Help me, im fat! social support in online weight loss networks

Journal of Consumer Behaviour, J. Consumer Behav. 10: 332–337 (2011)Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/cb.374 Help me, I’m fat! Social support in online weight loss networks PAUL W. BALLANTINE* and RACHEL J. STEPHENSONDepartment of Management, University of Canterbury, Private Bag 4800, Christchurch 8140, New Zealand Social networks have become an increasingly common way for people to share information and seek emotional support for issues surround-ing weight loss. This study aims to explore how users of a commercial social networking site who are focussed on weight loss give and/orreceive social support to/from other users. The authors use quantitative data from 145 members of the Weight Watchers Facebook page toexplore how social support is both given and received, and the communication style by which this is achieved. This study reveals threegroups—Passive Recipients, Active Supporters and Casual Browsers. Passive Recipients receive a high level of informational and emo-tional support but do so by being passive communicators. Active Supporters also receive a high level of both informational and emotionalsupport yet are more active in their communication style. Casual Browsers receive little social support and exhibit a passive communicationstyle. Thus, the authors find evidence that even though members of a social network may share a common interest, the way members chooseto participate and interact, and the benefits they accrue by doing so differ substantially.
Copyright 2011 John Wiley & Sons, Ltd.
networking sites have to offer. For example, Barak et al.
(2008) highlighted several attributes of the Internet that can The Internet has become an increasingly popular tool for indi- be described using a common concept, disinhibition, which viduals wishing to seek information, share experiences, ask allows us to understand why individuals differ in what they questions and provide emotional support about health issues say and do online, compared to what they say and do in a (Turner et al., 2001; Eysenbach et al., 2004; Jayanti and face-to-face setting. The reasons for this difference in behav- Singh, 2010). A key reason for this is that the Internet has iour include factors like anonymity, invisibility, delayed reac- many attributes that can help people feel more comfortable tions, solipsistic introjections and the neutralising of status.
with openly expressing their problems and concerns (Hwang All these factors can facilitate support for those users of an et al., 2010; Wu et al., 2010). Within the broad category of online community who are experiencing similar difficulties, health issues, online support groups and communities have such as issues surrounding weight loss.
become an increasingly common way for individuals to share The potential role of the Internet as a mechanism for information and seek emotional support for issues surrounding social support on health-related conditions was highlighted weight loss (Wright et al., 2010). Many of these online by Wangberg et al. (2007). The authors examined the rela- exchanges are facilitated by social networking sites such as tionship between Internet use, social support and subjective Facebook, which allow users to leave their network friends health, with their results suggesting that the Internet has both private and public messages, share photos and become mem- a direct positive relationship to subjective health, as well as bers of a wide range of interest groups.
an indirect positive relationship, mediated through social Social networking sites have also helped commercial orga- support. Moreover, the role of weak social ties, which occur nisations meet consumer demands by giving them the oppor- when people who are not interpersonally close interact in a tunity to become more personal, provide consumers with limited way within certain contexts, was examined by Wright support, and get direct information about their consumers’ et al. (2010) in terms of the role they have on the perceived wants (Moran and Gossieaux, 2010). Health organisations stress of participants in a health-related online support group.
such as Weight Watchers are also using these sites to provide Contrasted with strong social ties, which are present in close a means of communication for their members to give and re- relationships (i.e. family and friends), the authors found that ceive social support. The purpose of this research is to explore participants preferred support from weak ties while online, as how users of a commercial social networking site who are they provided benefits such as access to different viewpoints, focussed on weight loss give and/or receive social support objective feedback, reduced risk and reduced role obliga- to/from other users. We achieve this through a quantitative tions. Due to the sensitivity often associated with health- examination of users of the Weight Watchers Facebook page.
related issues, strong ties were found to act as a barrier togaining social support.
