physiological and emotional assessment of college students using wearable and mobile devices during the 2020 COVID-19 lockdown: an intensive, longitudinal dataset

This dataset was collected from university students before, during, and after the COVID-19 lockdown in Southern California. Data collection happened continuously for the average of 7.8 months (SD=3.8, MIN=1.0, MAX=13.4) from a population of 21 students of which 12 have also completed an exit survey, and 7 started before the California COVID-19 lockdown order. This multimodal dataset included different means of data collection such as Samsung Galaxy Watch, Oura Ring, a Life-logger app named Personicle, a questionnaire mobile app named Personicle Questions, and periodical and personalised surveys. The dataset contains raw data from Photoplethysmogram (PPG), Inertial measurement unit (IMU), and pressure sensors in addition to processed data on heart rate, heart rate variability, sleep (bedtime, sleep stages, quality), and physical activity (step, active calories, type of activity). Ecological momentary assessments were collected from participants on daily and weekly bases containing their Positive and Negative Affect Schedule (PANAS) questionnaire and their emotional responses to COVID-19 and their health. Subjective data was also collected through monthly surveys containing standard mood and mental health surveys such as Beck Depression Inventory II (BDI-II), Brief Symptom Inventory (BSI), GAD-7, Inclusion of Other in the Self Scale (IOS-Partner), Acceptability of Intervention Measure (AIM), Intervention Appropriateness Measure (IAM), Feasibility of Intervention Measure (FIM), Experiences in Close Relationships Scale Short Form (ECR-S), UCLA Three-Item Loneliness Scale (ULS), Multidimensional Scale of Perceived Social Support (MSPSS), Investment Model Scale (IMS), Conflict Management Scale (CMS), etc in addition to their response to important events and COVID-19. This dataset can be used to study emotions, mood, physical activity, and lifestyle of young adults through longitudinal subjective and objective measures. This dataset also contains valuable data regarding adjustment of lifestyle and emotions during the events of 2020 and 2021 including COVID-19 discovery and lockdown, Black Life Matter movement, 2020 US presidential elections, etc.

Methods
Experiment Design and Setup

This study was designed to create a longitudinal dataset of physiological and emotional assessments for emerging adults. This dataset can be used for studying affect and correlations of mental health, affect, physiology, sleep, and activity of emerging adults. 

However, as COVID-19 pandemic and lockdown started during the study, the purpose of this study was shifted toward studying the effects of the lockdown on participants' life and mental health. To adapt to the new conditions, the study design, recruitment materials, and questionnaires were updated as the study was ongoing.

This study collected objective ubiquitous data from wearable devices and life-logger apps and combined them with subjective ecological momentary assessments (EMAs) and surveys to create a spectrum of physiological and mental profile for each participant over time. To achieve this objective, this study was required to collect different modalities of data. First, the participants needed to be objectively monitored for their sleep and activity. While many wearable devices can provide accurate measurements on sleep and activity, collecting raw data can be useful in extracting additional features and creating opportunities for future research. Second, the participants' mood and mental state needed to be monitored by subjectively asking them about their mood and mental state. 

In order to collect this dataset, it was necessary to build an Internet platform that can ubiquitously monitor the participants over time, and provide tools for the study coordinator to monitor the progress and data collection remotely. To this end, ZotCare [5] was utilised. ZotCare is an online mHealth service that can be used to collect, store, and analyse the data on different levels of technology stack. Services used in this study include the collection service for collecting objective data from participants' wearable devices and collecting subjective EMAs through mobile app. The collected data then was processed by ZotCare’s cloud server and stored in a database. Besides that, ZotCare’s dashboard was utilised in this study to monitor the collected data and ensure participants' collaboration in data collection. 

To collect the subjective EMAs, ZotCare’s frontend app was customised in this study and was published as Personicle Questions in Google Play. In this application participants would answer daily EMA questionnaires and weekly ones. Both EMAs contain questionnaires that focus on participants’ emotions and their worries about their health in regards to COVID-19. The weekly EMAs additionally contains participants’ written responses about the highs and lows of the previous week. The emotion questions stem from an established measurement, the Positive and Negative Affect Scale (PANAS) [6], while others are brief self-developed, single-item questions that help get an understanding of participants’ overall emotions and worries as a result of the largely-impactful event at the time, the COVID-19 pandemic.  The participants could answer the daily questionnaire from 7:50 PM to 3:00 AM PST time. Every participant was notified at 8PM every night to respond to this questionnaire. If the participant did not respond to the EMA, two follow up reminders would be sent to them at 10PM and 11:59 PM. The weekly EMA was open for participants from 7:50 AM to 8 PM PST time every Sunday with the first reminder at 8 AM and two follow up reminders at 10 AM and noon in case they failed to respond to the EMA before the reminder.

