Karl Fua

Discovery of Behavioral Markers of Social Anxiety from Smartphone Sensor Data

Better understanding of an individuals smartphone use can help researchers to understand the relationship between behaviors and mental health, and ultimately improve methods for early detection, evaluation, and intervention. This relationship may be particularly significant for individuals with social anxiety, for whom stress from social interactions may arise repeatedly and unexpectedly over the course of a day. For this reason, we present an exploratory study of behavioral markers extracted from smartphone data. We examine fine-grained behaviors before and after smartphone communication events across social anxiety levels. To discover behavioral markers, we model the smartphone as a linear dynamical system with the accelerometer data as output. In a two-week study of 52 college students, we find substantially different behavioral markers prior to outgoing phone calls when comparing individuals with high and low social anxiety.

DEMONS: an integrated framework for examining associations between physiology and self-reported affect tied to depressive symptoms

Depression is a prevalent and debilitating disorder among college students. Advances in mobile technology afford the opportunity to collect heterogeneous data while people are in their natural settings. The aim of the current paper is to propose an integrated framework, DEMONS (DEpression MONitoring Study), for combining passive and active data sources using a wearable sensor and a smartphone application. The ability to combine passive and active longitudinal data with mobile devices allows for better understanding of the temporal relations between self-reported affect and physiological variables (e.g., heart rate variability) linked to depressive symptoms. Adoption of the proposed framework will provide crucial information regarding the development and maintenance of depression in college students, as well as increased opportunities for early detection and intervention.

Contextual Analysis to Understand Compliance with Smartphone-based Ecological Momentary Assessment

Mobile device-based ecological momentary assessment (mobile EMA) is increasingly utilized to capture in situ information about a persons physical and mental health states. Mobile EMA has methodological advantages over traditional survey methods (e.g., decreased recall bias); however, these advantages are reduced by participant noncompliance with EMA protocols. There is a dearth of information about how different participant contexts predict compliance. We examine how different spatiotemporal contexts and participant-phone interactions predict EMA response rate and response latency. Utilizing data from 65 participants during a two-week study, we first extract features from smartphone sensors that characterize participant context (location, social context, activity). We then build and evaluate a classifier to predict participant response rate and response latency for EMA-delivered prompts based on the context features, achieving 78% accuracy. We discuss the implications of our results for improving participant compliance in future health studies that deploy mobile EMAs.

I Did OK, but Did I Like It? Using Ecological Momentary Assessment to Examine Perceptions of Social Interactions Associated With Severity of Social Anxiety and Depression

Socially anxious and depressed individuals tend to evaluate their social interactions negatively, but little is known about the specific real-time contributors to these negative perceptions. The current study examined how affect ratings during social interactions predict later perceptions of those interactions, and whether this differs by social anxiety and depression severity. Undergraduate participants (N = 60) responded to a smartphone application that prompted participants to answer short questions about their current affect and social context up to 6 times a day for 2 weeks. At the end of each day, participants answered questions about their perceptions of their social interactions from that day. Results indicated that the link between negative affective experiences reported during social interactions and the end-of-day report of enjoyment (but not effectiveness) of those experiences was more negative when social anxiety was more severe. The link between negative affective experiences rated during social interactions and the end-of-day report of effectiveness (but not enjoyment) during those social encounters was more negative when depression was more severe. These findings demonstrate the importance of examining self-perceptions of social interactions based both on the extent to which individuals think that they met the objective demands of an interaction (i.e., effectiveness, mastery) and the extent to which they liked or disliked that interaction (i.e., enjoyment, pleasure). These findings also highlight how real-time assessments of daily social interactions may reveal the key experiences that contribute to negative self-evaluations across disorders, potentially identifying critical targets for therapy.

Predicting Social Anxiety From Global Positioning System Traces of College Students: Feasibility Study

