Imputing Missing Social Media Data Stream in Multisensor Studies of Human Behavior

2020 IEEE Second International Conference on Cognitive Machine Intelligence (CogMI)

Abowd

et al. 2020

Self Cite

In light of the ongoing COVID-19 pandemic, remote work styles have become the norm. However, these work settings introduce new intricacies in worker behaviors. The overlap between work and home can disrupt performance. The lack of social interaction can affect motivation. This elicits a need to implement novel methods to evaluate and enhance remote worker functioning. The potential to unobtrusively and automatically assess such workers can be fulfilled by social media and ubiquitous technologies. This paper situates recent research in the new context by extending our insights for increased remote interaction and online presence. We present implications for proactive assessment of remote workers by understanding day-level activities, coordination, role awareness, and organizational culture. Additionally, we discuss the ethics of privacy-preserving deployment, employer surveillance, and digital inequity. This paper aims to inspire pervasive technologies for the new future of work.

Section: Introductionmentioning

confidence: 99%

Social Media and Ubiquitous Technologies for Remote Worker Wellbeing and Productivity in a Post-Pandemic World

Swain

2020 IEEE Second International Conference on Cognitive Machine Intelligence (CogMI)

Abowd

et al. 2020

Self Cite

“…First, the past decade of computational social science research, which has repeatedly showcased how social media postings can provide rich insights about many real-world happenings, whether political, economic, social, or about health and well-being (Golder and Macy, 2011;Lazer et al, 2009Lazer et al, , 2020. Specifically, studies in psycholinguistics and crisis informatics have found promising evidence that the content shared on social media can help us to study mental health responses to crises, ranging from understanding how communities cope with protracted wars (Mark et al, 2012), community violence (De Choudhury et al, 2014;Saha andDe Choudhury, 2017), terrorism (Hoffman, 2018), homicides and mass shootings (Glasgow et al, 2014;Jones et al, 2017;Lin and Margolin, 2014). Second, with the growing adoption of social media among K-12 school communities including students, teachers, school administrators, and parents (Kimmons et al, 2018), social media constitutes a promising opportunity to study psychological states unobtrusively and passively.…”

Section: Methodsmentioning

confidence: 99%

“…We measured the probability (p-value) that the synthetic LC is greater than actual LC, which helps to quantify the statistical significance of our observations against chance or random observations. This method emulates permutation test frameworks applied in the prior work (Das et al, 2020;Saha, 2019a), and tests for the null hypothesis that outcome change around a randomly generated drill date is comparable to outcome change around actual drill date. If this p-value is found to be zero or significantly low (e.g., p < 0.05), then we can deduce that the treatment LC is indeed attributed to the effects of the activeshooter drill.…”

Section: Methodsmentioning

confidence: 99%

Impacts of school shooter drills on the psychological well-being of American K-12 school communities: a social media study

ElSherief

Gupta

et al. 2021

Humanit Soc Sci Commun

Self Cite

The toll from gun violence in American K-12 schools has escalated over the past 20 years. School administrators face pressure to prepare for possible active shootings, and often do so through drills, which can range from general lockdowns to simulations, involving masked “shooters” and simulated gunfire, and many variations in between. However, the broad and lasting impact of these drills on the well-being of school communities is poorly understood. To that end, this article applies machine learning and interrupted time series analysis to 54 million social media posts, both pre- and post-drills in 114 schools spanning 33 states. Drill dates and locations were identified via a survey, then posts were captured by geo-location, school social media following, and/or school social media group membership. Results indicate that anxiety, stress, and depression increased by 39–42% following the drills, but this was accompanied by increases in civic engagement (10–106%). This research, paired with the lack of strong evidence that drills save lives, suggests that proactive school safety strategies may be both more effective, and less detrimental to mental health, than drills.

“…In the same body of research, social media and online engagement platforms have facilitated an effective means to study employee behavior and satisfaction -a body of research that is extensive in CSCW and HCI area [5,33,105,130,132,135]. A variety of analytical and computational approaches on language and network dynamics have been applied to glean correlates of employee job satisfaction and wellbeing, such as engagement [66,105,130], employee affect [33,121], social pulse [131], reputation [74], organizational relationships [17,52,104], and workplace behavior [94]. Lee and Kang used Glassdoor data to study the influence job satisfaction factors, and their influence on employee retention and turnover [89].…”

Section: Social Media Technologies and Workplace Behaviormentioning

confidence: 99%

“…Participants authorized access to their social media data through an Open Authentication (OAuth) based data collection infrastructure that was developed in-house [119,121]. In particular, we asked permission from participants to provide their Facebook and LinkedIn data, unless they opted out, or did not have either of these accounts.…”

Section: Social Media (Linkedin) Datamentioning

confidence: 99%

LibRA

Proc. ACM Hum.-Comput. Interact.

Reddy

Mattingly

et al. 2019

Self Cite

Job roles serve as a boundary between an employee and an organization, and are often considered building blocks in understanding the behavior and functioning of organizational systems. However, a lack of clarity about one's role, that is, one's work responsibilities and degree of authority, can lead to absenteeism, turnover, dissatisfaction, stress, and lower workplace performance. This paper proposes a methodology to quantitatively estimate role ambiguity via unobtrusively gathered data from LinkedIn, shared voluntarily by a cohort of information workers spanning multiple organizations. After successfully validating this LinkedIn based measure of Role Ambiguity, or LibRA against a state-of-the-art gold standard, drawing upon theories in organizational psychology, we examine the efficacy and convergent validity of LibRA in explaining established relationships of role ambiguity with wellbeing and performance measures of individuals. We find that greater LibRA is associated with depleted wellbeing, such as increased heart rate, increased arousal, decreased sleep, and higher stress. In addition, greater LibRA is associated with lower job performance such as decreased organizational citizenship behavior and decreased individual task performance. We discuss how LibRA can help fill gaps in state-of-the-art assessments of role ambiguity, and the potential of this measure in building novel technology-mediated strategies to combat role ambiguity in organizations.CCS Concepts: • Human-centered computing → Empirical studies in collaborative and social computing; Social media; • Applied computing → Psychology.