In order to collect raw objective data of the study, ZotCare Tizen wearable app was used to collect the data directly from the participant’s watch. This app was set to collect raw data from sensors such as ppg, accelerometer, gyroscope with frequency of 20Hz, and pressure with frequency of 10Hz every 2 hours for 12 minutes. Besides the raw sensors, the samsung watch also collected data for pedometer and awake times. The pedometer data is supposed to contain step counts and more details of the participants' stepping pace. The frequency of pedometer data is random since it depends on the frequency of the steps taken and its detection by the watch. The awake times data shows the times that the watch detects that the participant is awake. 

In this experiment, Oura ring was used to monitor participants' activity and sleep. Oura is one of the most accurate commercially available wearable devices for sleep monitoring and its battery life is higher than smartwatches. The data collected by Oura is stored in Oura’s servers. ZotCare can use Open Authentication to collect these data from Oura and make them available to researchers on daily bases.

The Personicle Android mobile application was used to collect participants' data, which collected and processed the Google location, Google place, mobile device-specific measures like Calendar, and Ambient Light to identify points of interest and predict the activities performed by the user on a daily basis. Datapoints Personicle, as a mobile application, collected the phone data and sent it to the server for further processing, which, in turn, created the following data points: Lifelog: This was the raw data collected on every 5-minute interval from the phone. Point of interest (POI): This was the derived location/place of a user based on the state of the user (still/motion), Google location, and place API. Home and Work locations were also set as POI after collecting the data for 2 days from the date of app installation. Segment: Segments, in general, were the aggregated/filtered results from lifelog and POI between the occurrences of events. Changes in the state (still/motion) of a person or events like shopping or sleep were considered as events. We are providing access to Segments in this dataset.

The last data collection modality was subjective assessments. These assessments were collected upon participant’s enrollment, their exit, and every month of their participation in the study. Also another assessment was collected from participants about the events that have happened during their enrollment in the study and their effect on their mental state and emotions.These assessments were developed using Qualtrics software. Each assessment was a combination of different standard and self-developed questionnaires that is represented in the following table.

Measure

Description

Source

Technology Acceptance Model Measurement (Modified)

Modified to assess a user’s acceptance of the Personicle application/device. The constructs include perceived ease of use, perceived usefulness, attitude,  behavioural intention, self-efficacy, subjective norm, and system accessibility.  

Park, S. Y. (2009). An analysis of the technology acceptance model in understanding university students'

behavioural intention to use e-Learning. Educational Technology & Society, 12 (3), 150–162.

Acceptability of Intervention Measure (AIM-Modified), Intervention Appropriateness Measure (IAM-

Modified), and Feasibility of Intervention Measure (FIM-Modified)

Assesses the implementation

outcomes of the Personicle devices and application in terms of acceptability, appropriateness and feasibility. 

Weiner, B. J., Lewis, C. C., Stanick, C., Powell, B. J., Dorsey, C. N., Clary, A. S., ... & Halko, H. (2017).

Psychometric assessment of three newly developed implementation outcome

measures. Implementation Science, 12(1), 108.

Beck Depression Inventory II (BDI-II)

Assesses an individual’s depression severity ranging from mild to  severe.  

Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Beck depression inventory manual (2nd ed.). New York, NY: The Psychological Corporation.

Positive and Negative Affect Schedule (PANAS-SF)

Measures an individual’s positive and negative emotions.  

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive  and negative affect: the PANAS scales. Journal of personality and social psychology, 54(6), 1063. 

Brief Symptom Inventory (BSI)

Evaluates current or past level of symptomatology, intensity of symptoms,  and number of reported symptoms. The questionnaire covers nine dimensions of symptoms including  somatization, obsession-compulsion, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety,  paranoia, and psychosis. 

Derogatis, L. R., & Fitzpatrick, M. (2004). The SCL-90-R, the Brief Symptom Inventory (BSI), and the BSI 18. In M. E. Maruish (Ed.), The use of psychological testing for treatment planning and outcomes  assessment: Instruments for adults (pp. 1-41). Mahwah, NJ, US: Lawrence Erlbaum Associates Publishers. 

