Using Smartphone Sensor Paradata and Personalized Machine Learning Models to Infer Participants’ Well-being: Ecological Momentary Assessment

Im April haben Alexander Hart und Dorota Reis zusammen mit Elisabeth Prestele (Universität Koblenz-Landau) und Nicholas Jacobson (Dartmouth College) einen Artikel veröffentlicht, in dem die Smartphone Nutzung während Onlinebefragungen und das Wohlbefinden der Teilnehmenden im Mittelpunkt stehen. Wenn Studien erfordern, dass Teilnehmende täglich mehrfach an Befragungen teilnehmen müssen, kann dies unter Umständen zur Belastung werden. Die Studie untersuchte, ob sich Antworten der Teilnehmende stattdessen auch passiv, durch Nutzung der Smartphonesensoren während der Befragung, verlässlich vorhersagen ließen, ohne die Antworten der Teilnehmenden zu kennen.

Journal: Journal of Medical Internet Research

DOI: https://doi.org/10.2196/34015 [Volltext kostenfrei verfügbar]

Background: Sensors embedded in smartphones allow for the passive momentary quantification of people’s states in the context of their daily lives in real time. Such data could be useful for alleviating the burden of ecological momentary assessments and increasing utility in clinical assessments. Despite existing research on using passive sensor data to assess participants’ moment-to-moment states and activity levels, only limited research has investigated temporally linking sensor assessment and self-reported assessment to further integrate the 2 methodologies.
Objective: We investigated whether sparse movement-related sensor data can be used to train machine learning models that are able to infer states of individuals’ work-related rumination, fatigue, mood, arousal, life engagement, and sleep quality. Sensor data were only collected while the participants filled out the questionnaires on their smartphones.
Methods: We trained personalized machine learning models on data from employees (N=158) who participated in a 3-week ecological momentary assessment study. Results: The results suggested that passive smartphone sensor data paired with personalized machine learning models can be used to infer individuals’ self-reported states at later measurement occasions. The mean R2 was approximately 0.31 (SD 0.29), and more than half of the participants (119/158, 75.3%) had an R2 of ≥0.18. Accuracy was only slightly attenuated compared with earlier studies and ranged from 38.41% to 51.38%.
Conclusions: Personalized machine learning models and temporally linked passive sensing data have the capability to infer a sizable proportion of variance in individuals’ daily self-reported states. Further research is needed to investigate factors that affect the accuracy and reliability of the inference.

Research Without Borders: How to Identify and Overcome Potential Pitfalls in International Large-Team Online Research Projects

Im März haben Myriam Baum und Alexander Hart in Zusammenarbeit mit Wissenschaftlerinnen und Wissenschaftlern des Projektes Open Science Cross-Cultural (OSCC) einen Artikel veröffentlicht, in dem Hindernisse grenzübergreifender Forschung aufgezeigt werden. Der Artikel greift dabei vor allem die Erfahrung auf, die die Autorinnen und Autoren während des OSCC-Projektes gemacht haben und bespricht Lösungsansätze, wie Abläufe in großen Gruppen von global verteilten Forschenden interaktiv und reibungslos gestaltet werden können.

Journal: SAGE Research Methods

DOI: https://dx.doi.org/10.4135/9781529602074

Abstract: This case study is based on experiences with an international meta-scientific project across different fields of research, assessing the use of Open Science (OS) practices in these fields as well as individual researchers’ reasons for not engaging in OS practices. The project was conducted by a cross-cultural and diverse group of researchers. In the case, we share insights into potential pitfalls when conducting an international study, as well as possible solutions to overcome them. Specifically, we highlight various issues by focusing on two key phases, namely, designing an online survey with a focus on cross-cultural data collection and recruiting a cross-cultural online sample of researchers.

A community-sourced glossary of open scholarship terms

Im Februar haben Myriam Baum und Alexander Hart zusammen mit Autorinnen und Autoren des Framework for Open and Reproducible Research Training (FORRT) unter Leitung von Sam Parsons (Radboud University), Flávio Azevedo (Universität Köln, University of Cambridge), und Mahmoud Elsherif (University of Birmingham) einen Artikel veröffentlicht, der die Entwicklung eines Glossars für Begriffe aus dem Spektrum der transparenten Wissenschaft (Open Science) beschreibt. Der Glossar ist eine Gemeinschaftarbeit von über einhundert Wissenschaftlerinnen und Wissenschaftlern verschiedener Disziplinenen und Schwerpunkte, die die enthaltenen Begriffe recherchierten, definierten und miteinander diskutierten.

Journal: Nature Human Behavior

DOI: https://doi.org/10.1038/s41562-021-01269-4

Abstract: Open scholarship has transformed research, and introduced a host of new terms in the lexicon of researchers. The‘Framework for Open and Reproducible Research Teaching’ (FORRT) community presents a crowdsourced glossary of open scholarship terms to facilitate education and effective communication between experts and newcomers.

No party no joy?—Changes in university students' extraversion, neuroticism, and subjective well-being during two COVID-19 lockdowns

Im Januar haben Kai Krautter (Universtät des Saarlandes, Harvard Business School), Dorota Reis und Alexander Hart, in Zusammenarbeit mit Malte Friese (Universität des Saarlandes) einen Artikel veröffentlicht, in dem die Veränderungen, die Studierende während der COVID-19 durchlebt haben, untersucht wurden. Neben der Untersuchung des generellen Wohlbefindens legte die Studie dabei besonderen Augenmerk auf Veränderungen der Persönlichkeitsstruktur, die durch dieses außergewöhnliche Lebensereignis hervorgerufen wurden, und deren Stabilität. An der Studie nahmen zwei unterschiedliche Kohorten von Studierenden der Universität des Saarlandes teil.

