http://orcid.org/0000-0001-8586-8711
Figures
Abstract
Background
Emerging research is using global positioning system (GPS) and ecological momentary assessment (EMA) methods among sexual minority men (SMM), a population that experiences multiple health disparities. However, we are not aware of any research that has combined these approaches among SMM, highlighting the need for acceptability and feasibility research. The purpose of this study was to examine the acceptability of implementing GPS and EMA research protocols using smartphone applications among SMM as well as related socio-demographic correlates.
Methods
Data come from a sample of SMM on a popular geosocial-networking app in Paris, France (n = 580). We assessed the acceptability of implementing GPS and EMA research protocols on smartphone apps as well as socio-demographic characteristics (i.e., age, sexual orientation, country of origin, employment status, and relationship status). We examined the anticipated acceptability of GPS and EMA data collection methods as well as socio-demographic correlates of acceptability of GPS and EMA methods.
Results
We found that over half (54.1%) of the sample was willing to download a smartphone app for GPS-based research and we found that almost 60% of the participants were willing to download a smartphone app for EMA-based research. In total, 44.0% reported that they were willing to download both GPS and EMA apps. In addition, we found that older participants were less willing to download a smartphone app for EMA research than younger participants aged 18–24 (40–49 years: aPR = 0.40; 95% CI = 0.20, 0.78) and students were more willing to download smartphone apps for both GPS and EMA research (aPR = 1.41; 95% CI = 1.02, 1.95).
Citation: Duncan DT, Park SH, Goedel WC, Sheehan DM, Regan SD, Chaix B (2019) Acceptability of smartphone applications for global positioning system (GPS) and ecological momentary assessment (EMA) research among sexual minority men. PLoS ONE 14(1): e0210240. https://doi.org/10.1371/journal.pone.0210240
Editor: Iratxe Puebla, Public Library of Science, UNITED KINGDOM
Received: September 13, 2017; Accepted: December 19, 2018; Published: January 28, 2019
Copyright: © 2019 Duncan et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The data (along with the codebook) have been updated to the NYU Faculty Digital Archive: https://archive.nyu.edu/handle/2451/43759.
Funding: This work was funded in part by National Institutes of Health grants R01MH112406, R21MH110190, and R03DA039748 and the Centers for Disease Control and Prevention grant U01PS005122. This work was supported by Dr. Dustin Duncan’s New York University School of Medicine Start-Up Research Fund. Diana Sheehan was supported in part by a grant from the National Institute On Minority Health And Health Disparities (Award K01MD013770). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Ecological momentary assessment (EMA) includes a range of methods aimed at capturing data on health behaviors and outcomes in real-time from participants who self-report as they experience their daily lives [1]. These methods were developed in response to the limitations of self-report retrospective recall, as it can frequently be unreliable and often be systematically biased [2]. These methods recognize that behaviors and experiences are affected by context, so data must be collected on a given experience or behavior in its real-time setting for the data collected to be representative of one’s lived experiences [1]. Studies using these methods often involve repeated measures over varying durations, affording the temporal resolution needed to assess the dynamics of within-subject changes in behavior and experience over time and across contexts [3]. These methods have employed various modalities to collect data on behavior in real-time in real world environments, including emerging research using smartphone applications [4]. These methods of capturing information in real-time are useful tools in studying health behaviors—as they are often discrete and episodic [1]
The contexts of health behaviors, such as the neighborhood environments in which the behaviors are performed, are experienced continuously before, during, and after the performance of the health behavior being studied. [5] As such, integrating EMA methods with global positioning system (GPS) technologies may be an effective strategy for capturing both health behaviors in real time and in their environmental contexts; however, few applications of this combined approach exist in the literature and studies have used different combined GPS-EMA protocols [6, 7]. Epstein and colleagues [6] and Mitchell and colleagues [7], for example, both used a dedicated GPS unit in combination with a separate dedicated personal digital assistant (PDA) for EMA to assess the role of mood and geographic context in substance use. A single smartphone application could have the potential to leverage both a smartphone’s geolocation features to record the device’s latitude and longitude [8] and the ability to deliver prompted questionnaires via the smartphone’s notification system [9].
