Twitter as a Mental Health Support System for Students and Professionals in the Medical Field

Who posted #MeToo, why, and what happened: A mixed methods examination

Objectives

The #MeToo social media campaign raised awareness about sexual harassment. The purpose of the current study was to address three unexplored research questions. First, what factors influenced whether a person posted #MeToo? Second, how did posting (or not) influence participants' wellbeing? Finally, what motivated participants' posting (or not) #MeToo?

Method

This mixed-methods study explores how #MeToo was experienced by full-time employees (N = 395) who could have posted #MeToo (i.e., experienced a sexual harassment event), whether or not they did so. Participants completed surveys in July of 2018 assessing social media use, sexual harassment history, relational variables such as relative power and social support, and job and life satisfaction. Participants also responded to open-ended survey questions about the context of and decisions about #MeToo posting.

Results

Quantitative results indicated that sexual harassment history was the most powerful predictor of #MeToo posting, while power and interpersonal contact also contributed. Qualitative analyses (N = 74) using a grounded theory approach indicated themes associated with decisions to disclose, including feeling a responsibility to post, need for support, and affective benefits. Decisions not to disclose were event-related negative affect, posting-related negative affect, timing of the event, fit with the #MeToo movement, privacy concerns, and fear of consequences.

Conclusion

This study contributes to the literature on sexual harassment disclosure by focusing on informal means of disclosure and drawing on comparisons to formal reporting and implications for workplaces. Online sexual harassment disclosure, in many ways, reflects the impediments to formal reporting procedures. Given the increased use of social media for purposes of disclosure, these findings suggests that organizations should recognize the legitimacy of sexual harassment reports made online and consider the possible failings of their formal reporting systems as reasons for online disclosure.

Nowcasting unemployment rate during the COVID-19 pandemic using Twitter data: The case of South Africa

The global economy has been hard hit by the COVID-19 pandemic. Many countries are experiencing a severe and destructive recession. A significant number of firms and businesses have gone bankrupt or been scaled down, and many individuals have lost their jobs. The main goal of this study is to support policy- and decision-makers with additional and real-time information about the labor market flow using Twitter data. We leverage the data to trace and nowcast the unemployment rate of South Africa during the COVID-19 pandemic. First, we create a dataset of unemployment-related tweets using certain keywords. Principal Component Regression (PCR) is then applied to nowcast the unemployment rate using the gathered tweets and their sentiment scores. Numerical results indicate that the volume of the tweets has a positive correlation, and the sentiments of the tweets have a negative correlation with the unemployment rate during and before the COVID-19 pandemic. Moreover, the now-casted unemployment rate using PCR has an outstanding evaluation result with a low Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), Symmetric MAPE (SMAPE) of 0.921, 0.018, 0.018, respectively and a high R²-score of 0.929.

Social Media and the Patient - on Education and Empowerment

Social media has unprecedentedly impacted the world, and this includes patients and physicians alike. This article provides a glimpse of the pros and cons of social media to both parties, and how, despite its pitfalls, rheumatologists can put its use in daily practice to help bridge the gap between, and among, rheumatologists and patients to ultimately improve patient outcomes.

5-Year Analysis of Twitter Usage Among Physicians: 2016-2020 (Preprint)

Background

Physicians are increasingly using Twitter as a channel for communicating with colleagues and the public. Identifying physicians on Twitter is difficult due to the varied and imprecise ways that people self-identify themselves on the social media platform. This is the first study to describe a reliable, repeatable methodology for identifying physicians on Twitter. By using this approach, we characterized the longitudinal activity of US physicians on Twitter.

Objective

We aimed to develop a reliable and repeatable methodology for identifying US physicians on Twitter and to characterize their activity on Twitter over 5 years by activity, tweeted topic, and account type.

Methods

In this study, 5 years of Twitter data (2016-2020) were mined for physician accounts. US physicians on Twitter were identified by using a custom-built algorithm to screen for physician identifiers in the Twitter handles, user profiles, and tweeted content. The number of tweets by physician accounts from the 5-year period were counted and analyzed. The top 100 hashtags were identified, categorized into topics, and analyzed.

Results

Approximately 1 trillion tweets were mined to identify 6,399,146 (<0.001%) tweets originating from 39,084 US physician accounts. Over the 5-year period, the number of US physicians tweeting more than doubled (ie, increased by 112%). Across all 5 years, the most popular themes were general health, medical education, and mental health, and in specific years, the number of tweets related to elections (2016 and 2020), Black Lives Matter (2020), and COVID-19 (2020) increased.

Conclusions

Twitter has become an increasingly popular social media platform for US physicians over the past 5 years, and their use of Twitter has evolved to cover a broad range of topics, including science, politics, social activism, and COVID-19. We have developed an accurate, repeatable methodology for identifying US physicians on Twitter and have characterized their activity.

Traditional Machine Learning Models and Bidirectional Encoder Representations From Transformer (BERT)-Based Automatic Classification of Tweets About Eating Disorders: Algorithm Development and Validation Study

BACKGROUND

Eating disorders affect an increasing number of people. Social networks provide information that can help.

OBJECTIVE

We aimed to find machine learning models capable of efficiently categorizing tweets about eating disorders domain.

METHODS

We collected tweets related to eating disorders, for 3 consecutive months. After preprocessing, a subset of 2000 tweets was labeled: (1) messages written by people suffering from eating disorders or not, (2) messages promoting suffering from eating disorders or not, (3) informative messages or not, and (4) scientific or nonscientific messages. Traditional machine learning and deep learning models were used to classify tweets. We evaluated accuracy, F1 score, and computational time for each model.

RESULTS

A total of 1,058,957 tweets related to eating disorders were collected. were obtained in the 4 categorizations, with The bidirectional encoder representations from transformer-based models had the best score among the machine learning and deep learning techniques applied to the 4 categorization tasks (F1 scores 71.1%-86.4%).

CONCLUSIONS

Bidirectional encoder representations from transformer-based models have better performance, although their computational cost is significantly higher than those of traditional techniques, in classifying eating disorder-related tweets.