Towards Understanding Users’ Motivation to Share Health Information on Facebook by Sadegh Torabi B.Sc. in Computer Engineering, Kuwait University, 2005 M.Sc. in Computer Engineering, Kuwait University, 2009 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF Master of Applied Science in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Electrical and Computer Engineering) The University of British Columbia (Vancouver) April 2016 c Sadegh Torabi, 2016
157
Embed
Towards Understanding Users’ Motivation to Share Health ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Towards Understanding Users’ Motivation toShare Health Information on Facebook
by
Sadegh Torabi
B.Sc. in Computer Engineering, Kuwait University, 2005
M.Sc. in Computer Engineering, Kuwait University, 2009
Rashtian, Primal Wijesekera, Ivan Cherapau, and Ahmad Mahfouz. Also, I
would like to thank Hyoungshick Kim and Primal Asanka for their guidance
and feedback during different stages of the research. Third, I would like to
thank NSERC Internetworked Systems Security Network (ISSNet), for their
partial funding and support of the research. Last but not least, I would like to
thank my beloved family, especially my parents, my wife, and all those who
showed endless support and compassion through my journey.
xiv
Dedication
To my beloved parents, my dearwife, and all my family memberswho never stopped supporting me.
xv
Chapter 1
Introduction
Patients with chronic health condition(s) can benefit from sharing their health
information (HI)1 on SNSs in different ways: seeking or providing social
support, learning from the shared experiences, and self-management educa-
tion [43, 48, 56]. Furthermore, sharing HI on SNSs was shown to be ben-
eficial not only to the individuals themselves, but also to their online social
peers. Motivated by the two-way benefits, a variety of web-based services
have been used by patients to share their HI with other people (e.g., Blogs,
emails). Among these services, Facebook, which have attracted the largest
number of active online users, have emerged as a common place for sharing
different types of personal information, including HI [22, 46].
A variety of methodological approaches were used to understand patients’
HI sharing practices and the benefits of sharing HI among different patient
1Any type of information related to the health of an individual including personal healthinformation (PHI), electronic health records (EHRs), and personal health records (PHRs)
1
groups (e.g., breast cancer patients) [43, 48]. Despite the benefits, revealing
HI to other people was always associated with privacy concerns. In addition,
an implicit consensus among people was noted toward the idea that their HI
should be considered as “personal” and “private” information [42, 48]. Be-
sides that, Facebook was shown to be vulnerable toward different attacks,
resulting in unintended information disclosure, privacy invasion, and medical
data misuse [10, 25, 55]. One can argue that the general attitude toward
disclosing HI on Facebook might be influenced by the perceived privacy risks
and the gained benefits (privacy calculus) [40]. Moreover, while a number
of studies brought attention to the benefits and the privacy concerns related
to HI sharing on Facebook, there has been little done to understand the main
factors that influence patients’ motivation to engage in effective HI sharing
on Facebook [39, 43, 44]. In addition, given that existing SNSs’ security and
privacy features were not designed to support patients’ privacy requirements
while sharing HI on SNSs, it is important to design new features that enable
patients to communicate their HI effectively and without any concerns. This,
however, demands building a better understanding of patients’ practices and
preferences when sharing HI on SNSs.
In this study, we bridge the knowledge gap by answering the general re-
search questions: Why, how, and with whom patients share their HI on Face-
book? Following a series of qualitative and quantitative studies, we explored
Facebook users’ HI sharing practices and risk perceptions. We started by con-
ducting a preliminary online survey to collect quantitative data from a broad
2
number of active Facebook users. The survey aimed at corroborating prior
findings while removing the uncertainties that were caused by the limited
related work. After that, we adopted a grounded theory approach to inter-
view individuals with chronic health conditions [12]. We focused on exploring
users’ practices, preferences, and risk perceptions when sharing HI on SNSs.
The interviews helped us in developing a better understanding of the reason-
ing behind patients’ behaviors and privacy perceptions, while identifying the
key factors that influenced patients’ motivation to share HI on SNSs. We con-
ducted an online survey in order to confirm and extend upon our findings. The
online survey provided the opportunity to recruit a more representative sam-
ple of Facebook users. Moreover, analyzing the quantitative data resulted gen-
eralizable findings that characterized the main factors that influenced users’
motivation to share HI.
The results showed that participants who previously shared their HI on
Facebook, especially those who gained some benefits, were more willing to
share their HI. On the other hand, despite the perceived benefits, participants
who had strong privacy concerns (privacy Fundamentalists) were always un-
likely to share their HI, as compared to others who showed more flexibility in
the presence of different motivating factors (privacy Pragmatists and Uncon-
cerned). Along with the aforementioned factors, we also showed how partici-
pants’ perceived health status could influence their motivation toward sharing
HI with other social peers, especially strangers. Furthermore, we showed that
the recipients of the shared HI, and the type of the shared HI, could also play
3
a major role in motivating participants toward sharing their HI. Other than
the factors that influenced HI sharing motivation, our results highlighted a
number of features that could motivate users toward engaging in effective HI
sharing on Favebook (e.g., anonymous identity, search and recommendation
features). Armed with such an understanding, new privacy-preserving fea-
tures could be developed that might be more effective in practice. Ultimately,
while optimizing the gained benefits, these features would improve Facebook’s
privacy and make it less vulnerable toward privacy threats.
The rest of the thesis is structured as following: we summarized back-
ground and related work in Chapter 2. We presented the overall methodology
in Chapter 3. Details of the conducted user studies are presented in Chap-
ters 4, 5, and 6 respectively. The main findings of the research were discussed
in Chapter 7. Finally, we concluded the thesis by presenting conclusions and
future work (Chapters 8 and 9).