Importantly, GPS technology is a highly novel approach to studying the neighborhood contexts in which people engage/interact [10]. Indeed, GPS methods allow researchers to examine the multiple spatial contexts that an individual experiences (not just the residential neighborhood environment) and allow researchers to understand the timing and duration of exposures to these spatial contexts (addressing the “uncertain geographic context problem”) [11–13]. Emerging research has employed GPS [14, 15] and EMA methods [16–19] separately among sexual minority men (SMM), a population at high risk for HIV infection and whom experience numerous health disparities [20]. However, no published studies (to the best of our knowledge) have used EMA and GPS approaches simultaneously—a method sometimes referred to as geographically-explicit EMA (GEMA) [6, 21]–among SMM, highlighting the need for feasibility research prior to implementing these methods on larger scales.
As such, the purpose of this study was to examine the anticipated acceptability of using smartphone applications to implement both GPS and EMA research protocols among SMM. In addition, we examined whether socio-demographic characteristics were correlated with a participant’s willingness to download a smartphone application for GPS- and EMA-based research. We hypothesized the combined GPS-EMA research protocols using smartphone applications would be acceptable among SMM, however we also expected there to be socio-demographic variation.
Methods
Sample.
We used broadcast advertisements in October 2016 on a popular geosocial-networking smartphone application for SMM to recruit participants. We limited the advertisements to users in the Paris (France) metropolitan area. Users were shown an advertisement with text encouraging them to click through the advertisement to complete an anonymous web-based survey. We provided an incentive to encourage participation, which we described prior to survey administration and implemented after survey administration. In particular, the advertisement described that users who completed the survey were entered in a chance to win €65, which is approximately $70. The advertisement was placed during three consecutive 24-hour weekday periods. After implementing precautions to avoid and eliminate duplicate responses [22], we found no apparent duplicate responses. Our survey included 52 items; the survey was translated from English into French using an adaptation of the TRAPD (translate, review, adjudicate, pretest, document) translation protocol [23] and in this study five French speakers conducted the survey translation. The survey took an average of 11.4 minutes for users to complete. The survey was offered in French and English. Most participants (94.3%), however, took the survey in French. At the end of the recruitment period, 5,206 users had clicked on the advertisement and reached the landing page of the survey, 935 users provided informed consent and began the survey, and 580 users provided informed consent and completed the survey. This represents an overall completion rate of 62% and an overall response rate of 11.1%. All protocols were approved by the New York University School of Medicine Institutional Review Board prior to data collection.
Anticipated acceptability of GPS and EMA smartphone-based app research.
We asked participants, “Would you download a smartphone app that tracked where you went using global positioning system (GPS) technology for the purposes of a research study?” Response options were “Yes” and “No”. In addition, we asked participants “Would you download a smartphone app that asked you questions throughout the day about your current mood, surroundings, and feelings for the purposes of a research study?” Response options were “Yes” and “No”. For the analysis, we generated binary variables indicating the acceptance of only one of these applications (i.e., accepting of GPS but not EMA and accepting of EMA but not GPS) as well as made a composite variable of willingness to download a GPS and EMA app.
Socio-demographic characteristics.
We assessed age (categorized into five groups: 18–24 years, 25–29 years, 30–39 years, 40–49 years, 50 years and older), sexual orientation (gay, bisexual, other), whether or not they had been born in France (yes, no), employment status (employed, unemployed, student, retired), and current relationship status (coded as single or in a relationship).
Statistical methods.