Journal: Applied Psychology: Health and Well-Being

DOI: https://doi.org/10.1111/aphw.12336[Volltext kostenfrei verfügbar]

Abstract: The COVID-19 lockdowns represent a major life event with an immense impact on university students' lives. Findings prior to the pandemic suggest that changes in personality and subjective well-being (SWB) can occur after critical life events or psychological interventions. The present study examined how university students' extraversion, neuroticism, and SWB changed during two COVID-19 lockdowns in Germany. To this end, we conducted a partly preregistered, two-cohort study with four measurement points each from October 2019 to May 2021 (NStudy 1 = 81–148, NStudy 2 = 82–97). We used both multilevel contrast analyses and multi-group random-intercept cross-lagged panel models to examine within-person changes over time. Levels of life satisfaction, extraversion, and, unexpectedly, neuroticism were lower during both lockdowns. Students' affect improved during the first but deteriorated during the second lockdown, suggesting that similar experiences with the deceleration of daily life were associated with different affective outcomes during the two lockdown periods. Following the introduction or termination of a lockdown, changes in extraversion (neuroticism) were consistently positively (negatively) associated with changes in SWB. Our results stress the importance of disentangling between- and within-person processes and using pre-COVID baseline levels to examine changes in personality and SWB.

Promoting Open Science: A Holistic Approach to Changing Behaviour

Im Dezember hat Myriam Baum mit Samuel Robson (University of Queensland) und weiteren Autorinnen und Autoren einen Artikel veröffentlicht, der sich mit der Förderung transparenter wissenschaftlicher Praktiken befasst. In dem Artikel werden Möglichkeiten vorgestellt, wie nicht nur Wissenschaftlerinnen und Wissenschaftler, sondern u. a. auch Studierende, Universitäten und Verlage dazu motiviert werden können ihr wissenschaftliches Arbeiten transparenter zu gestalten.

Journal: Collabra Psychology

DOI: https://doi.org/10.1525/collabra.30137 [Volltext kostenfrei verfügbar]

Abstract: In this article, we provide a toolbox of recommendations and resources for those aspiring to promote the uptake of open scientific practices. Open Science encompasses a range of behaviours that aim to improve the transparency of scientific research. This paper is divided into seven sections, each devoted to different groups or institutions in the research ecosystem: colleagues, students, departments and faculties, universities, academic libraries, journals, and funders. We describe the behavioural influences and incentives for each of these stakeholders as well as changes they can make to foster Open Science. Our primary goal, however, is to suggest actions that researchers can take to promote these behaviours, inspired by simple principles of behaviour change: make it easy, social, and attractive. In isolation, a small shift in one person’s behaviour may appear to make little difference, but when combined, many shifts can radically alter shared norms and culture. We offer this toolbox to assist individuals and institutions in cultivating a more open research culture.

Promoting recovery in daily life: study protocol for a randomized controlled trial

Im April haben Dorota Reis und Alexander Hart, zusammen mit Malte Friese (Universität des Saarlandes) und Dirk Lehr (Leuphana Universität), ein Studienprotokoll veröffentlich, das ein in Vorbereitung befindliches Forschungsprojekt unserer Forschungsgruppe beschreibt. In diesem Projekt wird die Wirksamkeit von internetbasierten Trainingsmaßnahmen untersucht, die Berufstätigen dabei helfen sollen, nach der Arbeit besser abschalten zu können. Neben einer grundsätzlichen Bestimmung der Wirksamkeit, untersucht das Projekt ebenfalls, wie genau die Maßnahmen von Tag zu Tag wirken, und die Teilnehmenden dabei unterstützen, nach der Arbeit mehr "Freiraum im Kopf" zu schaffen.

Journal: BMC Psychology

DOI: https://doi.org/10.1186/s40359-021-00591-w [Volltext kostenfrei verfügbar]

Background: Work-related stress shows steadily increasing prevalence rates and has tangible consequences for
individual workers, their organizations, and society as a whole. One mechanism that may help offset the negative
outcomes of work-related stress on employees’ well-being is recovery. Recovery refers to the experience of unwinding
from one’s job when not at work. However, employees who experience high levels of work-related stress and are thus
particularly in need of recovery tend to struggle to switch-off. Due to the detrimental effects of this prolonged and
sustained mental representation of job stressors, interventions promoting recovery may contribute to improvements
in employees’ mental health.
Methods: In this randomized, waitlist controlled trial, we will investigate the effectiveness of two 6-week online train-
ing programs (cognitive behavioral and mindfulness-based). The sample will include employees working at least part-
time during regular work hours. Besides the pre-post-follow-up assessments, the trial will include measurement bursts
with the goal of examining the underlying mechanisms. We expect that both interventions will reduce work-related
perseverative thinking (PT) compared with the waitlist control groups (primary outcome). Also, we expect that both
interventions will result in similar improvements, but the underlying mechanisms will differ (process outcomes). In the
cognitive-behavioral intervention group, we expect that the main mechanism responsible for lower PT levels will be
an increase in recovery experiences across time. In the mindfulness-based group, we expect that the main mecha-
nism responsible for lower PT levels will be an increase in facets of mindfulness across time.
Discussion: In the present study, we will investigate mechanisms underlying assumed changes in work-related PT
in great detail. Besides evaluating the overall effectiveness of the two interventions in terms of pre-post-follow-up
changes, we will look at the underlying processes at different levels—that is, within days, within weeks, across weeks,
and between individuals. Accordingly, our study will offer a fine-grained approach to investigating potential determi-
nants, mediators, and moderators of the processes that may, in the end, be responsible for work-related strain. From a
public health perspective, if effective, the online training programs may offer valuable, low-threshold, and low-inten-
sity interventions for a broad range of occupations.