4
Chapter 2
Background and Related Work
2.1 Social networking sites (SNSs)
Social networking sites (SNSs) are web-based services that allow their users
to create visible profiles within the site, maintain connections with a list of
online users (including strangers), and interact with other users through the
created connections [8]. SNSs are classified as Web 2.0 sites that incorporate
strong social components (e.g., maintaining friends lists), and encourage shar-
ing user-generated contents (e.g., photos, comments) [15]. While a variety of
additional services are provided by different SNSs (e.g., online gaming), they
are mainly used to facilitate maintaining social relationships. Also, SNSs pro-
vide their users with the opportunity to establish new relationships with others
who share common interests or background [39]. It is important to recognize
that a number of web-based services might also incorporate properties similar
5
to SNSs, where they allow limited profile creation and collaborative communi-
cations among members of the site (e.g., blogs, discussion forums). However,
despite the similarities in terms of the implementation of social components,
there has been uncertainty about classifying these online services as SNSs due
to the adopted user interface technology and architecture [8, 15].
2.2 Sharing HI on SNSs
SNSs have been developed and used for a variety of purposes, including but
not limited to: socializing and content sharing (e.g., Facebook, Twitter), net-
working and maintaining professional profiles (e.g., LinkedIn), and peer-to-
peer communication (e.g., Skype). The highly inter-connected nature of ex-
isting SNSs, and the increasing number of users have encouraged people to
actively engage in sharing different types of information on SNSs. Several
studies have indicated that HI sharing is becoming a leading habit among a
considerable number of SNSs users [22, 41, 49]. The results of the Pew In-
ternet survey [21] indicate that a considerable number internet users in the
US went online to follow their friends’ personal health experience, with a no-
ticeable increase when compared to the previous reports. Meanwhile, 16% of
the surveyed participants reported going online to find others who had sim-
ilar health concerns [22]. Moreover, people with health concerns have been
shown to visit their SNSs to seek support from other online peers [46].
Despite their practices, users’ motivation to share HI on SNSs was shown
to be influenced by a number of factors:
6
2.2.1 Perceived benefits
The benefits of using SNSs for HI sharing has been discussed in a number of
studies [42, 43, 56]. Lederman et al. [38] discussed the benefits of addressing
socio-technical needs by utilizing SNSs and developing engaging therapeutic
solutions for mentally ill patients. Following a user-centered design approach,
Skeels [48] captured breast cancer patients’ HI sharing requirements and de-
signed an online interactive technology to facilitate HI sharing and manage-
ment. Kamal [30] also used a similar approach to design a SNS prototype
for promoting healthy behavior changes. Despite the reported benefits for
patients, literature reviews showed that only a small number of studies have
focused on the effects of using SNSs on patients’ health management. For in-
stance, Merolli et al. [43] reviewed the literature and found that among all the
identified studies (N=19), only five studies focused on SNSs (classified as web
2.0 sites). For them, this was a clear indication that research on SNSs usage in
chronic disease management was still in its early stage. Moreover, Moorhead
et al. [44] surveyed primary research and identified the lack of information
about the uses, benefits, and limitations of social media for health communi-
cation among the general public, patients, and health professionals. Similar
conclusions were also drawn by Lefebvre and Bornkessel [39]. Furthermore,
they suggested addressing a number of general research questions in order to
better understand how SNSs can be effectively and efficiently used to improve
health across the population.
7
2.2.2 Perceived health status
Patients’ motivation to engage in protective health behaviors was shown to be
influenced by the severity and the likelihood of their health conditions [52]. In
the context of HI sharing on SNSs, a number of studies have shown that people
who suffered from chronic health conditions were likely to visit SNSs to seek
or share their HI with other social peers [21, 46]. There are different ways
to asses one’s overall health status and the severity of the health conditions.
The self-reported perceived health status was used as a reliable measurement
of patients’ overall health status in a number of studies [28, 54]. Also, they
found a correlation between the perceived health status and the number of
health conditions, with patients who had “poor” health to have more health
conditions. Nevertheless, the subjective assessment of the health conditions
might not always accurately describe the overall health status. For instance,
one might suffer from a number of severe health conditions and yet consider
his health to be stable or good, while another person might have a minor
health issue and feels completely devastated by his health issues.
2.2.3 Perceived privacy
The nature of SNSs can lead to the diffusion of personal information beyond
its intended targets, while resulting in the lack of subsequent control over
its exposure [10, 27, 44]. In general, information revelation in SNSs was
shown to be influenced by the raised privacy concerns due to both the per-
sonal experiences and the negative reports in the media [55]. In the context
8
of SNSs, privacy concerns have been always associated with sharing HI among
users [16, 51]. A survey of 1060 US adults found that 63% raised concerns
related to publicly sharing their HI on SNSs, while 57% were concerned that
their HI might be hacked or leaked from the SNSs [3]. Morris et al. [45]
surveyed different types of questions that SNSs users asked their social peers
about and found that “health” was a type of topic that people tend to consider
too personal. Moreover, SNSs were considered as inappropriate places for dis-
cussing overly personal information such as information related to individual’s
health [16, 45].
It has been shown that internet users’ privacy concerns and their attitudes
toward privacy could highly influence their motivation to disclose personal
information to online sites [11, 16, 49]. The Westin privacy index was intro-
duced as a way to meaningfully classify internet users based on their attitudes
toward privacy and their motivations to disclose personal information on the
internet [33]. According to the Westin privacy index, internet users could be
divided into three groups: (1) Privacy Fundamentalists, who highly value pri-
vacy and feel very strongly about it; (2) Privacy Pragmatists, who have strong
feelings about privacy but can also see the benefits from surrendering some
privacy in situations where they believe they can prevent the misuse of their
information; and (3) Privacy Unconcerned, who have no real concerns about
privacy or about how other people and organizations use their information.
A number of studies showed that the Westin privacy index could be used as a
reasonable indicator of internet users’ attitudes toward privacy and their moti-
9
vation to disclose personal information online [11, 14, 33, 34]. Nevertheless,
Woodruff et al. [53] argued that the prescribed classification of internet users
using the Westin privacy index might not always correlate with their motiva-
tion to disclosure information online. Despite their argument, we found that
the literature provides sufficient evidence to reflect on the overall correlations
between internet users’ privacy attitudes and their motivation to disclose per-
sonal information online.