For this study, we decided on the four groups: neither, EMA alone, GPS alone and GPS and EMA because in addition to combined GPS-EMA protocols, it might be useful to know whether there are specific correlates of willingness to use each method alone versus using both simultaneously, which aligns with the goals of the study. We examined the anticipated acceptability of GPS and EMA data collection methods as well as the socio-demographic characteristics associated with such acceptance. First, a descriptive analysis of the study sample was performed. A chi-square or Fisher’s exact test, as appropriate, was used to examine 1) differences in participant characteristics for each outcome, independently, based on acceptability of the GPS and EMA methods (e.g. GPS only vs. all other); and 2) differences in participant characteristics across the four groups (GPS only, EMA only, GPS and EMA and neither GPS nor EMA). Second, using the log-binomial regression model, we assessed adjusted associations to examine the correlates of anticipated acceptability of the GPS and EMA methods. The adjusted prevalence ratio (aPR) and 95% confidence intervals (CIs) were computed. We estimated the prevalence ratio (as opposed to the odds ratio) because the outcomes were relative common and it is well known in epidemiologic studies that odds ratios overestimate the effect in the context of a non-rare outcome [24–26]. Socio-demographic variables, including age, sexual orientation, origin (born in France), employment status, and current relationship status, were adjusted for. All statistical analyses were conducted using Stata 14.0 (StataCorp, College Station, TX, USA).
Results
Study participants
The median age was 34 years old (interquartile range [IQR] = 27.0, 42.0), and most participants identified their sexual orientation as either gay (84.0%) or bisexual (11.9%) (Table 1). Most participants (77.6%) were born in France. Furthermore, most respondents were employed (66.9%), and 14.0% were students. The majority of participants reported being single (65.2%).
Acceptability of GPS and EMA smartphone app-based research.
A total of 18 people did not answer either the GPS question or EMA question: 17 people were missing responses to both questions and 1 participant did not provide a response for the EMA question. In our analytical sample, more than half (54.1%) the sample was willing to download a smartphone-based GPS application for research purposes. Almost 60% of participants were willing to download a smartphone-based EMA application. In total, 44.0% reported that they were willing to download both the GPS and EMA applications; 10.2% were willing to download the GPS but not the EMA application; 14.1% were willing to download the EMA but not the GPS application; and 166 people (28.6%) were not willing to download either the GPS or the EMA application for research purposes. The tetrachoric correlation between these two dichotomous outcomes (i.e., downloading GPS and downloading EMA) was computed (r = 0.6978).
Socio-demographic correlates of acceptability of GPS and EMA smartphone app-based research
Individuals born outside France were more willing than those born in France to download an app for a GPS protocol but not for an EMA protocol, whereas unemployed participants were more likely to download an app for an EMA protocol but not for a GPS protocol (also found in our multivariate models described below). Differences were found in the ages of participants according to their willingness to download both GPS and EMA applications (Table 1). In addition, the age distribution across the four groups varied according to individual characteristics (Table 1). Multivariate analyses revealed that, after adjusting for sociodemographic characteristics, older individuals were less willing than the youngest group to download an app for an EMA protocol but not for a GPS protocol (Table 2). For example, compared to those aged 18–24 years, participants aged 40–49 years had an aPR of 0.40 (95% CI = 0.20, 0.78) for downloading an app for an EMA protocol but not for a GPS protocol. Unemployed participants were more willing to download an app for an EMA protocol but not for a GPS protocol than employed participants (aPR = 1.82; 95% CI = 1.11, 2.99). No such relationship was found for the anticipated willingness to download an app for GPS methods but not an app for EMA methods. There was an increasing probability to accept an app for both purposes with increasing age. For example, participants aged 50 years and older were more willing than the youngest group (aPR = 2.24; 95% CI = 1.37, 3.66), but for a given age, students were more likely to accept the app than non-student (aPR = 1.41; 95% CI = 1.02, 1.95). No differences were found in sexual orientation and current relationship status in terms of anticipated acceptability of GPS and EMA protocols.
