Computer-Mediated Communication and Social Support.

An extensive breadth of research has analyzed social support in face-to-face (FtF) conditions and reported its ample spectrum of positive effects on well-being by allowing recipients to cope better with stressful and problematic events in their lives (e.g., Burleson, 2009; High & Dillard, 2012). Furthermore, more recent studies have been evaluating how social support theories and its effects equate in online environments due to an increasing number of people turning to online social support networks due to the anonymity, heterogeneity, objectivity and asynchronicity such environments provide (e.g. Kim et al, 2012; Yoo et al., 2014; Wright et al., 2013). Anonymity and privacy provided by computer-mediated environments have been found to be key factors to facilitate self-disclosure and motivate individuals to reach out for support and information without fears of being judged, secluded or victimized (Rains, 2013; DeHaan et al., 2013). Moreover, computer-mediated social support (CMSS) has allowed individuals – especially those with health-related and stigmatized conditions – to compensate for a lack of access to information and ability to form offline relationships with others (often weak-tied) who share the same condition. In addition, CMSS allows individuals to seek and provide support more easily without feelings of embarrassment or limitations related to gender attributions and time-space constraints (Spottsswood et al., 2013; DeHaan et al., 2013, McKenna & Bargh, 1998; Robinson et al., 2009).

Overall, CMSS research has put forward many interesting findings. For example, it has been found that emotional well-being and social support satisfaction are moderated by communication competence (how skillful a person is in communicational interactions). Yoo et al. (2013) and Wright et al. (2014) found that social support satisfaction and well-being are positively affected by recipient’s communication competence so that the higher communication competence, the more social support satisfaction and well-being. Another research reports that the level of emotional expression (emotional bandwidth) in support seeking affects support provision in the sense that emotionally charged self-disclosures (high level emotional bandwidth) from the part of the support seeker do not necessarily equate to receiving more supportive messages; but rather, that the willingness to provide support is mediated by the sex, level of perceived sense of community and preference for online social interaction (POSI) from the part of the support provider. Results of this research suggest that only females and those with a strong POSI and a sense of community are willing to comfort individuals who disclose their distress through high levels of emotional bandwidth (High et al., 2014). Regarding preferences for CMSS and based in self-consciousness theory, Lim et al. (2012) posit that support seekers seem to prefer CMC mediums over FtF when their perceived inter-personal cost is rather high. In other words, support seekers would prefer CMC when they perceive that they may fail to meet others’ expectations.

Although CMSS research has been able to assess many effects of online network use on supportive communication, sampling mechanisms and self-report data collection instrumentation have been constantly mentioned as two of the most important limitations in this type of research. Problems in sampling have arisen due to the limited number of participants and the difficult task to enroll and analyze individuals who are dealing with stressors, especially if the stressors are related to health debilitating or stigmatized conditions (e.g. Lawlor et al., 2014; Crowson & Goulding, 2013). One way study designs have addressed this limitation is by having participants fill online surveys anonymously (e.g. Bockting et al., 2013). Yet, researchers have mentioned that this type of online recruitment may be contacting individuals who are already coping fairly well with their conditions and thus limit the generalizability of findings. Another problem in sampling is that the majority of participants in many studies have been female which may mean that recruitment mechanisms are underrepresenting men in samples (e.g. Wright & Rains, 2013). Finally, data collection has relied heavily on self-report instrumentation (i.e. surveys, questionnaires) which have been linked to error and lack of objectivity which may lead in turn to inaccuracies in results  (DeHaan & Mustanski, 2013; Bockting et al., 2013; Wright et al, 2013; Berger et al., 2005).

In order to address a few of these limitations, efforts have been made to conduct data analysis employing automated methodologies in online networks that do not rely on traditional recruitment procedures and self-report data. In this sense, social support communication scholars have employed content analytic software like InfoTrend (e.g. Yoo et al., 2013; Kim et al., 2012) in order to automatize the measurement of support exchange levels (by identifying the number of supportive messages provided and received in a forum for a particular patient) in an online community, but then again, these methodologies have been limited by the constraints imposed by the use of self-report instruments at the time of assessing impact of level of support on sentiment change in patients which has, in turn, limited the number of participants (to about 236 at best).

