bareback porn and engaging in bareback sex in real life; however, there is insufficient data to determine causality. In essence, is this association a result of men who prefer bareback sex gravitating toward bareback porn?

Although the mobile web presents innovative potential with regard to research and health outreach, it simultaneously provides a new set of limitations and challenges. There is some evidence to suggest that gay and bisexual men might be adopting mobile technologies faster than the general population. A study conducted in 2011 with 660 NYC gay and bisexual men in bars/clubs and bathhouses noted that 72% of participants owned a smart device and an additional 8.6% said they planned to buy one within the next 12 months (Grov, Ventuneac, Rendina, Jimenez, & Parsons, in press). In contrast, a Pew Research Center study of U.S. adults estimated only 35% own smart phones (Smith A. for the Pew Resarch Center, 2011). However; a sole emphasis on emerging technologies by clinicians, outreach workers, and academics has the potential to create disparities rooted in access issues. Although the technological gap is closing, there still remains a divide between users of these products, with those of lower socioeconomic status, lower educational attainment, and those who are older being less likely to own smartphones and other mobile devices (Buente & Robbin, 2008; Chinn & Fairlie, 2007; Zickuhr, 2011). In addition to inequities, utilization of these technologies for the purposes of prevention or treatment by health providers raises questions about how to virtually ensure protection of patient's privacy (although, similar considerations exist in academic and research settings).

To improve access to online communities, partnerships between researchers and website owners might be necessary (Wohlfeiler, Hecht, Raymond, Kennedy, & McFarland, 2011; Wohlfeiler et al., 2012). Online communities, and the corporate entities that facilitate them, have the potential to play important roles in our ability to attain public health goals. Although balancing public health/research and corporate goals seems contradictory, the formation of academic-community-corporate partnerships could offer new and innovative mechanisms for improving the sexual health of MSM. An example of such a partnership is highlighted in the work conducted by Rosenberger et al. over the past several years in collaboration with Online Buddies Inc. (owners and operators of Manhunt.net). Since 2009, data have been collected from the membership base of Manhunt.net with the goal of documenting and understanding the sexual behaviors of gay and bisexual men who use online websites for social and sexual interactions (Calabrese, Rosenberger, Schick, Novak, & Reece, 2013; Hensel, Rosenberger, Novak, & Reece, 2012; Jozkowski et al., 2010; Rosenberger, Reece, Novak, & Mayer, 2011a; Rosenberger et al., 2011b, 2012a; Rosenberger, Schick, Herbenick, Novak, & Reece, 2012b; Stupiansky et al., 2010). Given the number of users of manhunt.net, this partnership provided an opportunity to capture sexuality data from sample sizes in the tens of thousands. The studies utilized both online cross sectional surveying and longitudinal daily diaries to examine a variety of sexual practices including measurements of sexual behavior during specific events, and condom use over time. The data revealed some interesting information on the types of sexual behavior that MSM reported, including the most commonly reported behavior was kissing a partner on the mouth (74.5%), followed by oral sex (72.7%), and partnered masturbation (68.4%). Anal intercourse occurred among less than half of participants (37.2%) and was most common among men ages 18–24 (42.7%) (Rosenberger et al., 2011b). In terms of condom use, age, race/ethnicity, partner status, and location of sexual event were all significantly related to the likelihood of condom use during men's most recent penile anal intercourse with another man (p < .001) (Rosenberger et al., 2012a).

Conclusions and Recommendations

Over the last three decades, gay and bisexual men have rapidly taken to using the Internet for sexual purposes. These include sexual health information seeking, finding sex partners, dating, cybersex, pornography, and for sex work (to name a few). In addition to adoption, gay and bisexual men have adapted to the ever-evolving technological advances that have been made in connecting users to the Internet. This has evolved from logging into the World Wide Web via dial-up modem, to engaging in anonymous instant messaging in a chat room, to geo-social networking and sharing erotic content with others over a handheld device.