10
Chapter 3
Methodology
The research was conducted following a three steps methodology: (1) a broad
survey of active SNS users HI sharing practices and risk perceptions in order to
corroborate prior findings and identify important research directions; (2) in-
depth interviews to explore patients’ motivation to share HI on SNSs; and (3)
a confirmatory survey in order to test our previous findings and identify the
main factors that influenced HI sharing motivations on SNSs. In what follows,
we provide more details about the methodology used for each conducted user
which consisted of information that were not identifiable by themselves but
could be used to describe properties of an individual in real life (e.g., age,
gender, hobbies); and (3) Identity information, which represented informa-
tion that could lead to revealing one’s real identity (e.g., picture, first/last
name). We also created an index variable for each component by averaging
participants’ rating for every information item within that component.
The results in Table 6.2 indicates that considerably more participants were
“(Very) Likely” to hide information related to their Contact and location and
Identity, with an average score of 91% and 81% respectively. Demographic
information on the other hand resulted the least score among all components
(average score of 38%), with statistically significantly less number of partici-
pants who were likely to hide their demographic information on Facebook as
compared to identity, contact, or location information. It was also interesting
to see that “health conditions” were categorized as Demographic information,
with about 30% of participants who were likely to hide information related to
their health conditions on Facebook.
78
Table 6.2: PCA results for different personal information items. The lastcolumn represents the percentage of participants who were likely tohide each information item.
ComponentFactor
loading
Agreement
(%)
Contact and location information — 91
Phone number 0.87 94
Residential address 0.80 95
Email address 0.74 89
Future location information 0.73 89
Current location information 0.72 86
Demographic information — 38
Gender 0.83 21
My health condition(s) 0.80 30
Hobbies and interests 0.69 38
Age and date of birth 0.62 64
Identity information — 81
Identifiable profile picture 0.79 87
Last name 0.75 89
First name 0.72 66
Information items that did not
conform to any component — —
Occupation and employment NA 76
City where I live NA 72
79
Comparing participants based on their privacy attitudes, we found statisti-
cally significant differences in participants motivation to hide different infor-
mation items with the exception of the following two cases: (1) Pragmatists
and Unconcerned participants’ motivation to hide Contact and location infor-
mation, which confirms the relatively higher level of concerns raised by all
participants toward revealing their Contact and location information; and (2)
Fundamentalists and Pragmatists participants’ motivation to hide Demographic
information, which indicates less concerns toward revealing information re-
lated to participants’ demographics. In addition, we found a statistically sig-
nificant difference in participants’ motivation to hide Demographic information
when compared based on their prior HI sharing experiences (p= 0.003). The
tests results also indicated that participants with Positive experiences in the
past were more willing to reveal their Demographic information as compared
to participants with other experiences (e.g., Negative).
6.2.9 Willingness to trust SNSs with HI
We asked participants to indicate their willingness to trust different SNS providers
with their submitted HI. About 27% of all participants trusted an SNS with
their HI if it was provided by a governmental health authority, while slightly
less than 20% of participants trusted a governmental agency (non-health re-
lated) and a recognized private company (Figure 6.11). On the other hand,
about 58% of participants did not trust an SNS that was provided by a govern-
mental agency (non-health related), which was relatively more than the per-
80
centage of participants who did not trust other SNS providers. Within-subjects
comparisons showed that significantly more number of participants were will-
ing to trust an SNS with their HI if it was provided by a governmental health
agency as compared to other providers.
0% 20% 40% 60% 80% 100%
A governmental agency (non-‐health related)
A recognized private company
A governmental health authority (e.g., city, state/province, federal/naDonal)
[3] Social media “likes” healthcare: From marketing to social business.PwC HRI Social Media Consumer Survey, 2012. URLhttp://download.pwc.com/ie/pubs/2012_social_media_likes_healthcare.pdf.→ pages 9
[4] S. A. Adams. Revisiting the online health information reliability debatein the wake of “web 2.0”: an inter-disciplinary literature and websitereview. International journal of medical informatics, 79(6):391–400,2010. → pages 85
[5] I. Altman. Privacy regulation: culturally universal or culturally specific?Journal of Social Issues, 33(3):66–84, 1977. → pages 19
[6] S. B. Barnes. A privacy paradox: Social networking in the united states.First Monday, 11(9), 2006. → pages 14
[7] T. Bodenheimer, K. Lorig, H. Holman, and K. Grumbach. Patientself-management of chronic disease in primary care. Jama, 288(19):2469–2475, 2002. → pages 51
[8] D. Boyd and N. Ellison. Social network sites: Definition, history, andscholarship. Journal of Computer-Mediated Communication, 13(1):210–230, October 2008. ISSN 10836101.doi:10.1111/j.1083-6101.2007.00393.x. URLhttp://dx.doi.org/10.1111/j.1083-6101.2007.00393.x. → pages 5, 6
[9] B. Bulgurcu. Understanding the Information Privacy-Related Perceptionsand Behaviors of an Online Social Network User. PhD thesis, The Facultyof Graduate Studies, Business Administration, University of BritishColumbia, August 2012. → pages 32