Working in this direction, informatics researchers have tried to improve this automated methodologies and tackle the need for self-report data, by employing text-mining and machine learning theories in the analysis of user behavior in online social networks (e.g. Zhao et al., 2014; Wang et al, 2014). For example, Zhao et al. (2014) has employed text-mining and data mining classifiers (artificial intelligence) to identify influential supporters in an entire online social support community based on message features and emotional outcomes of support seekers. More specifically, the methodology of Zhao’s study consisted in training classifiers with an example of 298 randomly selected forum posts with negative and positive sentiments. Positive and negative sentiment posts were differentiated by several lexical and style features of the posts (exclamation marks, question marks, paralinguistic symbols, slang, etc.). For example, the words “happy”, “joy” and “lol” as well as the emoticon “J” would belong to a positive sentiment message whereas the words “disappointed” and “painful” as well as the emoticon “L”would belong to a negative sentiment message. Next, a comparison between sentiment strength between the thread originator’s first post and the subsequent first self-reply was used to set a metric (influential responding reply or IRR metric) that would determine the strength of intermediate responding replies (support messages from other users of the forum in response to the thread originator’s post) in order to identify influential users in a forum in the sense that their posts would be the ones that generated the most significant sentiment change in threads overall. Consequently, this approach helps identify providers whose enacted support leads to the greatest recipient’s emotional improvement.

A limitation of this automated methodology, however, is that it relies on observational data and therefore is not able to fully address causal effects regarding original support seeker’s sentiment. In other words, the change in seeker’s sentiment could have been not caused by the intermediate thread posts necessarily, but rather by other factors such as communication through other channels like private messaging, e-mail, offline interventions, etc. Nonetheless, the implications of this type of research are interesting and have already laid down the foundations for strategies that could be used by online support network managers in order to identify potential good supporters in an efficient way at any given moment and state of an ever-changing online social network structure. Once good supporters are identified, online community managers could further encourage their involvement and participation in forums via public recognition of their contributions to the online community (e.g. virtual badges, awards, prizes, etc.)

Future research in this line of research may serve to strengthen and broaden the CMSS communication literature. For example, effects of features of supportive messages (verbal person centeredness, length, etc.) as well as gender attributions (e.g. High & Solomon, 2014; Spottswood et al., 2013) could be further assessed and validated by analyzing the posts generated by the most influential support providers. In a similar way, studies analyzing the most common types of support based on the optimal matching model and type of stressor (e.g. Rains et al., 2015) could be validated by the rapid identification of type of support of the most influential user’s posts in online forums that deal with different type of stressors. Finally, the scarce social support literature on support seeking could be strengthened by further analyzing support seeker’s initial posts in order to detect message features that provoke the best supportive responses (those that draw the attention of the most influential users) and that ultimately led to the greatest and most positive emotional changes.

Although automated massive data extraction/analysis mechanisms can lead to potentially interesting findings, another way researchers from communication and informatics fields could advance CMSS literature is by exploring supportive communication patterns in mobile communication platforms. Based on social awareness streams (SAS) literature (Naaman et al., 2010), which describes the way information is produced and consumed by mobile devices (e.g. semi-public short messages) and theories that describe selective attention and multitasking (e.g. limited capacity model, staccato attention), communication scholars have started to analyze social supportive communication enacted in SASs; also known as staccato social support (SSS). This type of social support refers to mobile-mediated supportive communication that is characterized by individuals who are processing information while multitasking (doing multiple activities at the same time). This cognitive processing of information is known as staccato attention (Hembrooke, 2003) and is characterized by unobtrusive bursts of short mediated behaviors such as skimming a Facebook newsfeed or sending a tweet that do not disrupt daily activities. This type of behavior is expected from individuals using smartphone technology which allows them to communicate online while on the move.

