Facebook language predicts depression in medical records

Eichstaedt, Johannes C.; Smith, Robert J.; Merchant, Raina M.; Ungar, Lyle H.; Crutchley, Patrick; Preoţiuc-Pietro, Daniel; Asch, David A.; Schwartz, H. Andrew

doi:10.1073/pnas.1802331115

Cited by 488 publications

(405 citation statements)

References 43 publications

(46 reference statements)

Supporting

Mentioning

353

Contrasting

Unclassified

Order By: Relevance

“…For example, text analysis can map the lifecycle of social movements (Ahmed, Jaidka, & Cho, 2017) and indicate mental health diagnoses (Eichstaedt et al, 2018). The current paper adopts this approach, paired with social science theory, empirical evidence, and exploratory techniques, to predict the amount of time a pet will spend available online (Study 1) and the verbal correlates of pet adoption (e.g., text features that define adopted vs. unadopted profiles; Study 2).…”

Section: Language Patterns and Psychological Dynamicsmentioning

confidence: 99%

Putting your best pet forward: Language patterns of persuasion in online pet advertisements

Markowitz

2019

J Applied Social Pyschol

View full text Add to dashboard Cite

This paper evaluates persuasion dynamics of animal adoption using text data from a large archive of online pet advertisements. In Study 1, 184,805 adoption profiles from Petfinder indicated how long a pet will remain online and unadopted. Consistent with evidence from related persuasion settings such as peer‐to‐peer lending, pets spent less time online if profile writers had an analytic thinking style and advertisements contained few peripheral processing cues such as social words. Study 2 (N = 676,004 adoption profiles) replicated Study 1 patterns and found that adopted pet profiles contained more markers of analytic thinking and fewer social words than unadopted pet profiles. In an experiment (Study 3, N = 987), participants read an adoption advertisement typical of adopted or unadopted pets. Participants self‐reported that they would be more likely to adopt a pet and visit its shelter after reading a more analytic and less social adoption profile (indicators of adopted pets) than a less analytic and more social profile (indicators of unadopted pets). Finally, Study 4 (N = 3,245 Tweets) demonstrated that more analytic and less social word patterns relate to increased engagement online, such as likes and retweets. These data suggest pet adoption that begins online is a social and psychological process, enhanced by messages with markers of complex thinking and few humanizing references. Advances to persuasion theory are discussed, underscored by the implications for pet adoption and how language patterns in online advertisements can reflect influence at scale.

show abstract

Section: Language Patterns and Psychological Dynamicsmentioning

confidence: 99%

Putting your best pet forward: Language patterns of persuasion in online pet advertisements

Markowitz

2019

J Applied Social Pyschol

View full text Add to dashboard Cite

show abstract

“…There is growing interest in building a “social mediome,” using data derived from social media platforms that reveal individual‐ and population‐level health information to gain greater insight into patient health habits and facilitate treatment. Public health researchers now utilize data from websites such as Google, Twitter, and Facebook to more precisely examine health trends, diagnose illnesses, and even predict behavior such as depression and suicide …”

Section: Articlementioning

confidence: 99%

Social Media, E‐Health, and Medical Ethics

Terrasse¹,

Gorin²,

Sisti³

2019

Hastings Center Report

View full text Add to dashboard Cite

show abstract

“…Topic models have been shown to be able to differentiate social affections in web-based texts (Bao et al, 2009), to extract emotional components, such as anger and anxiety, from a psychotherapy corpus (Imel et al, 2015), and to estimate a writer's emotional states from Facebook language to predict future diagnoses of depression better than clinical screening tests (Eichstaedt et al, 2018). Thus, topic model is a promising approach to extract and compare the meaning of emotional words.…”

Section: Awe In the Japanese Contextmentioning

confidence: 99%

Meaning of awe in Japanese (con)text: Beyond fear and respect

Nakayama¹,

Uchida²

2019

Preprint

View full text Add to dashboard Cite

Awe is theorized as an emotion appraised by perceived vastness and need for accommodation. This theoretical framework was developed with a review of spatially and temporally distributed literature mostly in the American and European cultural context, and is assumed to be culturally universal. However, awe as described by Japanese literature, was not explicitly included in the original theorization. We tested whether this framework generalized to the Japanese context by analyzing how Japanese awe-related words (e.g., “畏敬/ikei”) are used in Japanese text. A topic model was used to extract topics in contexts as an index of meaning. Results show that (1) the meaning of awe was statistically dissociable from similar but distinct meanings of fear and respect, and (2) the dissociating topics included transcendent entities such as god, spirits/ghosts, and powerful beings. Japanese meaning of awe includes vastness (i.e., transcendence) that goes beyond typical respect (i.e., power distance) requiring an accommodation of one’s mental framework.

show abstract

Facebook language predicts depression in medical records

Cited by 488 publications

References 43 publications

Putting your best pet forward: Language patterns of persuasion in online pet advertisements

Putting your best pet forward: Language patterns of persuasion in online pet advertisements

Social Media, E‐Health, and Medical Ethics

Meaning of awe in Japanese (con)text: Beyond fear and respect

Contact Info

Product

Resources

About