[10] J. A. Calandrino, A. Kilzer, A. Narayanan, E. W. Felten, andV. Shmatikov. ’you might also like:’ privacy risks of collaborativefiltering. In Proceedings of the 2011 IEEE Symposium on Security andPrivacy, SP ’11, pages 231–246, Washington, DC, USA, 2011. IEEEComputer Society. ISBN 978-0-7695-4402-1. doi:10.1109/SP.2011.40.URL http://dx.doi.org/10.1109/SP.2011.40. → pages 2, 8
[11] F. Chanchary and S. Chiasson. User Perceptions of Sharing, Advertising,and Tracking. In Eleventh Symposium On Usable Privacy and Security(SOUPS 2015). USENIX Association, 2015. → pages 9, 10, 65, 83
[13] V. Choudhury and E. Karahanna. The relative advantage of electronicchannels: a multidimensional view. Mis Quarterly, pages 179–200,2008. → pages 86
[14] S. Consolvo, I. E. Smith, T. Matthews, A. LaMarca, J. Tabert, andP. Powledge. Location disclosure to social relations: why, when, & whatpeople want to share. In Proceedings of the SIGCHI conference on Humanfactors in computing systems, pages 81–90. ACM, 2005. → pages 10, 65,83
[15] G. Cormode and B. Krishnamurthy. Key differences between web 1.0and web 2.0. First Monday, 13(6), 2008. → pages 5, 6
[16] M. De Choudhury, M. R. Morris, and R. W. White. Seeking and sharinghealth information online: Comparing search engines and social media.In Proceedings of the SIGCHI Conference on Human Factors in ComputingSystems, CHI ’14, pages 1365–1376, New York, NY, USA, 2014. ACM.ISBN 978-1-4503-2473-1. doi:10.1145/2556288.2557214. URLhttp://doi.acm.org/10.1145/2556288.2557214. → pages 9
[17] M. Duggan, N. B. Ellison, C. Lampe, A. Lenhart, M. Madden, L. Rainie,and S. Aaron. Social Media Update 2014. Pew Research Center,January 2015. → pages 55, 60, 71
[18] N. B. Ellison, C. Steinfield, and C. Lampe. The benefits of Facebook“friends:” social capital and college students’ use of online socialnetwork sites. Journal of Computer-Mediated Communication, 12(4):1143–1168, July 2007. ISSN 10836101.doi:10.1111/j.1083-6101.2007.00367.x. URLhttp://dx.doi.org/10.1111/j.1083-6101.2007.00367.x. → pages 55
[19] T. W. Feeley and K. I. Shine. Access to the medical record for patientsand involved providers: Transparency through electronic tools. Annalsof Internal Medicine, 155(12):853–854, 2011.doi:10.1059/0003-4819-155-12-201112200-00010. URL+http://dx.doi.org/10.1059/0003-4819-155-12-201112200-00010. → pages12
[20] S. Fox. Health topics. Pew Research Center Report, February 2011. →pages 107
[21] S. Fox. The social life of health information. Pew Research CenterReport, May 2011. → pages 6, 8, 20, 24
[22] S. Fox and M. Duggan. Health online 2013. Pew Research Center’sInternet and American Life Project, January 2013. → pages 1, 6, 12,20, 24
[23] S. Fox and M. Duggan. The Diagnosis Difference. Pew Research Center,November 2013. → pages 71, 83
[24] D. Gefen, E. Karahanna, and D. W. Straub. Trust and TAM in OnlineShopping: An Integrated Model. MIS Q., 27(1):51–90, Mar. 2003. ISSN0276-7783. URL http://dl.acm.org/citation.cfm?id=2017181.2017185. →pages 86
[25] R. Gross and A. Acquisti. Information revelation and privacy in onlinesocial networks. In Proceedings of the 2005 ACM Workshop on Privacy inthe Electronic Society, WPES ’05, pages 71–80, New York, NY, USA,2005. ACM. ISBN 1-59593-228-3. doi:10.1145/1102199.1102214. URLhttp://doi.acm.org/10.1145/1102199.1102214. → pages 2
[26] A. Hartzler, M. Skeels, M. Mukai, C. Powell, P. Klasnja, and W. Pratt.Sharing is caring, but not error free: Transparency of granular controlsfor sharing personal health information in social networks. In AMIAAnnual Symposium, volume 2011, pages 559–568, 2011. → pages 12
[27] C. Hawn. Take two aspirin and tweet me in the morning: how twitter,facebook, and other social media are reshaping health care. Healthaffairs, 28(2):361–368, 2009. → pages 8
[28] T. K. Houston and J. J. Allison. Users of internet health information:differences by health status. Journal of Medical Internet Research, 4(2),2002. → pages 8, 62, 83
[29] G. Iachello and J. Hong. End-user privacy in human-computerinteraction. Found. Trends Hum.-Comput. Interact., 1(1):1–137, 2007.ISSN 1551-3955. doi:http://dx.doi.org/10.1561/1100000004. → pages 14
[30] N. Kamal. Designing online social networks to motivate health behaviourchange. PhD thesis, University of British Columbia, 2013. → pages 7
[31] R. Kang, S. Brown, L. Dabbish, and S. Kiesler. Privacy attitudes ofmechanical turk workers and the us public. In Symposium on UsablePrivacy and Security (SOUPS), 2014. → pages 65
[32] H. Krasnova, S. Spiekermann, K. Koroleva, and T. Hildebrand. Onlinesocial networks: why we disclose. Journal of Information Technology, 25(2):109–125, 2010. → pages 46
[33] P. Kumaraguru and L. F. Cranor. Privacy indexes: a survey of Westin’sstudies. Technical report, Carnegie Mellon University, 2005. → pages 9,10, 59, 65, 83
[34] M. Kwasny, K. Caine, W. A. Rogers, and A. D. Fisk. Privacy andTechnology: Folk Definitions and Perspectives. In CHI’08 ExtendedAbstracts on Human Factors in Computing Systems, pages 3291–3296.ACM, 2008. → pages 10
[35] C. Lampe, N. Ellison, and C. Steinfield. A face(book) in the crowd:social searching vs. social browsing. In Proceedings of the 2006 20thanniversary conference on Computer supported cooperative work, CSCW’06, pages 167–170, New York, NY, USA, 2006. ACM. ISBN
[36] C. Lampe, N. B. Ellison, and C. Steinfield. Changes in use andperception of facebook. In Proceedings of the 2008 ACM conference onComputer supported cooperative work, CSCW ’08, pages 721–730, NewYork, NY, USA, 2008. ACM. ISBN 978-1-60558-007-4.doi:http://doi.acm.org/10.1145/1460563.1460675. URLhttp://doi.acm.org/10.1145/1460563.1460675. → pages 55
[37] R. Lazarus. Coping theory and research: Past, present, and future. FiftyYears of the Research and Theory of RS Lazarus: An Analysis of Historicaland Perennial Issues, pages 366–388, 1993. → pages 32