Social support can be described as a type of interpersonal Coulson et al. (2007) observed that there has been a prolifera- exchange that can make an individual feel either loved, tion of health-related information on the Internet. Reasons why esteemed, accepted, valued or motivated (Teoh et al., 2009).
individuals seek health-related information on the Internet can In the context of weight loss, social support has been linked be attributed to some qualities that the Internet and social to better health outcomes and as having a positive effect onweight loss behaviour and weight maintenance (Teoh et al., *Correspondence to: Paul W. Ballantine, Department of Management, Uni- 2009). For example, Wing and Jeffery (1999) explored the versity of Canterbury, Private Bag 4800, Christchurch 8140, New Zealand.
E-mail: paul.ballantine@canterbury.ac.nz benefits of social support for weight loss and maintenance Copyright 2011 John Wiley & Sons, Ltd.
Social support in online weight loss networks by assigning a standard behavioural weight loss treatment to Hwang et al. (2010) identified four main types of social sup- participants who were grouped with friends and family, and port: informational, emotional, instrumental and appraisal, the same treatment to those who participated alone. They and this study will focus on the two most frequent types given found that those with social support (the grouped participants) and received within an online support group (Buchanan lost more weight and maintained their weight loss, compared and Coulson, 2007): informational and emotional. Infor- to those who participated alone. More recently, Moisio and mational support includes activities such as advice giving, Beruchashvili (2010) described weekly Weight Watchers sup- referral to experts, situation appraisal and teaching (Coulson port groups as being both a spiritual and therapeutic compan- et al., 2007), while emotional support includes aspects such ion, which gave them an aura of indispensability in members’ as empathy, concern, caring, love and trust (Dalgard, 2010).
lives. However, the above studies focussed on the role of Beyond the types of social support given and received, face-to-face support groups, and it is not yet known if social users within an online community also differ with regard to support plays a similar role in online environments.
their preferred type of communication, in that users can The Internet can be considered both an active communica- adopt either active or passive roles. The purpose of this study tion medium and a passive one (Wangberg et al., 2007). Sim- is to explore how users of a commercial social networking ilarly, the way that individuals give and/or receive support site who are focussed on weight loss give and/or receive so- online can further be described as being active or passive.
cial support to/from other users. We do this by exploring Active social support occurs when participants are interact- how users of the Weight Watchers Facebook page give ing with others in their online social network. An example and/or receive both informational and emotional support to/ of this interaction is when a participant comments on a mes- from other members and the active or passive communica- sage another participant has written on the ‘wall’ of an online tion roles they take in doing so. By addressing this aim and social network. Within online communities, authors have in keeping with the theme of this special issue, we hope to shown that the reciprocity of social interactions positively shed further light on the different types of interaction that affects loyalty towards an online community (Chan and Li, can occur in online social networks.
2010; Shen et al., 2010), and how the shared values of com-munity members can enhance both trust and relationshipcommitment (Wu et al., 2010). In contrast to those memberswho actively participate, online communities can also pro- vide a learning function (and means of social support) forthose individuals who decide to read and not contribute to This study took a quantitative approach, where an online the social interactions taking place, and who are often survey was administered to users of the Weight Watchers Facebook page. To help recruit participants, an introductory Lurking behaviour, wherein people browse websites in a message was posted on the wall of the Weight Watchers read-only mode, usually occurs because people want to learn Facebook page that outlined the purpose of this study. The about a community or topic, or want to gain a sense of message was posted approximately two times each day dur- belonging (Rafaeli et al., 2004). In essence, a person who ing the data collection period to ensure the continuous expo- lurks still receives social support, albeit passively, suggesting sure of the introductory message, given the high volume of that they should be considered when investigating how peo- posting activity by members. The introductory message also ple receive support in an online social network. The effec- included a survey link, which provided further information tiveness of passive support has been evidenced by authors on the survey and a consent form. To encourage participa- like Hwang et al. (2010), who found that the weight loss tes- tion, a random prize draw was held where three partici- timonies of others played a prominent role in participants’ pants who had completed the survey were awarded a $50 Amazon.com voucher. After a survey period of two weeks, The notion of passive support is similar to the idea of the introductory message was removed from the wall of the parasocial relationships, a one-sided relationship that can oc- cur between a media user and the media being consumed,which has been examined in the context of online communi-ties (Ballantine and Martin, 2005). Passive support is also conceptually similar to the description of two community The survey was formulated using Qualtrics, which provides member types, devotees (i.e. members who lack an interest a platform for designing, distributing and evaluating online in other users, yet have considerable interest in the focal surveys. The survey was divided into three main sections activity) and tourists (i.e. members who have only a passing and was designed to take approximately 5 minutes to com- interest in the focal activity and little interest in other users), plete. The first section covered general questions about why which Kozinets (1999) outlined in his seminal research on and how participants use the Weight Watchers Facebook virtual communities of consumption. With both devotees page. The second section covered questions based on the and tourists having weak social ties to an online community, participants’ communication style and the types of support it can be argued that many social network users may prefer to they sought from an online weight loss network. The last observe rather than interact in network discussions.