Background Social anxiety is highly prevalent among college students. Current methodologies for detecting symptoms are based on client self-report in traditional clinical settings. Self-report is subject to recall bias, while visiting a clinic requires a high level of motivation. Assessment methods that use passively collected data hold promise for detecting social anxiety symptoms and supplementing self-report measures. Continuously collected location data may provide a fine-grained and ecologically valid way to assess social anxiety in situ. Objective The objective of our study was to examine the feasibility of leveraging noninvasive mobile sensing technology to passively assess college students’ social anxiety levels. Specifically, we explored the different relationships between mobility and social anxiety to build a predictive model that assessed social anxiety from passively generated Global Positioning System (GPS) data. Methods We recruited 228 undergraduate participants from a Southeast American university. Social anxiety symptoms were assessed using self-report instruments at a baseline laboratory session. An app installed on participants’ personal mobile phones passively sensed data from the GPS sensor for 2 weeks. The proposed framework supports longitudinal, dynamic tracking of college students to evaluate the relationship between their social anxiety and movement patterns in the college campus environment. We first extracted the following mobility features: (1) cumulative staying time at each different location, (2) the distribution of visits over time, (3) the entropy of locations, and (4) the frequency of transitions between locations. Next, we studied the correlation between these features and participants’ social anxiety scores to enhance the understanding of how students’ social anxiety levels are associated with their mobility. Finally, we used a neural network-based prediction method to predict social anxiety symptoms from the extracted daily mobility features. Results Several mobility features correlated with social anxiety levels. Location entropy was negatively associated with social anxiety (during weekdays, r=−0.67; and during weekends, r=−0.51). More (vs less) socially anxious students were found to avoid public areas and engage in less leisure activities during evenings and weekends, choosing instead to spend more time at home after school (4 pm-12 am). Our prediction method based on extracted mobility features from GPS trajectories successfully classified participants as high or low socially anxious with an accuracy of 85% and predicted their social anxiety score (on a scale of 0-80) with a root-mean-square error of 7.06. Conclusions Results indicate that extracting and analyzing mobility features may help to reveal how social anxiety symptoms manifest in the daily lives of college students. Given the ubiquity of mobile phones in our society, understanding how to leverage passively sensed data has strong potential to address the growing needs for mental health monitoring and treatment.

Remembering or knowing how we felt: Depression and anxiety symptoms predict retrieval processes during emotional self-report.

Researchers and clinicians routinely rely on patients’ retrospective emotional self-reports to guide diagnosis and treatment, despite evidence of impaired autobiographical memory and retrieval of emotional information in depression and anxiety. To clarify the nature and specificity of these impairments, we conducted two large online data collections (Study 1, N = 1,983; Study 2, N = 900) examining whether depression and/or anxiety symptoms would uniquely predict the use of self-reported episodic (i.e., remembering) and/or semantic (i.e., knowing) retrieval when rating one’s positive and negative emotional experiences over different time frames. Participants were randomly assigned to one of six time frames (ranging from at this moment to last few years) and were asked to rate how intensely they felt each of four emotions, anxious, sad, calm, and happy, over that period. Following each rating, they were asked several follow-up prompts assessing their perceived reliance on episodic and/or semantic information to rate how they felt, using procedures adapted from the traditional “remember/know” paradigm (Tulving, 1985). Across both studies, depression and anxiety symptoms each uniquely predicted increased likelihood of remembering across emotion types, and decreased likelihood of knowing how one felt when rating positive emotion types. Implications for the theory and treatment of emotion-related memory disturbances in depression and anxiety, and for dual-process theories of memory retrieval more generally, are discussed.

Dynamically tracking anxious individuals’ affective response to valenced information.

Past research has shown that an individual’s feelings at any given moment reflect currently experienced stimuli as well as internal representations of similar past experiences. However, anxious individuals’ affective reactions to streams of interrelated valenced information (vs. reactions to static stimuli that are arguably less ecologically valid) are rarely tracked. The present study provided a first examination of the newly developed Tracking Affect Ratings Over Time (TAROT) task to continuously assess anxious individuals’ affective reactions to streams of information that systematically change valence. Undergraduate participants (N = 141) completed the TAROT task in which they listened to narratives containing positive, negative, and neutral physically- or socially-relevant events, and indicated how positive or negative they felt about the information they heard as each narrative unfolded. The present study provided preliminary evidence for the validity and reliability of the task. Within scenarios, participants higher (vs. lower) in anxiety showed many expected negative biases, reporting more negative mean ratings and overall summary ratings, changing their pattern of responding more quickly to negative events, and responding more negatively to neutral events. Furthermore, individuals higher (vs. lower) in anxiety tended to report more negative minimums during and after positive events, and less positive maximums after negative events. Together, findings indicate that positive events were less impactful for anxious individuals, whereas negative experiences had a particularly lasting impact on future affective responses. The TAROT task is able to efficiently capture a number of different cognitive biases, and may help clarify the mechanisms that underlie anxious individuals’ biased negative processing.

MoCHA: Designing Games to Monitor Cognitive Health in Elders at Risk for Alzheimer's Disease

MoCHA (Monitoring Cognitive Health using Apps) is a set of tablet-based games designed to provide convenient, low-stress, affordable monitoring of cognitive health for elders at risk of developing Alzheimers disease. Conducting psychological measurement via gameplay poses unique game-design challenges, and there are additional factors to consider when designing games for non-gamer elders who may be, or become, cognitively impaired. In this paper we briefly describe the MoCHA system, identify key design challenges, and show how specific features of the game contribute to meeting these challenges.