[38] R. Lederman, G. Wadley, J. Gleeson, S. Bendall, and M. ÁlvarezJiménez. Moderated online social therapy: Designing and evaluatingtechnology for mental health. ACM Trans. Comput.-Hum. Interact., 21(1):5:1–5:26, Feb. 2014. ISSN 1073-0516. doi:10.1145/2513179. URLhttp://doi.acm.org/10.1145/2513179. → pages 7
[39] R. C. Lefebvre and A. S. Bornkessel. Digital social networks and health.Circulation, 127(17):1829–1836, 2013. → pages 2, 5, 7
[40] Y. Li. Theories in online information privacy research: A critical reviewand an integrated framework. Decision Support Systems, 54(1):471–481, 2012. → pages 2, 46, 83
[41] Social Media Survey Report. Liquid Grids, May 2014. → pages 6
[42] F. Lupiáñez-villanueva, W. Lusoli, M. Bacigalupo, I. Maghiros,N. Andrade, and C. Codagnone. Health-related information as personaldata in Europe: Results from a representative survey in Eu27. Medicine2.0: Ethical and legal issues, confidentiality and privacy, 2011. → pages2, 7, 12, 36
[43] M. Merolli, K. Gray, and F. Martin-Sanchez. Methodological review:Health outcomes and related effects of using social media in chronicdisease management: A literature review and analysis of affordances.Journal of Biomedical Informatics, 46(6):957–969, 2013. → pages 1, 2,7, 85
[44] A. S. Moorhead, E. D. Hazlett, L. Harrison, K. J. Carroll, A. Irwin, andC. Hoving. A new dimension of health care: Systematic review of theuses, benefits, and limitations of social media for healthcommunication. J Med Internet Res, 15(4):e85, Apr 2013.doi:10.2196/jmir.1933. URL http://www.ncbi.nlm.nih.gov/pubmed/23615206.→ pages 2, 7, 8
[45] M. R. Morris, J. Teevan, and K. Panovich. What do people ask theirsocial networks, and why?: a survey study of status message q&abehavior. In Proceedings of the SIGCHI conference on Human factors incomputing systems, pages 1739–1748. ACM, 2010. → pages 9
[46] H. J. Oh, C. Lauckner, J. Boehmer, R. Fewins-Bliss, and K. Li.Facebooking for health: an examination into the solicitation and effectsof health-related social support on social networking sites. Computers inHuman Behavior, 29(5):2072–2080, 2013. → pages 1, 6, 8
[47] P. Ressler, Y. Bradshaw, L. Gualtieri, and K. Chui. Communicating theexperience of chronic pain and illness through blogging. Journal ofMedical Internet Research, 14(5):e143, 2012. → pages 12
[48] M. M. Skeels. Sharing By Design: Understanding and supporting personalhealth information sharing and collaboration within social networks. PhDthesis, University of Washington, 2010. → pages 1, 2, 7, 12, 20, 24, 107
[49] S. Torabi and K. Beznosov. Privacy Aspects of Health RelatedInformation Sharing in Online Social Networks. In USENIX Workshopon Health Information Technologies (HealthTech ’13). USENIXAssociation, August 2013. URLhttp://0b4af6cdc2f0c5998459-c0245c5c937c5dedcca3f1764ecc9b2f.r43.cf2.rackcdn.com/12452-healthtech13-torabi.pdf. → pages 6, 9
[50] B. Ur, P. G. Kelley, S. Komanduri, J. Lee, M. Maass, M. L. Mazurek,T. Passaro, R. Shay, T. Vidas, L. Bauer, et al. How does your passwordmeasure up? the effect of strength meters on password creation. InProc. USENIX Security, 2012. → pages 18, 55, 60
[51] M. Velden and K. El Emam. “not all my friends need to know”: aqualitative study of teenage patients, privacy, and social media. Journal
of the American Medical Informatics Association, 20(1):16–24, July2012. → pages 9, 12
[52] N. D. Weinstein. Perceived probability, perceived severity, andhealth-protective behavior. Health Psychology, 19(1):65, 2000. → pages8
[53] A. Woodruff, V. Pihur, S. Consolvo, L. Schmidt, L. Brandimarte, andA. Acquisti. Would a privacy fundamentalist sell their DNA for $1000...if nothing bad happened as a result? the westin categories, behavioralintentions, and consequences. In Symposium on Usable Privacy andSecurity (SOUPS), 2014. → pages 10
[54] N. Xiao, R. Sharman, H. Rao, and S. Upadhyaya. Factors influencingonline health information search: An empirical analysis of a nationalcancer-related survey. Decision Support Systems, 57:417–427, 2014. →pages 8, 62, 71, 83, 86
[55] A. L. Young and A. Quan-Haase. Information revelation and internetprivacy concerns on social network sites: a case study of facebook. InProceedings of the fourth international conference on Communities andtechnologies, pages 265–274. ACM, 2009. → pages 2, 8
[56] D. M. Zulman, K. M. Nazi, C. L. Turvey, T. H. Wagner, S. S. Woods, andL. C. An. Patient interest in sharing personal health record information:A web-based survey. Annals of Internal Medicine, 155(12):805–810,2011. doi:10.1059/0003-4819-155-12-201112200-00002. URL+http://dx.doi.org/10.1059/0003-4819-155-12-201112200-00002. → pages1, 7, 12
3. What is your highest level of completed education?
• Less than high school
• High school (secondary school)
• Diploma (post-secondary courses)
• Undergraduate university degree (Bachelor’s)
• Graduate university degree (Masters’s or PhD)
• Community college or professional school (College degree)
• Other (please specify)
4. Employment category: [Select a category that best fits your current job]
• A range of employment categories (e.g., Education, Business, Engi-
neering, ...), or
101
• Other (please specify)
5. Current country of residence?
• Select from list, or
• Other (please specify)
A.1.2 Questions About SNS Usage
1. How many active SNS profiles do you maintain? [Active accounts are
those which you regularly use to connect to people, share information,
and perform online activities]
• 1
• 2
• 3
• 4
• 5 or more
2. For your most frequently used SNS account(s), provide the name of the
SNS and specify how regularly do you log into your account. [Informa-
tion for at least one SNS is required, you may skip the rest if does not
apply]
The following questions will be repeated for each specified active SNS
account:
102
• SNS name [Select from list, or specify other SNS names]
• How regularly do you log into your above SNS account?