section concluded with questions about the participants’ Online communities and social networks have the ability demographics (i.e. gender and age), Facebook usage and to provide different types of social support to their members.
Copyright 2011 John Wiley & Sons, Ltd.
fully complete the questionnaire. In terms of demographic Several scales were developed specifically for this research, as characteristics, 98 per cent of participants were female. Parti- the constructs were not developed in the literature. An eight- cipants aged 31–40 were most common (37%), followed by item measure was developed to capture the informational sup- 21–30 (23%), 41–50 (21%) and 51–60 (14%). Most partici- port construct based on the exploratory research of Hwang pants used Facebook for 1–10 hours per week (58%), followed et al. (2010). Emotional support was also measured with an by 11–25 hours (30%), with 85 per cent of participants rating eight-item scale developed from Hwang et al. (2010). Com- their Internet ability as being good or very good.
munication style was measured with an eight-item scale devel- Principal components analysis (with Varimax rotation) oped from the work of Rafaeli et al. (2004), who examined was used to assess the underlying structure of the measures active and passive users within online communities. All items used. Based on this analysis, a three-factor solution emerged.
were measured using a five-point Likert scale anchored These factors explained 65 per cent of the variance, with all strongly agree to strongly disagree, with exploratory factor eigenvalues being over 1, all items loading heavily onto analysis being used to refine the scales. The scale items used one of the factors, and with all factors also being easily inter- in this study are provided in Table 1.
pretable. Table 2 provides the factor loadings for each of thescale items. The dimension of informational support wasfound to contain eight items, emotional support containedfive items, while communication style contained three items.
Measures were then assessed for their internal consistencyusing the Cronbach’s alpha reliability procedure (Cronbach, The online survey website was left open for a period of two 1951). Table 3 reports the means, standard deviations and weeks during September 2010, and at the end of this time, reliabilities of all the measures used in this study.
168 surveys were submitted. Of these, 145 surveys were suit- To determine if participants held common perceptions able for inclusion in the final sample, with 23 surveys being about informational support and emotional support, and com- removed due to systematic response patterns or a failure to munication style, cluster analysis was used to uncover any sim-ilarities that might be used to help identify distinct groupswithin the sample. These groups were identi two-step procedure (e.g. Milligan, 1980; Hair et al., 2006).
First, hierarchical cluster analysis using Ward’s method was I use this page to gain information about how I should be exercising applied to the mean item scores of the three factors. Adopting I find out valuable dietary information on this page.
‘stopping rule’ (e.g. Hair et al., 2006), the changes in the This page helps me understand which foods I should and shouldn’t within-cluster sum of squares suggested three clusters as an initial solution. Second, K-means cluster analysis was then This page provides me with effective weight loss information.
used to fine-tune the clusters assuming three groups. The mean If I have a question related to losing weight, I can usually find the factor scores for each of these three groups are provided in I use this page for information about suggested activities to lose were developed based upon these mean ratings, and the results The information provided by other members of this page helps me of a series of analyses of variance (using post-hoc Scheffe tests) to identify any pairwise differences in mean factor scores.