In spite that this type of frequent online communication may be replacing traditional local social support networks (Stefanone et al, 2012), few initial attempts to study SSS have been published, Adams et al. (2014), for instance, developed a mobile application (VERA) that encourages users to post their health decisions in a SAS. When making a health-related decision, the user would start VERA and take a photo that would document the decision. The user then records a subjective healthiness rating, enters an optional text caption, sets a binary option which tell whether the depicted action was performed or not, and finally selects a picture describing how the user feels about the health decision. Other users can see the post on their newsfeed and write messages to the original poster. The pilot of VERA was tested with participants who downloaded the app from Google Play or Apple store and their interactions were monitored for a span of 4 weeks. Message content was analyzed and categorized according to the type of support provided (informational, tangible, esteem, emotional or network). Results of this study report a significant frequency of informational and esteem support production. Given that the great majority of users posted about health concerns which were fairly under the recipient’s control (e.g. weight-loss, general well-being habits, etc.), the high production of informational support validates the optimal matching model proposed by Cutrona and colleagues. However, the high production of esteem support contrasts with the same optimal matching model which posits that nurturant types of support (emotional and esteem) are more adequate for uncontrollable stressors. Furthermore, the high prevalence of esteem support also differs with findings of meta-analytic reviews of supportive messages in online social networks, which indicate that emotional and informational support messages are produced more frequently than esteem support messages (Rains et al., 2015). A possible explanation for the results could be that SSS may preclude certain styles of interaction due to staccato attention (e.g. no long narratives, users trained to “like” statuses). Another explanation could be that features of the medium (e.g. small keyboards of mobile phones) make the production of esteem support more prevalent in mobile communication platforms. Future research on SSS should validate these findings and further assess the impact of messages (both positive and negative) on recipients. In addition, lurking (not actively participating in an online community) could also be analyzed as to determine whether lurkers may still benefit from SSS environments as seems to be the case in traditional CMSS platforms given that it has been found that lurking may have similar empowering and positive effects to those accrued to active participants without having to actually communicate (post messages) with others in the online community (Phoenix & Coulson, 2010; Mickelson, 1997).

Evidently, interdisciplinary research is already leading to a better understanding of how people communicate support. Furthermore, findings are aiding design and development of customized social support computer applications as well as more efficient automated data collection/analysis mechanisms that address and minimize the limitations of traditional methodologies.

References

Adams, Phil, Eric PS Baumer, and Geri Gay. “Staccato social support in mobile health applications.” Proceedings of the 32nd annual ACM conference on Human factors in computing systems. ACM, 2014.

Albrecht, Terrance L., and Mara B. Adelman. Communicating social support. Sage Publications, Inc, 1987.

Berger, M., Wagner, T. H., & Baker, L. C. (2005). Internet use and stigmatized illness. Social science & medicine, 61(8), 1821-1827.

Bockting, W. O., Miner, M. H., Swinburne Romine, R. E., Hamilton, A., & Coleman, E. (2013). Stigma, Mental Health, and Resilience in an Online Sample of the US Transgender Population. American journal of public health, 103(5), 943-951.

Burleson, B. R. (2009). Understanding the outcomes of supportive communication: A dual-process approach. Journal of Social and Personal Relationships, 26(1), 21-38.

Cutrona, Carolyn E., and Daniel W. Russell. “Type of social support and specific stress: Toward a theory of optimal matching.” (1990)

Crowson, M., Goulding, A., (2013), Virtually Homosexual: Technoromanticism, demarginalisation and identity formation among homosexual males. Computers in Human Behavior.

DeHaan, S., Kuper, L. E., Magee, J. C., Bigelow, L., & Mustanski, B. S. (2013). The interplay between Online and offline explorations of identity, relationships, and sex: A Mixed-methods study with LGBT youth. Journal of Sex Research, 50(5), 421-434.