Experiences in Close Relationships Scale Short Form (ECR-S)

Assesses attachment avoidance and  attachment anxiety.  

Brennan, K., Clark, C., & Shaver, P. R. (1998). Self-report measurement of adult attachment: An integrative  overview. In J. A. Simpson (Ed.), Attachment theory and close relationships (pp. 46–76). New York:  Guilford Press. 

UCLA Three-Item Loneliness Scale (ULS)

Assesses loneliness. 

Hughes, M. E., Waite, L. J., Hawkley, L. C., & Cacioppo, J. T. (2004). A short scale for measuring loneliness  in large surveys: Results From two population-based studies. Research on going, 26(6), 655-672. 

Multidimensional Scale of Perceived Social Support (MSPSS)

Assesses social support from family, friends, and  significant other.  

Zimet, G.D., Dahlem, N.W., Zimet, S.G. & Farley, G.K. (1988). The multidimensional scale of perceived  social support. Journal of Personality Assessment, 52, 30-41. 

Investment Model Scale (IMS)

Measures four constructs (commitment level, satisfaction level, quality of  alternatives, an investment size).  

Rusbult, C. E., Martz, J. M., & Agnew, C. R. (1998). The Investment Model Scale: Measuring commitment  level, satisfaction level, quality of alternatives, and investment size. Personal Relationships, 5(4), 357-391. 

Inclusion of Other in the Self Scale (IOS-Partner)

Measures how close the respondent feels with another person (romantic partner).  

Aron, A., Aron E. N., & Smollan, D. (1992). Inclusion of other in the self scale and the structure of  interpersonal closeness. Journal of Personality and Social Psychology, 63, 596-612.  

Conflict Management Scale (CMS)

Assesses conflict management in romantic relationships.

Mannarini, S., Balottin, L., Munari, C., & Gatta, M. (2017). Assessing conflict management in the couple: The  definition of a latent dimension. The Family Journal, 25(1), 13–22. 

GAD-7

A brief scale that is used to help identify cases of Generalized Anxiety Disorder. 

Spitzer, R. L., Kroenke, K., Williams, J. B., & Löwe, B. (2006). A brief measure for assessing generalized anxiety disorder: the GAD-7. Archives of internal medicine, 166(10), 1092-1097.

Social Isolation

Assesses social isolation in relation to the COVID-19 pandemic. 

Manuscript in progress

Racial Profiling/Bias 

Measures how the respondent interprets interactions with others based on race after the COVID-19 pandemic. 

Manuscript in progress

Health Exposure

Measures the degree of exposure to the coronavirus.

Manuscript in progress

Health Mindset

Measures an individual’s opinion on the effect size of changes in one’s physical health. 

Manuscript in progress

Health Behavioral Changes & Health Impact

Assesses behavioral changes relating to health in conjunction to the COVID-19 pandemic. 

Manuscript in progress

Health Anxiety

Measures how stressed and/or anxious an individual is over their health regarding the COVID-19 pandemic. 

Manuscript in progress

Health Perceived Control

Measures an individual’s perceived control over their health in relation to the COVID-19 pandemic. 

Manuscript in progress

Health Perceived Threat

Assesses the degree that the respondent believes the Coronavirus has or will impact their health.  

Manuscript in progress

Belief in Conspiracy Theories

A brief scale assessing an individual’s belief in conspiracy theories in relation to the COVID-19 pandemic. 

Manuscript in progress

Political Orientation

A single question assessing the degree of political orientation from liberal to conservative. 

Manuscript in progress

Emotion Regulation

Measures how respondents manage emotions in relation to the COVID-19 pandemic and the events surrounding it. 

Manuscript in progress

Brief COPE

Measures efforts of managing stressful events in relation to the COVID-19 pandemic. 

Carver, C. S. (1997). You want to measure coping but your protocol’too long: Consider the brief cope. International journal of behavioral medicine, 4(1), 92-100.

Recruitment and Enrollment

Participants were recruited through UCI faculty and through posted flyers around UCI campus. Participants were between the ages of 18-22, were able to fluently speak and write in English, used an Android operating system of 6.0 or higher on their primary phone (to be compatible with wearable devices), was not a parent, was not married, did not come back to school after more than three years, and was not diagnosed with or met criterias for depression. These criterias were screened for via a phone call after participants expressed interest in participating via email. 