Discussion
The purpose of this study was to examine the anticipated acceptability of SMM downloading smartphone applications for GPS and EMA protocols. As previously discussed, emerging GPS and EMA research has been conducted among SMM, however, to our best knowledge, no study has combined GPS and EMA methods among any SMM population. In addition, none of the existing GPS studies have used a smartphone app for data collection in a SMM population. But rather, existing research has utilized a dedicated GPS device [14, 15]. A recent study that evaluated the anticipated acceptability of text message- and voice-based EMA methods among a sample of 74 young SMM found that almost all participants (96%) reported that they would be willing to accept texts on their smartphone to answer questions about their current mood, surroundings, or feelings [27]. In addition, a very large majority (89%) reported being willing to accept phone calls to answer these questions [27]. Among existing studies that implemented EMA methods among SMM, two studies have used smartphone applications on a non-participant smartphone [17, 18] and one allowed participants to use their own smartphones to participate [19]. Willingness to download an app to a personal phone may increase compliance to research protocols.
We recognize that use of an EMA application increase cellular data; however, use of an EMA application as compared to EMA texts might be more feasible in population-based research health research, in part because participants can be charged for text messages—making them less likely to engage in research protocols that involve texts being sent to them. In addition, apps might provide more flexibility for the researcher in terms of questions, responses and other special features and apps may be more visually appealing and consequently encourage participation in health-related research. In this study, willing to downloading a smartphone app for GPS methods alone and downloading a smartphone app for EMA protocols alone were moderately correlated and we found that most SMM were willing to download a smartphone app for either purpose. However, there was socio-demographic variation, where, for example, older participants were more willing to download a smartphone application for both GPS and EMA protocols.
The variation in willingness to download these apps by age is interesting: the youngest are more reluctant to use an app for GPS-based research but are more open to using an app for EMA-based research, underlining that these two approaches have different implications in terms of privacy. Perhaps, there is a stronger feeling of distrust among out-of-school young SMM. Future studies could examine concerns related to the use of smartphone apps for GPS and EMA protocols with SMM using qualitative methods such as semi-structured interviews. Future studies could also test the feasibility of implementing GPS and EMA studies where SMM download smartphone apps to track their mobility and examine their health behaviors in real-time, including with large samples of SMM across geographies as well as assessing the feasiblity of collecting GPS and EMA data in SMM populations for different research topics (e.g., research on sexual behavior, drug use and/or mental health), which may impact feasibility. These studies would advance the field substantially by providing a nuanced spatio-temporal link between environment and risk behavior, also because there may be differences in hypotheticals for engagement in a research study versus real life behaviors in this particular population. This future research on neighborhoods and the health of gay, bisexual and other SMM, is warranted in light of increasing research demonstrating that neighborhoods can influence HIV outcomes in SMM populations [28] as well as other health outcomes/behaviors [29, 30]. The implications for this next-generation work are that they can potentially revolutionize our researchers and practitioners’ views of biomedical prevention, including pre-exposure prophylaxis (PrEP) interventions.
Limitations
This study has limitations. Social desirability bias is possible. In addition, this study was conducted in a single urban geographic location in Western Europe among a sample of SMM recruited from a single geosocial-networking application, which limits the generalizability of our findings. For example, SMM who use geosocial-networking applications might be different from those who do not. In particular, SMM who use geosocial-networking applications may be more willing than SMM who do not download other applications, including those for GPS tracking, as these geosocial-networking applications already utilize a smartphone’s geolocation features to connect their users with other nearby users [31]. Furthermore, these data were collected two years ago and the expansion of smartphones and attitudes toward technology may have changed since then.
Conclusion
We found that over half (54.1%) of the sample was willing to download a smartphone app for GPS-based research and almost 60% of the participants were willing to download a smartphone app for research using EMA methods. We also found socio-demographic variation in willingness to download these apps. Therefore, care should be taken when conducting this research—as segments of SMM are less likely to be willing to engage in this type of research. Future studies using GPS and EMA methods together will advance the field substantially by providing a nuanced spatio-temporal link between environment and risk behavior in SMM, a population that experiences multiple health disparities.