– Daily
– Weekly
– Monthly
– Every few months
– Other
• Select the device that you use the most to connect to the specified
SNS account.
– Personal Computer/Laptop/Desktop
– Tablet (e.g., iPad)
– Smartphone (e.g., iPhone)
– Other (please specify)
A.1.3 Questions About HI Sharing Behaviours in SNSs
1. For each HI category, how often do you share information in SNSs? [The
survey presented 8 HI categories with corresponding responses in the
form of a 5-point Likert scale with anchor points specified as: never,
sometimes, and always]
• Refer to Table A.1 for HI categories.
103
2. Specify other types of HI that you share and do not appear in the above
list.
3. Why do you share your HI in SNSs? [Select all that apply]
• Help others by sharing personal experience and knowledge
• Get benefits by receiving useful feedback from online contacts
• Seek help or social support
• Alleviate anxiety (Sharing HI makes me feel better and less stressed)
• Seek online interactions and make discussions
• Promote healthy living
• Other (please specify)
4. For HI that you do not share in SNSs, describe your choice by selecting
all reasons that apply from below.
• I do not want to be treated as “the sick” person by my online con-
tacts
• I do not want my online contacts to worry about me by receiving
bad news about my health
• I have different people in my online contact list and I prefer not to
share my HI with all of them
• My HI is personal and I do not share it with anyone
• I prefer to share my HI with my doctor
104
• I do not want my online contacts to know about my HI
• My online contacts are not interested in my HI
• I prefer to share my HI offline
• Other (please specify)
A.1.4 Questions About Privacy Risk Perceptions andBehavioural Responses
1. When sharing HI in SNSs, how much each of the following factors con-
tribute to the increase of your perceived privacy risks: [Responses vary
on a 5-point Likert scale with anchor points specified as: does not affect,
slightly affects, and strongly affects]
• The receiver(s) of the HI
• HI type and category
• The SNS platform where the HI is shared
• Your current physical or mental health condition (e.g., depressed,
healthy, suffer from chronic pain, etc.)
2. For every HI category (Table A.1), how do you evaluate your perceived
privacy risk level when sharing HI with the given four user categories:
• Select individual(s)
• Select group(s)
• Your entire contact list
105
• All other SNS users
[Select the perceived privacy risk level from a 5-point Likert scale with
anchor points: very low (or none), moderate, and extreme]
3. “Behavioural responses” refer to different actions undertaken by an in-
dividual to mitigate the perceived privacy risks.
Assume you want to share some HI in SNS: For each assumed privacy
risk level (very low, acceptable, moderate, high, extreme), what would
be your preferred behavioural response(s)? [select all that apply for
each risk level]
• Avoid risks by not sharing HI
• Minimize risks by manipulating shared HI (e.g., change or remove
personal and identifiable information)
• Minimize risks by filtering recipients (e.g., share with specific indi-
vidual(s) or group(s))
• Minimize risks by changing the used SNS (e.g., share HI in other
trusted SNS)
• Accept risks and share HI without any preemptive actions
106
A.2 Categorization of HI examples
Prior to conducting our survey, we reviewed the literature and found more
than 30 different examples of HI that people often shared online [20, 48].
They include but not limited to the following: mental or physical conditions,
healthy eating and dietary information, daily exercise routines, experience
with a specific doctor, previous surgeries, symptoms and side effects, insur-
ance company information, useful online articles, information about children
and elderly family members. We categorized the identified examples of HI
sharing according to their similarities and characteristics. To triangulate the
categorization, we used an online card sorting tool (OptimalSort)1 to run a
closed card sorting exercise, where we asked 11 participants, mostly friends
and colleagues, to group different HI examples into the predefined 8 cate-
gories. Seven participants grouped over 75% of the HI examples into cate-
gories similar to our predefined ones. Participants’ feedback was also incorpo-
rated toward refining/rephrasing the HI categories and the given HI examples
in order to avoid further ambiguity in the survey questionnaire and study ma-
terials. The finalized HI categories are presented in Table A.1.
1www.optimalworkshop.com
107
Table A.1: HI categories and examples.
Category HI Examples
Healthy livingdietary and healthy eating, physical exercise,
mental exercise
Own experienceexperience with: previous surgeries, treat-
ments, and symptoms
Useful found information books, articles, websites
Mental and emotional
health conditionsstressed, excited, depressed
Physical health conditions sick, injured, in good shape
Medical health recordspersonal information and address, physician
information, insurance policy, x-rays, reports
Experience of somebody
else
experience with: previous surgeries, treat-
ments, and symptoms
HI of people in your custody parents, children, and others
108
Appendix B
Supplementary Materials for the
Exploratory Study (Interviews)
The study was reviewed and approved by UBC’s BREB. Participants also re-
viewed the recruitment ad before sending their participation request through
the study webpage. After reviewing the study related information, participants
had to submit their information and consent to participate in the study.
109
! Recruitment!page,!v1.2!!!!!29/11/2013!1!
Sharing Health Related Information In Online Social Networks The purpose of this exploratory research is to investigate different aspects of sharing health related information (HRI) in online social networks (OSNs). While most people consider HRI to be private and personal in nature, recent studies showed increasing number of users who shared information about their health in OSNs (e.g., Facebook). In fact, sharing HRI is becoming a new application of OSNs. Whether if it is for the gained benefits of sharing HRI, or simply to help others and provide social support, there is a lack of understanding about different aspects of sharing HRI in OSNs.