I get good tips on how to lose weight from this page.
The members of the first group (Passive Recipients) were found to have the highest informational support ratings. They Using this page makes me feel important.
I feel that members of this page care about me as they can relate to had similar ratings on emotional support to the Active Sup- porters but with high communication style scores (indicating I gain a feeling of acceptance from using this page.
a preference for a passive communication style). Overall, I use this page to receive comfort from others when I am people in this group received a high level of informational disappointed with my weight loss outcomes.
and emotional support from the Weight Watchers Facebook When I want to express my feelings, I use this page.
People on this page give me encouragement to lose weight.
page but did so by being passive members of this social net- People using this page are sympathetic towards me.
work. There were 68 participants in this group (46.9% of the Seeing the success of others on this page helps me stay on my sample). The second group (Active Supporters) also per- ceived high informational and emotional support benefits but had the lowest communication style scores (indicating a I prefer to observe rather than post messages on this page.
I use this page by observing discussions that are taking place.
preference for an active communication style). Thus, they I would classify myself as an interactive user of this page.
received a high level of both informational and emotional I like to express my opinions on this page.
support from the Weight Watchers Facebook page but did I do not interactively communicate on this page as I have nothing to so by being active members of this social network. There were 47 participants in this group (32.4% of the sample).
If I have advice to give in regards to what someone has posted, Iwill comment.
The final group (Casual Browsers) had the lowest informa- I would classify myself as a user who browses this page.
tional and emotional support scores, while also having the If I have a weight related question I want answered, I will post a highest communication style scores (indicating a preference question on this page to get a response.
for a passive communication style). Overall, they received Copyright 2011 John Wiley & Sons, Ltd.
Social support in online weight loss networks Table 3. Means, standard deviations and reliabilities of measuresused understand which foods Ishould and shouldn’t beeating to lose weight.
This page provides me Table 4. Final cluster centres (mean factor scores for each cluster) with effective weightloss information.
information about how Ishould be exercising tolose weight.
little social support from the Weight Watchers Facebook page, while being passive users of this social network. There were 30 participants in this group (20.7% of the sample).
To further understand the characteristics of each of the three groups, the demographic information provided by par- ticipants was used to see if any demographic differences existed between the groups. No differences were found in terms of participant gender, age, Facebook usage or level of Internet ability. Additional comparisons were made be- answers on this page.
I gain a feeling of tween the groups in terms of their Facebook usage. When asked how long they had been a member of the Weight Watchers Facebook page, Active Supporters were found to have been members the longest (w2 = 12.733, p = 0.047) when compared to both Passive Recipients and Casual Brow- as they can relate to whatI am experiencing.
sers. When asked to rate their level of satisfaction with the Weight Watchers Facebook page, Passive Recipients were the most positive, followed by Active Supporters and Casual Browsers (F = 9.432, p = 0.000). Active Supporters were found to be the most likely to post messages on the Weight question I want answeredI will post a question Watchers Facebook page, followed by Passive Recipients and Casual Browsers (F = 27.346, p = 0.000). Similarly, Active Supporters were found to be more likely to leave comments on the messages posted by others, followed by Passive Recipients and Casual Browsers (F = 21.185, p = 0.000).
when I am disappointedwith my weight lossoutcomes.
I use this page by observing discussionsthat are taking place.
The Weight Watchers Facebook page examined in this study provides an online social network for members to give and/or receive social support to/from other users. When the type of social support given and/or received was examined in con- junction with communication style, three distinct groups of users were found to exist. Passive Recipients received a high level of informational and emotional support from theWeight Watchers Facebook page but did so by being passivemembers. Active Supporters also received a high level ofboth informational and emotional support, but unlike Passive Copyright 2011 John Wiley & Sons, Ltd.
