Meta-analytic procedures for social research

Depression, the most prevalent mental illness, is underdiagnosed and undertreated, highlighting the need to extend the scope of current screening methods. Here, we use language from Facebook posts of consenting individuals to predict depression recorded in electronic medical records. We accessed the history of Facebook statuses posted by 683 patients visiting a large urban academic emergency department, 114 of whom had a diagnosis of depression in their medical records. Using only the language preceding their first documentation of a diagnosis of depression, we could identify depressed patients with fair accuracy [area under the curve (AUC) = 0.69], approximately matching the accuracy of screening surveys benchmarked against medical records. Restricting Facebook data to only the 6 months immediately preceding the first documented diagnosis of depression yielded a higher prediction accuracy (AUC = 0.72) for those users who had sufficient Facebook data. Significant prediction of future depression status was possible as far as 3 months before its first documentation. We found that language predictors of depression include emotional (sadness), interpersonal (loneliness, hostility), and cognitive (preoccupation with the self, rumination) processes. Unobtrusive depression assessment through social media of consenting individuals may become feasible as a scalable complement to existing screening and monitoring procedures.

https://www.pnas.org/content/pnas/115/44/11203.full.pdf

Facebook language predicts depression in medical records.

Depressive symptomatology is manifested in greater first-person singular pronoun use (i.e., I-talk), but when and for whom this effect is most apparent, and the extent to which it is specific to depression or part of a broader association between negative emotionality and I-talk, remains unclear. Using pooled data from N = 4,754 participants from 6 labs across 2 countries, we examined, in a preregistered analysis, how the depression-I-talk effect varied by (a) first-person singular pronoun type (i.e., subjective, objective, and possessive), (b) the communication context in which language was generated (i.e., personal, momentary thought, identity-related, and impersonal), and (c) gender. Overall, there was a small but reliable positive correlation between depression and I-talk (r = .10, 95% CI [.07, .13]). The effect was present for all first-person singular pronouns except the possessive type, in all communication contexts except the impersonal one, and for both females and males with little evidence of gender differences. Importantly, a similar pattern of results emerged for negative emotionality. Further, the depression-I-talk effect was substantially reduced when controlled for negative emotionality but this was not the case when the negative emotionality-I-talk effect was controlled for depression. These results suggest that the robust empirical link between depression and I-talk largely reflects a broader association between negative emotionality and I-talk. Self-referential language using first-person singular pronouns may therefore be better construed as a linguistic marker of general distress proneness or negative emotionality rather than as a specific marker of depression. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Depression, negative emotionality, and self-referential language: A multi-lab, multi-measure, and multi-language-task research synthesis.

The study investigates the relationship between air pollution and the status of human mental health using the data of the China Migrant Dynamic Survey (CMDS), 2014. Various econometric methods were used. Particularly, the ordinary least square (OLS) was used to estimate the parametric impact of mental health. Moreover, a robustness test was used to estimate negative impact of air pollution on mental health. Results found that air pollution significantly increased psychological disorder. The concentration of PM2.5 in the air leads to higher prominence of the four negative emotions such as nervous, depression, powerless, and restless or fidgety. After introducing the instrumental variable to solve the endogenous problem, the impact of air pollution on the level of the mental illnesses was still significant and negative. Results are also evident that air pollution has heterogeneity to the negative status of residents’ mental health. Specifically, it has a significant and negative impact on women, rural and urban population, and middle- and low-income groups. The study results suggest that the government must have to revise the environmental regulation policy reduce the negative impact on human health.

Air pollution risks human mental health: an implication of two-stages least squares estimation of interaction effects

The words that people use have been found to reflect stable psychological traits, but less is known about the extent to which everyday fluctuations in spoken language reflect transient psychological states. We explored within-person associations between spoken words and self-reported state emotion among 185 participants who wore the Electronically Activated Recorder (EAR; an unobtrusive audio recording device) and completed experience sampling reports of their positive and negative emotions 4 times per day for 7 days (1,579 observations). We examined language using the Linguistic Inquiry and Word Count program (LIWC; theoretically created dictionaries) and open-vocabulary themes (clusters of data-driven semantically-related words). Although some studies give the impression that LIWC's positive and negative emotion dictionaries can be used as indicators of emotion experience, we found that when computed on spoken language, LIWC emotion scores were not significantly associated with self-reports of state emotion experience. Exploration of other categories of language variables suggests a number of hypotheses about substantive everyday correlates of momentary positive and negative emotion that can be tested in future studies. These findings (a) suggest that LIWC positive and negative emotion dictionaries may not capture self-reported subjective emotion experience when applied to everyday speech, (b) emphasize the importance of establishing the validity of language-based measures within one's target domain, (c) demonstrate the potential for developing new hypotheses about personality processes from the open-ended words that are used in everyday speech, and (d) extend perspectives on intraindividual variability to the domain of spoken language. (PsycINFO Database Record (c) 2020 APA, all rights reserved).

https://minerva-access.unimelb.edu.au/bitstream/handle/11343/253830/2019-18075-001.pdf

The language of well-being: Tracking fluctuations in emotion experience through everyday speech.

A meta-analysis of correlations between depression and first person singular pronoun use.

Depressive symptomatology is manifested in greater first-person singular pronoun use (i.e., I-talk), but when and for whom this effect is most apparent, and the extent to which it is specific to depression or part of a broader association between negative emotionality and I-talk, remains unclear. Using pooled data from N = 4,754 participants from 6 labs across 2 countries, we examined, in a preregistered analysis, how the depression–I-talk effect varied by (a) first-person singular pronoun type (i.e., subjective, objective, and possessive), (b) the communication context in which language was generated (i.e., personal, momentary thought, identity-related, and impersonal), and (c) gender. Overall, there was a small but reliable positive correlation between depression and I-talk (r = .10, 95% CI [.07, .13]). The effect was present for all first-person singular pronouns except the possessive type, in all communication contexts except the impersonal one, and for both females and males with little evidence of gender differences. Importantly, a similar pattern of results emerged for negative emotionality. Further, the depression–I-talk effect was substantially reduced when controlled for negative emotionality but this was not the case when the negative emotionality–I-talk effect was controlled for depression. These results suggest that the robust empirical link between depression and I-talk largely reflects a broader association between negative emotionality and I-talk. Self-referential language using first-person singular pronouns may therefore be better construed as a linguistic marker of general distress proneness or negative emotionality rather than as a specific marker of depression.

Depression, Negative Emotionality, and Self-Referential Language: A Multi-Lab, Multi-Measure, and Multi-Language-Task Research Synthesis_Tackman et al_Preprint

To'Meisha S. Edwards

Papers

A meta-analysis of correlations between depression and first person singular pronoun use.

Depression, negative emotionality, and self-referential language: A multi-lab, multi-measure, and multi-language-task research synthesis.

Depression, Negative Emotionality, and Self-Referential Language: A Multi-Lab, Multi-Measure, and Multi-Language-Task Research Synthesis_Tackman et al_Preprint