How to participate? Your participation is important for our research. You will be interviewed about your opinions, thoughts, and experiences when using OSNs (e.g., Facebook). If you have one or more chronic health condition(s) (e.g., spinal cord injury), or you are a caregiver who provides healthcare support to an individual with chronic health condition(s), then you can participate in this study by clicking on the “Participate” button below:
Compensation: $25 per interview (approximately 1 hour)
Please review the following information carefully and provide your consent by clicking on "Submit" button at the end of the page
Principal Investigator The principal investigator of this research is Dr. Konstantin Beznosov from the Department of Electrical and Computer Engineering at the University of British Columbia. You can contact him at [email protected] or (604) 822 9181. Co-Investigator Sadegh Torabi ([email protected]) The co-investigator is a PhD student from the Department of Electrical and Computer Engineering at UBC. You can contact him at 604 827 3410. Purpose The purpose of this exploratory user study is to address the lack of understanding about different aspects of health related information (HRI) in online social networks (OSNs). While most people consider HRI to be private and personal in nature, recent studies showed increasing number of users who shared information about their health in OSNs. In fact, sharing HRI is becoming a new application of OSNs. Whether if it is for the gained benefits of sharing HRI, or simply to help others and provide social support, there is a lack of understanding about the privacy aspects of sharing HRI in OSNs. We will interview individuals with similar health conditions (e.g., spinal cord injury patients) to identify privacy and usability requirements. We will eventually suggest a set of privacy-preserving guidelines that could be used by OSN designers and application developers to enhance HRI sharing experience in OSNs. Study Procedures In this study, we will collect qualitative data by means of individual interviews. As a person with chronic health condition(s) (e.g., spinal cord injury), or a caregiver who is providing healthcare support for an individual with chronic health condition(s), you will take part in an individual interview about your OSN usage, online information sharing behaviours, health condition and health management activities, perceived privacy risks, and corresponding behavioural responses. The interview will be conducted in a mutually convenient time and location. It will last approximately 60 minutes, and with your permission, it will be audio-recorded and transcribed. The interview will also include the collection of some demographic 111
information (e.g., age, gender, education, etc.) through a designated online form, or by completing a printed version of the form upon request. A follow-up telephone call or email communication may be necessary to clarify issues arising from the discussion. Confidentiality All information will be kept strictly confidential. To ensure complete anonymity, pseudonyms (false names) will be substituted for your real name and the names of anyone else mentioned in the interview discussion. All stored media, transcripts and consent forms will be identified by code and kept in a locked filing cabinet in the project office. The only people that will hear the recorded interviews and/or see the transcripts will be the Principal investigator (Dr. Konstantin Beznosov), the Co-Investigator (Sadegh Torabi), and the transcriptionist. The information gained from this research will be anonymized and written up in publications and/or reports and will be shared with interested parties. You may request copies of these publications or reports about the findings of the study. The recorded interviews will be destroyed 5 years after publication of the original analysis and findings. Compensation, Benefits, and Potential Risks Your participation is important for our research. As a result of participating in this research, you will be compensated with the amount of $25 (CAD). Also, there might be some other benefits to you as a result of participating in this research, as discussion with others about your thoughts and experiences may help you to clarify your own thinking. Moreover, the outcomes of the study may help increase the awareness and gained knowledge about the studied topic. There might be some potential risks related to involvement in this research. It is possible that some topics discussed in the interview may raise new and/or previous emotional issues but you need only answer questions or express your views when you wish to do so. The investigator(s) will be respectful of your preferences. However, the investigator(s) are not equipped to provide counseling or professional advice. Contact for information If you have any questions or require further information about the project, you may contact the co-investigator (Sadegh Torabi at +1 604 827 3410). If you have any concerns about your treatment or right as a research subject, you may contact the Research Subject Information Line in the UBC Office of Research Services at 604 822 8598 or e-mail [email protected].
It is very important that your participation in this study is entirely voluntary and based on clear understanding. Furthermore, you may refuse to participate or withdraw from the study at any time. By clicking on the “Submit” button below you accept to participate in this study and indicate the following:
1. You are 19 years and older, 2. You are living in the Greater Vancouver region, 3. You maintain at least one active OSN account that you use
regularly, 4. You have one or more chronic health conditions (e.g., spinal cord
injury), or you are a caregiver who provides healthcare support to an individual with chronic health condition(s),
5. Your participation in this study is entirely voluntary and you may refuse to participate or withdraw from the study at any time without giving a reason and without any negative impact on received benefits and compensation, and
6. You have read and understood the provided information, and you consent to participate in this study [you can print or save a copy of this consent form for your records].
Provide your Name and Contact information and submit your request [Required fields are marked with * ] By clicking on "Submit" button below, you consent to participate in this study. Your contact information will be recorded and you will be contacted shortly. *"Name: " *"Email: " Phone: "
!
113
B.1 Interview guide and questions
Will collect qualitative data by means of semi-structured interviews. The inter-
view lasted between 60-90 minutes, and were audio-recorded and transcribed.
The interviews started by reviewing the consent form and the collection of
demographic information (age, gender, education, etc.). Then, a number of
research-related questions were asked, as shown in the following subsections.
A follow-up telephone call or email communication was made when neces-
sary to clarify issues arising from the discussion. The interview questions are
presented as following:
B.1.1 Health condition background
• What is the health condition you have?
• How/When did the health condition appeared or started the first time?
• How does the health condition affect your daily life?
• What are the challenges that you face due to the health condition you
have?
• How does the health condition you have affect your social life?
• Is there anything specific about your health condition that is of your
concern?
114
B.1.2 Health management
• How do you manage your health condition?
• How others (if any) are involved in your health management process?
• What is your relationship with doctors, physicians, and nurses?
• Do you have any concerns regarding your health management?
B.1.3 SNS usage and background
• How many SNS accounts do you maintain?
• How often do you log into your SNS accounts and what do you usually
do there?
• Who do you connect to using the SNSs? Who are your online friends?
• What do the SNS environments mean to you (e.g., Facebook)?
B.1.4 HI sharing on SNSs
• Have you ever shared HI in your SNSs? Why?
• Whom do you usually share your HI with?
• How do you think sharing HI could be helpful/beneficial to you or others
in your social network?
• When it comes to sharing HI, do you have specific preferences about the
type of the SNS where you share your HI in? Why?
115
• How do you think about existing privacy settings in SNSs?
B.1.5 A specialized SNSs for managing health conditions
• Have you ever considered using an SNS to manage, share, and seek HI?