Recipients, they did so by being active members of this social network initially perceived the introductory message social network. Finally, the Casual Browsers received little as spam, which may also have led some users to decide not social support from the Weight Watchers Facebook page and were passive users of this online forum.
While this study provides an initial understanding of how No demographic differences were found between any of users of a commercial social networking site who are the three groups. However, Active Supporters were found focussed on weight loss give and/or receive social support to have been members of the Weight Watchers Facebook to/from other users, some constructs were not included that page the longest. Similarly, this group of users was also could shed further light on this topic. For example, dimen- found to be the most likely to post messages in this social sions of trust (e.g. Wu et al., 2010) could be included in order network, as well as being the most likely to respond and to understand how members of the three groups identified in provide comments on the messages posted by others. In this this study perceive the information provided by others. This respect, the actions of this group support the findings of Chan would allow researchers to understand why some members and Li (2010) and Shen et al. (2010), who showed that the did (or did not) value the informational and emotional support reciprocity of social interactions (i.e. more involvement with members of the group) can positively affect loyalty in anonline community.
This study also provides some support to the findings of Wright et al. (2010), as two of the groups in this study (Passive Recipients and Casual Browsers) exhibited a styleof communication which suggests that they did not feel The results of this study highlight how a social network can strong social ties to the Weight Watchers Facebook page.
provide informational and emotional support to its members, Thus, rather than actively post and comment on the messages even though users may differ in how they provide this support posted by other group members, both of these groups pre- to other members (if indeed they do). Thus, while some mem- ferred to browse and observe the messages posted by others.
bers of a social network may take an active role in providing However, even though they preferred to take a passive com- as well as receiving social support, the results of this study munication role, the Passive Recipients clearly valued the suggest that many members also accrue informational and informational and emotional support they experienced from emotional benefits by taking the role of a passive recipient.
this social network. In this respect, the Passive Recipients Indeed, the benefits enjoyed by being a member of a social group took the role of lurkers, in that the social network network can still be obtained by those users who choose to provided a learning function, as well as a means of social act as passive observers of the social exchanges of others.
support, even though they typically did not return the social This study also adds to the literature on how the Internet can support they received from the Weight Watchers Facebook serve as a mechanism for health-related outcomes. Although page. This finding shares similarities with the arguments pro- the focus of this study was on how users of a commercial social vided by Ballantine and Martin (2005) regarding parasocial networking site who are focussed on weight loss give and/or interaction in online communities. Even though the relation- receive social support to/from other users, the findings of this ship between Passive Recipients and those members who study might also be applied in other online contexts where posted on the social network was typically one-sided and social support can play a key role. For marketers interested in non-reciprocated, it was evident that they valued the infor- understanding how consumers use social media, this study mational and emotional support that was available. This highlights how consumers may obtain beneficial outcomes usage of a social network is also similar to the devotees iden- from using an online social network, even though they may tified by Kozinets (1999): members who have little interest in appear to be taking a non-active role. With many companies other users yet have considerable interest in the activity being choosing to have an online presence through social networking discussed. In contrast, the smallest group identified in this sites such as Facebook, it is important that such companies are study, the Casual Browsers, was conceptually similar to the aware of the different ways that consumers may choose to tourists described by Kozinets (1999). Thus, while they had interact and provide support to each other, and the benefits they little interest in taking an active communication role, the results suggest that they gained little social support fromthe Weight Watchers Facebook page.
Limitations and future researchThere were two main difficulties associated with this study Ballantine PW, Martin BAS. 2005. Forming parasocial relation- that may have acted as limitations. First, the high number ships in online communities. Advances in Consumer Research of postings on the Weight Watchers Facebook page meant that the introductory message posted to recruit participants Barak A, Boniel-Nissim M, Suler J. 2008. Fostering empowerment was often pushed off the bottom of the wall after a relatively in online support groups. Computers in Human Behavior 24(5): short period of time. To resolve this issue, the message was Buchanan H, Coulson NS. 2007. Accessing dental anxiety online posted twice daily over a two-week period, yet it cannot be support groups: an exploratory qualitative study of motives known how many members of this social network were and experiences. Patient Education & Counseling 66(3): exposed to the invitation. Secondly, some members of this Copyright 2011 John Wiley & Sons, Ltd.