If interested participants were eligible to participate, they were then scheduled for an enrollment session. Participants visited Dr. Borelli’s THRIVE lab to complete the pre-assessment. During this enrollment session, participants completed a battery in which they were asked specific questions that re-confirmed eligibility for the study. Additionally, we collected select demographic information from them (i.e., age, year in school, gender, ethnicity). The research assistant provided information about the study and reviewed the consent form. Once the participant consented, the participant then completed psychological assessments (e.g., BDI-II, PANAS, BSI). If during the pre-assessment, the participant happened to score high on the BDI-II (i.e., a score that indicates moderate depression) or if the participant indicated suicidality, the PI, Dr. Borelli who is a liscensed clinical psychologist, contacted the participant to perform additional psychological assessments on suicide, depression, and social support screening before deciding if the participant can continue in the study or should be withdrawn. In case the participant needed additional assistance, health referrals and resources were provided to them and they were compensated in a prorated way.

If the participants were selected for the study, they had to go through an enrollment session. During the enrollment session, the research team helped set up the wearable smartwatch and smartring, download all necessary applications onto the participants’ phone (Personicle,  Personicle Questions, Oura, Galaxy Wearable), and review device care and study expectations. After the session, participants would then receive an email with $30 of compensation in the form of an Amazon gift-code.  

The majority of the participants were recruited during the COVID-19 pandemic. To address the social distancing orders, we modified our procedures so that the enrollment process could be completed 100% remotely. Participants would be asked to measure their ring at home (this was done using a measuring tape) to get an estimation of their ring size for the Oura ring. Participants were then mailed a shipping package, which contained the monitoring devices (watch and Oura ring). In this package, they were also given a prepaid postage stamp and an  address label for them to use when returning their devices. Participants were given instructions for how to use the devices during a remote pre-assessment visit with a research member, which was accomplished via Zoom. During this remote session, the research team would guide the participants through the same procedures as an in-person visit and participants were compensated the same amount after the session. 

Data Collection

Participation was originally set for 3 months, but participants had the option to continue for an additional 3 months afterwards, up to 3 more times for a maximum of 12 months of total participation. Participants completed a followup assessment at the end of the initial 3-month study period, and completed additional follow ups after each 3-month period if they opted in. Our research team would monitor for incoming data daily and would send reminders to participants to complete surveys if 2-3 were incomplete in a row. If watch or ring data was not submitted regularly, our research team would email the participant a reminder to wear the device and would follow-up with troubleshooting methods if necessary. Participants were compensated $15 per week via email in the form of an Amazon gift-code. 

Exit

At the end of the 12 week period, participants were emailed a Qualtrics link with their ID# for them to complete their exit assessment which took approximately 30 minutes. The exit assessment consisted of the same questions as the baseline assessment in addition to event-mining questions that asked participants to rate how large events impacted their life (i.e.,the impact of the first COVID-19 vaccine rollout). Participants were then given instructions to delete all the phone applications that were downloaded during the start of the study (Personicle,  Personicle Questions, Oura, Galaxy Wearable). 

All participants who were enrolled in the study between January 2020 and March 2020 were invited to an extension of the current study. At the end of the initial 12-week period, participants completed their exit survey (30 minutes). After completion of the survey, the research coordinator would then review the participant’s scores on the BDI-II and follow the same protocol for screening and professional assistance as during the enrollment session. Participants were then given the opportunity to electronically sign a new consent form that asked them to consent to be involved in a continuation of the study (this occurred via DocuSign). If participants agreed to participate in the study continuation, they then continued wearing the devices and completing the surveys for an extended period of 8 weeks. These 8 weeks would have the same on-going data collection, daily surveys, and weekly surveys. At the end of the extended 8 weeks, participants were provided with a Qualtrics link for their second and final exit survey. Participants were then given instructions to delete all the phone applications that were downloaded during the start of the study (Personicle, Personicle Questions, Oura, Galaxy Wearable) and how to delete the watch’s MAC address off of their OIT account. They were then asked to return the devices and chargers by using the packaging,  postage stamp, and address label that was provided to them when they first received the devices. Participants were then compensated $50 via email in the form of Amazon gift-codes. Participants who decide to continue participation into the study’s extension would receive an additional $50 after the second exit assessment was completed.