The Language of Social Support in Social Media and Its Effect on Suicidal Ideation Risk

De Choudhury M, Kıcıman E. The Language of Social Support in Social Media and its Effect on Suicidal Ideation Risk. Proc Int AAAI Conf Weblogs Soc Media. 2017;2017:32-41.

---

Summary

• Introduction
  Online social support is known to play a significant role in mental well-being. However, current research is limited in its ability to quantify this link.
  The study proposes a method to measure how the language of comments in Reddit mental health communities influences the risk to suicidal ideation in the future.

• Methodology

The study examined publicly shared longitudinal post and comment data on mental health communities in the social media Reddit. Then, it identified individuals who proceeded to post on a Reddit suicide support forum at a later point in the time.

Utilizing comments as a proxy for social support, the study developed a human-machine hybrid statistical methodology that models and quantifies the effects of the language of these comments in individuals who do and do not post on the suicide support forum.

Applying stratified propensity score matching, authors first identified linguistic features of comments that show significant effects. Then, they obtained human assessments on the presence of suicidal ideation risk markers, filtered the features, and included these assessments in computing the local average treatment effect, so as to assess the effects of specific linguistic features of comments in future risk to suicidal ideation.

• Result
  The study finds that the positive impact of esteem and network support play a more prominent role in reducing forthcoming risk.

• Conclusion
  The research contributed to addressing methodological gaps in assessing the role of online social support in future risk to suicidal ideation. The result of the study can be used to improve support provisions in online communities.

---

Comments

Things to study more

• What is the stratified propensity score matching approach?
  Propensity score matching (PSM) is a quasi-experimental method in which the researcher uses statistical techniques to construct an artificial control group by matching each treated unit with a non-treated unit of similar characteristics. Using these matches, the researcher can estimate the impact of an intervention.
  "Propensity" in this study can be interpreted as "how users are likely to be affected by specific tokens".

  reference: https://dimewiki.worldbank.org/Propensity_Score_Matching#:~:text=Propensity%20score%20matching%20(PSM)%20is,the%20impact%20of%20an%20intervention.

  Rosenbaum, P. R., and Rubin, D. B. 1983. The central role of the propensity score in observational studies for causal effects.Biometrika70(1):41–55

• Interesting preceding research

  Oh, H. J.; Lauckner, C.; Boehmer, J.; Fewins-Bliss, R.; and Li, K.2013. Facebooking for health: An examination into the solicitation and effects of health-related social support on social networking sites.Computers in Human Behavior29(5):2072–2080.
  

Possible Future Research Ideas

• Utilizing features that distinguish the people who are likely to benefit from online social support from those who don't to personalize support provisions in online communities
  The study reveals that not all individuals posting in the Reddit mental health communities are equally likely to be influenced by support received through comments. It is said that those who benefit from online social support tend to show greater social and futuristic orientation, interpersonal awareness, and lower cognitive impairment. These observations align with observations in the prior study.

  Kaplan, B. H.; Cassel, J. C.; and Gore, S. 1977. Social support and health.Medical care15(5):47–58.

• How to improve intervention messages to suicidal risk users
  The study suggests that receiving acknowledgment of one's feelings, or advice around mental health issues does not translate to reduced future expressions of suicidal ideation. The coded comments in MH -> SW (people who possibly developed suicidal ideation) involve acknowledgments (40%) and informational support (23%).
  How can we utilize this to improve intervention messages for suicidal risk?

• Finding out why certain types of social support are relatively more beneficial
  The research could not show why certain forms of social support tend to be associated with a reduced likelihood of suicidal ideation in the future, or why certain others tend to show converse effects.

Questions

• How using the causal inference method can prevent confounding effects?
  Understanding that participation in online support communities is inherently self-selected in nature, if traditional statistical approaches like regression models are used to measure the effectiveness of support it can lead to confounding effects. To close this gap, the authors said they used methods from causal inference literature. It is said that causal analysis is commonly used for observational studies.

  • Rubin, D. B. 2011. Causal inference using potential outcomes.Journal of the American Statistical Association.
  • Imbens, G. W., and Rubin, D. B. 2015.Causal Inference in Statistics, Social, and Biomedical Sciences. Cambridge University Press.

---

Things to note

• The idea of utilizing Reddit mental health communities and suicide support data is so creative.

• Z scores stand for how many standard deviations a data point is from the mean in a distribution.

• Human-computer interaction