Hembrooke, H., and Gay, G. The laptop and the lecture: The effects of multitasking in learning environments. J of Computing in Higher Ed. 15, 1 (2003), 46–54.

High, A. C., & Solomon, D. H. (2014). Communication Channel, Sex, and the Immediate and Longitudinal Outcomes of Verbal Person-centered Support. Communication Monographs, 81(4), 439-468.

High, A. C., & Dillard, J. P. (2012). A review and meta-analysis of person-centered messages and social support outcomes. Communication Studies, 63, 99-118. doi: 10.7080/10510974.2011.598208

High, A. C., Oeldorf-Hirsch, A., & Bellur, S. (2014). Misery rarely gets company: The influence of emotional bandwidth on supportive communication in Facebook. Computers in Human Behavior, 34, 79-88. doi: 10.1016/j.chb.2014.01.037

Kim, E., Han, J. Y., Moon, T. J., Shaw, B., Shah, D. V., McTavish, F. M., & Gustafson, D. H. (2012). The process and effect of supportive message expression and reception in online breast cancer support groups. Psycho‐Oncology, 21(5), 531-540.

Lawlor, A., & Kirakowski, J. (2014). Online support groups for mental health: A space for challenging self-stigma or a means of social avoidance?. Computers in Human Behavior, 32, 152-161.

Lim, V. K., Teo, T. S., & Zhao, X. (2013). Psychological costs of support seeking and choice of communication channel. Behaviour & Information Technology, 32(2), 132-146.

McKenna, K. Y., & Bargh, J. A. (1998). Coming out in the age of the Internet: Identity” demarginalization” through virtual group participation. Journal of personality and social psychology, 75(3), 681-694.

Mickelson, K. Seeking social support: Parents in electronic support groups. In Culture of the Internet,S. Kiesler, Ed., Erlbaum (Mahwah, NJ, 1997), 157–178.

Naaman, M., Boase, J., and Lai, C.-H. Is it really about me?: message content in social awareness streams. InProc. CSCW, ACM (2010), 189–192.

Rains, S. A., Peterson, E. B., & Wright, K. B. (2015). Communicating social support in computer-mediated contexts: A meta-analytic review of content analyses examining support messages shared online among individuals coping with illness. Communication Monographs, 1-28. doi: 10.1080/03637751.2015.1019530

Rains, S. A. (2013). The Implications of Stigma and Anonymity for Self-Disclosure in Health Blogs. Health communication, (ahead-of-print), 1-9.
Robinson, J. D., & Turner, J. (2009). Impersonal, interpersonal, and hyperpersonal social support: Cancer and older adults. Health communication, 15(2), 227-234.

Spottswood, E. L., B Walther, J., Holmstrom, A. J., & B Ellison, N. (2013). Person‐centered emotional support and gender attributions in computer‐mediated communication. Human Communication Research, 39, 295-316.

Tan, Ah-Hwee. “Text mining: The state of the art and the challenges.” Proceedings of the PAKDD 1999 Workshop on Knowledge Disocovery from Advanced Databases. Vol. 8. 1999.

Witten, Ian H., and Eibe Frank. Data Mining: Practical machine learning tools and techniques. Morgan Kaufmann, 2005.

Wright, K. B., Rosenberg, J., Egbert, N., Ploeger, N. a, Bernard, D. R., & King, S. (2013). Communication competence, social support, and depression among college students: A model of Facebook and face-to-face support network influence. Journal of Health Communication, 18, 41-57. doi:10.1080/10810730.2012.688250.

Yoo, W., Namkoong, K., Choi, M., Shah, D. V., Tsang, S., Hong, Y., … & Gustafson, D. H. (2014). Giving and receiving emotional support online: Communication competence as a moderator of psychosocial benefits for women with breast cancer. Computers in human behavior, 30, 13-22.

Zhao, K., Greer, G. & Yen, J. (2014) Finding influential users of an online health community: a new metric based on sentiment influence. American Medical Informatics Association.

Leave a Reply

Your email address will not be published. Required fields are marked *