Acknowledgments
We thank the translators and participants of this study who contributed to the project. We also thank Yazan Al-Ajlouni for conducting background research for this article and Yazan Al-Ajlouni and Hayden Mountcastle for his help in preparing this article. We thank Noah Kreski and Jace Morganstein for assisting in the development, translation and management of the survey used in the current study.
References
- 1.
Stone AA, Shiffman S. Ecological momentary assessment (EMA) in behavorial medicine. Annals of Behavioral Medicine. 1994.
- 2. Coughlin SS. Recall bias in epidemiologic studies. Journal of clinical epidemiology. 1990;43(1):87–91. pmid:2319285
- 3. Schwartz JE, Stone AA. Strategies for analyzing ecological momentary assessment data. Health Psychology. 1998;17(1):6. pmid:9459065
- 4. Runyan JD, Steenbergh TA, Bainbridge C, Daugherty DA, Oke L, Fry BN. A smartphone ecological momentary assessment/intervention “app” for collecting real-time data and promoting self-awareness. PLoS One. 2013;8(8):e71325. pmid:23977016
- 5. Hektner JM, Csikszentmihalyi M. The experience sampling method: Measuring the context and content of lives. Handbook of environmental psychology. 2002;8:233–43.
- 6. Epstein DH, Tyburski M, Craig IM, Phillips KA, Jobes ML, Vahabzadeh M, et al. Real-time tracking of neighborhood surroundings and mood in urban drug misusers: application of a new method to study behavior in its geographical context. Drug and alcohol dependence. 2014;134:22–9. pmid:24332365
- 7. Mitchell JT, Schick RS, Hallyburton M, Dennis MF, Kollins SH, Beckham JC, et al. Combined ecological momentary assessment and global positioning system tracking to assess smoking behavior: a proof of concept study. Journal of dual diagnosis. 2014;10(1):19–29. pmid:24883050
- 8. Wu X, Brown KN, Sreenan CJ. Analysis of smartphone user mobility traces for opportunistic data collection in wireless sensor networks. Pervasive and Mobile Computing. 2013;9(6):881–91.
- 9. Torous J, Kiang MV, Lorme J, Onnela J-P. New tools for new research in psychiatry: a scalable and customizable platform to empower data driven smartphone research. JMIR mental health. 2016;3(2).
- 10.
Duncan DT, S R, Chaix B.. Operational Neighborhood Definitions in Health Research: Spatial Misclassification and Other Issues. In: Duncan DT, K I, editor. Neighborhoods and Health 2nd Edition Oxford, UK: Oxford University Press; 2018.
- 11. Chaix B, Merlo J, Evans D, Leal C, Havard S. Neighbourhoods in eco-epidemiologic research: delimiting personal exposure areas. A response to Riva, Gauvin, Apparicio and Brodeur. Social science & medicine. 2009;69(9):1306–10.
- 12. Kwan M-P. The uncertain geographic context problem. Annals of the Association of American Geographers. 2012;102(5):958–68.
- 13. Kwan M-P. How GIS can help address the uncertain geographic context problem in social science research. Annals of GIS. 2012;18(4):245–55.
- 14. Duncan DT, Kapadia F, Regan SD, Goedel WC, Levy MD, Barton SC, et al. Feasibility and acceptability of global positioning system (GPS) methods to study the spatial contexts of substance use and sexual risk behaviors among young men who have sex with men in New York City: A p18 cohort sub-study. PloS one. 2016;11(2):e0147520. pmid:26918766
- 15. Duncan DT, Chaix B, Regan SD, Park SH, Draper C, Goedel WC, et al. Collecting Mobility Data with GPS Methods to Understand the HIV Environmental Riskscape Among Young Black Men Who Have Sex with Men: A Multi-city Feasibility Study in the Deep South. AIDS and Behavior. 2018:1–14
- 16. Rowe C, Hern J, DeMartini A, Jennings D, Sommers M, Walker J, et al. Concordance of Text Message Ecological Momentary Assessment and Retrospective Survey Data Among Substance-Using Men Who Have Sex With Men: A Secondary Analysis of a Randomized Controlled Trial. JMIR mHealth and uHealth. 2016;4(2).