Researchers too have been adopting and adapting to the Internet, though perhaps not at the same rapid pace at which technology (and its users) have advanced. Studies have carefully considered the ethics, feasibility, and acceptability of using the Internet to conduct research with gay and bisexual men. Researchers have evaluated the use of the Internet to enroll individuals in face-to-face studies, to engage in pure web-based studies, delivering intervention content, and studying the efficacy of health interventions in digital environments. Over the years, the Internet and technology have evolved the methods we have used to study gay and bisexual men have evolved, as have some of our research questions. However, it is also clear that more work needs to be done, particularly with regard to intervention development. For example, it is necessary to strengthen process and outcome evaluation procedures for online technologies. These include the measurement of intervention dosage, message exposure, and competing tasks in an online environment. Given the evolution of the Internet as an increasingly social (user-driven) environment, more work is also needed on effective methods for tracking Intervention diffusion.

The growth of the Internet has also resulted in its diversification. Websites, and now apps, tap into a wide range niche markets and sub-populations of gay and bisexual men. As a result, considerations are needed to address potential sampling bias when recruiting participants off a single site. While it is known that websites may attract a ‘type’ of user, researchers may benefit from determining if individual's behaviors change depending on the particular web forum, thus evaluation within-, versus between-, individual characteristics..

In contrast to technological advances, many of our research questions remained grounded in models of disease prevention. Perhaps this has been warranted, as gay and bisexual men continue to bear disproportionate burden of the HIV/AIDS epidemic and are among the only at-risk groups who have seen recent increases in new diagnoses (CDC, 2012). Likewise, the urgent need to reduce HIV in this population has been a driving force to develop innovative research and intervention methodologies. In essence, it may be that this population would have otherwise been ignored and some of the innovative research methodologies would have otherwise been undeveloped. Yet, we can only wonder what other questions might have been explored were we not so focused on preventing HIV. Now over three decades into the epidemic, it is clear that some of the progress made in HIV prevention may be stalled (Stall et al., 2009), thus presenting researchers and community outreach providers with new challenges in deploying Internet-based HIV prevention, treatment, and education. It may be that our lack of understanding of sexual behaviors other than unprotected anal sex has contributed to currently stalled efforts. For example, it may be that previous efforts have resulted in a body of literature about gay and bisexual men that is disease focused and that has not fully allowed for an exploration of the manner in which these men construct their sexual lives (Reece & Dodge, 2004). Particularly missing from the literature on gay and bisexual men has been work related to the variation in men's sexual repertoires, the extent to which these men practice certain behaviors, and the true motivations for engaging in particular motivations. Alternate approaches, such as models of resilience, may help to fill gaps in our understand of gay and bisexual men's sexuality as well as achieving goals to reduce HIV and STI transmission (Herrick et al., 2011).

It is clear that the future of the Internet will be via mobile devices; whether on a smart phone or tablet, or integrated into everyday products like eyewear (e.g., Google Glass). Likewise, it is clear that the Internet has become an increasingly social resource. And, as has been the case with new technologies historically, mobile products and the mobile Internet will be widely adopted for purposes related to sexuality. Several obstacles face researchers seeking to understand the Internet and sex. These include adapting research methods to constantly shifting technologies, and the significant delay between when data are collected and when the results are finally available for others to learn from.

Geo-social networking apps like Grindr have been around for several years, yet only a handful of peer-reviewed publications on the topic are currently available. It may be that, by the time we fully understand how gay and bisexual men are using geo-social networking apps, a newer technology has already replaced what exists today. As consumers move to mobile devices and a greater number of applications are marketed to gay and bisexual consumers, it is clear that our attention span has become increasingly shorter. Although it may be possible to have an individual complete a two-hour long assessment at a research office using Audio Computer-Assisted Self Interview (ACASI) software, most Internet based surveys are designed to be brief in order to avoid attrition. Incentives can increase the amount of time participants would be willing to engage in a research study, but they are also magnets for spam, duplicate respondents, and participants otherwise trying to earn money, even if it means misrepresenting themselves or their data (Bauermeister et al., 2012). As devices shrink, the amount of content that can be presented to a user is also reduced. To date, there are limited data regarding participation and completion rates, or data quality across studies administered via a computer web browser, versus a tablet, versus a smart phone. It is also necessary to explore which online recruitment techniques and recruitment messages are most effective in reaching targeted populations (Parsons et al., 2013). These are empirical questions we hope researchers will explore in the years to come.