Social support in online weight loss networks Chan KW, Li SY. 2010. Understanding consumer-to-consumer Moran E, Gossieaux F. 2010. Marketing in a hyper-social world: the interactions in virtual communities: the salience of reciprocity.
tribalization of business study and characteristics of successful Journal of Business Research 63(9/10): 1033–1040.
online communities. Journal of Advertising Research 50(3): Coulson NS, Buchanan H, Aubeeluck A. 2007. Social support in cyberspace: a content analysis of communication within a Rafaeli S, Ravid G, Soroka V. 2004. De-lurking in virtual Huntington’s disease online support group. Patient Education communities: a social communication network approach to measuring the effects of social and cultural capital. Pro- Cronbach L. 1951. Coefficient Alpha and the internal structure of ceedings of the 37th Hawaii International Conference on tests. Psychometrika 16(3): 297–334.
Dalgard OS. 2010. Social support: definition and scope. Available Shen Y, Huang C, Chu C, Liao H. 2010. Virtual community loyalty: an interpersonal-interaction perspective. International Journal of Electronic Commerce 15(1): 49–73.
Eysenbach G, Powell J, Englesakis M, Rizo C, Stern A. 2004. Health Teoh AN, Chia MSC, Mohanraj V. 2009. The comparison between related virtual communities and electronic support groups: system- active and passive types of social support: the emotional atic review of the effects of online peer to peer interactions. British responses. Journal of Applied Biobehavioral Research 14(2): Medical Journal 328. doi:10.1136/bmj.328.7449.1166.
Hair JF, Black WC, Babin BJ, Anderson RE, Tatham RL. 2006.
Turner JW, Grube JA, Meyers J. 2001. Developing an optimal match Multivariate Data Analysis. Pearson Prentice Hall: Upper Saddle within online communities: an exploration of CMC support communities and traditional support. Journal of Communication Hwang KO, Ottenbacher AJ, Green AP, Cannon-Diehl MR, Richardson O, Bernstam EV, Thomas EJ. 2010. Social support Wangberg SC, Andreassen HK, Prokosch H, Santana SMV, in an Internet weight loss community. International Journal of Sorensen T, Chronaki CE. 2007. Relations between Internet use, socio-economic status (SES), social support and subjective Jayanti RK, Singh J. 2010. Pragmatic learning theory: an inquiry- health. Health Promotion International 23(1): 70–77.
action framework for distributed consumer learning in online Wing RR, Jeffery RW. 1999. Benefits of recruiting participants with communities. Journal of Consumer Research 36(6): 1058–1081.
friends and increasing social support for weight loss and mainte- Kozinets RV. 1999. E-tribalized marketing?: the strategic implica- nance. Journal of Consulting and Clinical Psychology 67(1): tions of virtual communities of consumption. European Man- Wright KB, Rains S, Banas J. 2010. Weak-tie support network Milligan GW. 1980. An examination of the effect of six types of preference and perceived life stress among participants in error perturbation on fifteen clustering algorithms. Psychometrika health-related, computer-mediated support groups. Journal of Computer-Mediated Communication 15(4): 606–624.
Moisio R, Beruchashvili M. 2010. Questing for well-being at Wu J, Chen Y, Chung Y. 2010. Trust factors influencing virtual Weight Watchers: the role of the spiritual-therapeutic model in community members: a study of transaction communities. Jour- a support group. Journal of Consumer Research 36(5): 857–875.
nal of Business Research 63(9/10): 1025–1032.
Copyright 2011 John Wiley & Sons, Ltd.

Source: http://ils.unc.edu/bmh/courses/706/Ballantine2011-SocialSupportWeightLoss_review.pdf