- 17. Wray TB, Kahler CW, Monti PM. Using ecological momentary assessment (EMA) to study sex events among very high-risk men who have sex with men (MSM). AIDS and Behavior. 2016;20(10):2231–42. pmid:26746212
- 18. Yang C, Linas B, Kirk G, Bollinger R, Chang L, Chander G, et al. Feasibility and Acceptability of Smartphone-Based Ecological Momentary Assessment of Alcohol Use Among African American Men Who Have Sex With Men in Baltimore. JMIR Mhealth Uhealth. 2015;3(2):e67. Epub 2015/06/19. pmid:26085078
- 19. Rendina HJ, Millar BM, Parsons JT. The critical role of internalized HIV-related stigma in the daily negative affective experiences of HIV-positive gay and bisexual men. Journal of affective disorders. 2017;227:289–97. pmid:29128785
- 20. Stall R, Mills TC, Williamson J, Hart T, Greenwood G, Paul J, et al. Association of co-occurring psychosocial health problems and increased vulnerability to HIV/AIDS among urban men who have sex with men. American journal of public health. 2003;93(6):939–42. pmid:12773359
- 21. Kirchner TR, Shiffman S. Spatio-temporal determinants of mental health and well-being: advances in geographically-explicit ecological momentary assessment (GEMA). Social psychiatry and psychiatric epidemiology. 2016;51(9):1211–23. pmid:27558710
- 22. Duncan DT, Goedel WC, Mayer KH, Safren SA, Palamar JJ, Hagen D, et al. Poor sleep health and its association with mental health, substance use, and condomless anal intercourse among gay, bisexual, and other men who have sex with men. Sleep Health. 2016;2(4):316–21. pmid:29073390
- 23.
Harkness JA, Van de Vijver FJ, Mohler PP. Cross-cultural survey methods: Wiley-Interscience Hoboken,NJ; 2003.
- 24. Behrens T, Taeger D, Wellmann J, Keil U. Different methods to calculate effect estimates in cross-sectional studies a comparison between prevalence odds ratio and prevalence ratio. Methods Archive. 2004;43(5):505–9.
- 25. Thompson ML, Myers J, Kriebel D. Prevalence odds ratio or prevalence ratio in the analysis of cross sectional data: what is to be done? Occupational and environmental medicine. 1998;55(4):272–7. pmid:9624282
- 26. McNutt LA, Wu C, Xue X, Hafner JP. Estimating the relative risk in cohort studies and clinical trials of common outcomes. American journal of epidemiology. 2003;157(10):940–3. Epub 2003/05/15. pmid:12746247.
- 27. Duncan DT, Kapadia F, Kirchner TR, Goedel WC, Brady WJ, Halkitis PN. Acceptability of ecological momentary assessment among young men who have sex with men. Journal of LGBT Youth. 2017:1–9
- 28. Bauermeister JA, Connochie D, Eaton L, Demers M, Stephenson R. Geospatial Indicators of Space and Place: A Review of Multilevel Studies of HIV Prevention and Care Outcomes Among Young Men Who Have Sex With Men in the United States. The Journal of Sex Research. 2017:1–19.
- 29. Cerdá M, Nandi V, Frye V, Egan JE, Rundle A, Quinn JW, et al. Neighborhood determinants of mood and anxiety disorders among men who have sex with men in New York City. Social Psychiatry and Psychiatric Epidemiology. 2017:1–12.
- 30. Duncan DT, Park SH, Goedel WC, Kreski NT, Morganstein JG, Hambrick HR, et al. Perceived Neighborhood Safety Is Associated with Poor Sleep Health among Gay, Bisexual, and Other Men Who Have Sex with Men in Paris, France. Journal of Urban Health. 2017:1–9
- 31.
Toch E, Levi I, editors. Locality and privacy in people-nearby applications. Proceedings of the 2013 ACM international joint conference on Pervasive and ubiquitous computing; 2013: ACM.