• What do you expect from a specialized SNS that is used to help you
manage your health conditions and get connected to others?
• How do you define your privacy?
B.1.6 Study related feedback
Do you have any comments, suggestions or concerns related to this study? We
appreciate your constructive feedback?
B.2 Supplementary results
Participants came with different health issues. Nine participants suffered from
chronic pain and arthritis in different parts of their body. We interviewed two
quadriplegic participants with limited physical mobility, among whom one had
also suffered from chronic lung and heart diseases. We also interviewed an
HIV positive patient, who was infected as a results of an accidental needle
poke while doing his job as a paramedic. Finally, one participant had Neu-
romyelitis Optica (NMO), which is a rare disease that attacks the central nerve
system and causes blindness, paralysis, and other health issues. The remaining
participants suffered from a combination of mental and/or physical illnesses
116
(e.g., eating disorder and depression, arthritis and lung disease). More details
about participants’ health conditions are presented in Table B.1.
117
Table B.1: Participants demographics and health conditions. The firstcolumn represents participants’ ID.
ID Gender Age Health condition(s)
P1 M 38 chronic sciatica due to an accident
P2 M 59 back fracture and defective left knee
P3 M 31 severe arthritis in right hand due to a car accident
P4 F 68C4-C5 incomplete quadriplegic due to damaged
neck in a sport accident
P5 F 30 chronic depression
P6 F 21 curved spine and chronic back pain
P7 M 54C5-C6 quadriplegic due to a motor accident, and
chronic heart/lung disease
P8 M 38 chronic back pain
P9 F 42Neuromyelitis optica (NMO), episodes of blindness,
headaches, and fatigue
P10 M 37 osteoarthritis (deformed leg) and defective knee
P11 M 40L3-L4 fusion due to a work-related accident and
COPD (lung problem)
P12 M 59degenerative disk and brain injury (lost senses of
balance, taste, and smell)
P13 M 51 osteoarthritis in all joints
P14 F 39 eating disorder and post-traumatic stress disorder
P15 M 37 bipolar depression and anxiety
P16 M 48 post-traumatic stress disorder
P17 M 48 arthritis in hands and knees
P18 F 38degenerative arthritis in foot and ankle, anemia,
and depression
P19 M 50 HIV due to an accidental needle poke
P20 M 50 depression and chronic pain from broken neck
P21 F 35 herniated disks (L4-L5) with chronic pain
118
Appendix C
Supplementary Materials for the
Confirmatory Study (Online
Survey)
The study was reviewed by UBC’s BREB and a certificate of approval was
issued. To take part in the study, participants reviewed the recruitment ad
through MTurk and reviewed the study related information. Participants were
required to consent to participate in the online survey before answering survey
Sharing Health Information on Social Networking Sites
The Department of Electrical and Computer Engineering at University of British Columbia invites participants for a study on health information sharing behaviours on Facebook.
We require volunteers to participate in our online survey that takes no more than 30 minutes. All participants will receive the amount of 1 US dollar through Amazon Mechanical Turk.
The study requires the participant to satisfy the following conditions: (1) you must be 19 years and older; and (2) you must be an active Facebook user, that is a person who maintains a Facebook profile and uses the service to share/seek information with/from online contacts regularly.
If you would like to participate in this study, please follow the link below:
https://survey.ubc.ca/surveys/sadeq/phealth-v1-pilot/ For further information, please contact us at [email protected]
We asked participants to explain in their words why they think their experi-
ences were Positive. As presented in Table C.2, a total of 272 text responses
were analyzed and coded to represent participants’ positive experiences. Pos-
itive emotional and social support in the form of sympathy, empathy, and
139
prayers, were identified as the most common positive experiences among par-
ticipants. Useful recommendations and advice came second in the list, with
participants receiving feedback that positively helped them toward managing
their health conditions. Participants also described their positive experiences
by indicating that Facebook was used as an effective communication channel
for broadcasting information related to their health, while receiving timely
feedback from other social peers. Furthermore, participants benefitted from
their conversations with others in order to bring awareness to their health is-
sues and justify their behaviors whenever necessary. By sharing their HI on
Facebook, participants were able to find other social peers who had mutual
health experiences. Communicating with these social peers provided partici-
pants with valuable information/experiences while making them feel that they
belong to a group of understandable and easy to communicate people. Finally,
the two-way benefits of sharing HI on SNSs was easy to identify by going
through participants’ positive experiences in trying to help other people when-
ever possible.
140
Table C.2: Positive HI sharing experience. The first two columns repre-sent the coded category and related sub-categories. The last twocolumns represent the total number/percentage of positive codedevents under each category (272 total references).
As shown in Table C.3, participants identified a number of reasons for de-
scribing their prior HI sharing experiences to be Negative. Participants were
frustrated by the responses they received from their social peers who over-
reacted to their health problems and showed overwhelming and unnecessary
concerns. Participants were also agitated by the social peers who used their
shared HI in order to make judgments, spread rumours, gossip, or partici-
141
pated in insulting discussions. Furthermore, participants raised some privacy
concerns with respect to discussing their health issues in a semi-public en-
vironment lie Facebook, which occasionally led to oversharing their health
information without their permissions. Finally, while participants did not ap-
preciate the impractical recommendations and advice given to them by some
social peers, they felt lonely and unimportant when they received no support-
/replies from other social peers.
Table C.3: Negative HI sharing experience. The first two columns rep-resent the coded category and related sub-categories. The last twocolumns represent the total number/percentage of negative codedevents under each category (86 total references).
Category Sub-categoriesCoded
events(%)
People don’t un-
derstand
people overreact on health issues, feel
pity, create unnecessary worry, provide
responses that may increase anxiety
27 31
Negative social
impact
gossips, rumours, insulting discussions
and trolls, judgements, condescending
responses
22 26
Privacy concerns
public/open environment, people get
too involved/nosy, over sharing one’s
health information, receive spam/junk
17 20
Impractical
advice
impractical recommendations, advice,
and information12 14
Ignored postno replies to posts, no social support/in-