About

Timothy Brick is an Assistant Professor of Human Development & Family Studies and a Graduate Faculty member at the Social Data Analytics (SoDA) program at Pennsylvania State University. He is also a faculty affiliate of the Center for Social Data Analytics (C-SoDA). His research focuses on social data analytics, applying data-driven methods to understand human development and family studies. Brick is actively involved in the academic community, contributing to the advancement of social data analytics through his teaching, research, and collaboration with other faculty and students at Penn State.

Research topics

• Computer Science
• Psychology
• Artificial Intelligence
• Developmental psychology
• World Wide Web
• Human–computer interaction
• Psychotherapist
• Cognitive psychology
• Data science
• Social psychology
• Embedded system
• Statistics

Selected publications

• Adaptive Intervention Designs for Studies of Human Development

  2026-02-20

  other1st authorCorresponding

  Observational and experimental designs for the study of human development almost always follow a single common pattern: Data analysis begins only after data collection has been completed. Deviations from this pattern are extremely rare. Yet a number of options exist to enable active adaptation of study design, using insights from the data to adapt the design of the study while data collection is still active. These approaches can increase power, lower costs, and enable completely new capabilities. Although such methods often require careful planning and more advanced statistics, the increasing availability of software and technology has made them more and more accessible to modern scientists. This chapter provides a high-level overview of adaptive design strategies, including the history, challenges, and promise of these approaches, with a specific focus on those methods that are most applicable to the types of longitudinal studies common in the field of human development.

  PublisherDOI

• CRAVE: Cross-Reality Analytics for Virtual Environments — deepening insights into human food selection and eating behavior

  Computers in Human Behavior Reports · 2026-04-01

  articleOpen access

  Virtual reality (VR) has long been hailed as an effective tool for behavioral research studies combining experimental control with ecological validity. Two frequently used categories of behavioral studies using VR can be distinguished: behavior in VR is studied as a proxy for behavior in physical reality, or behavior change in physical reality is studied after experiencing interventions in VR. In this paper, we explore and discuss a third category: behavior in VR and behavior in physical reality are analytically combined to leverage the experiential differences between the two, and to better understand human behavior. For this approach, we coin the term cross-reality analytics , where behavior in multiple realities is observed, combined and analyzed. We place these three categories in a proposed conceptual framework, and discuss similarities and differences between cross-reality analytical studies, proxy and interventions studies. To illustrate cross-reality analytics, we also present hitherto unpublished results from a cross-reality study, analyzing food selection data collected in VR and connecting it with eating behavior data in physical reality. Results show that data on virtual food choices (e.g., virtual fruit first) are associated with physical eating behavior (e.g., total physical food consumed), and that additional insights were gained about human behavior by analyzing unique behavior in VR (e.g., physically impossible portion size manipulation) and combining it with data from physical reality (e.g., total physical food consumed). The results illustrate the potential of cross-reality analytics for developing a deeper understanding of behavioral traits, and show promise for future behavioral research. • We introduce cross-reality analytics for VR behavioral research. • VR studies fall into proxy, intervention, or cross-reality analytical categories. • Cross-reality analytics enables joint analysis of behavior from multiple realities. • A cross-reality virtual buffet-real eating study revealed novel behavioral insights. • Behaviors exclusive to VR and sensory mismatches reveal new behavioral insights.

  PublisherDOI

• ByteTrack: a deep learning approach for bite count and bite rate detection using meal videos in children

  Frontiers in Nutrition · 2025-10-03 · 2 citations

  articleOpen access

  Introduction: Assessing eating behaviors such as eating rate can shed light on risk for overconsumption and obesity. Current approaches either use sensors that disrupt natural eating or rely on labor-intensive video coding, which limits scalability. Methods: We developed ByteTrack, a deep learning system for automated bite count and bite-rate detection from video-recorded child meals. The dataset comprised 1,440 minutes from 242 videos of 94 children (ages 7-9 years) consuming four meals, spaced one week apart, with identical foods served in varying amounts. ByteTrack operates in two stages: (1) face detection via a hybrid Faster R-CNN and YOLOv7 pipeline, and (2) bite classification using an EfficientNet convolutional neural network combined with a long short-term memory (LSTM) recurrent network. The model was designed to handle blur, low light, camera shake, and occlusions (hands or utensils blocking the mouth). Performance was compared with manual observational coding. Results: On a test set of 51 videos, ByteTrack achieved an average precision of 79.4%, recall of 67.9%, and F1 score of 70.6%. Agreement with the gold-standard coding, assessed by intraclass correlation coefficient, averaged 0.66 (range 0.16-0.99), with lower reliability in videos with extensive movement or occlusions. Discussion: This pilot study demonstrates the feasibility of a scalable, automated tool for bite detection in children's meals. While results were promising, performance decreased when faces were partially blocked or motion was high. Future work will focus on improving robustness across diverse populations and recording conditions. Clinical trial registration: https://clinicaltrials.gov/study/NCT03341247, identifier NCT03341247.

  PublisherOA PDFDOI

• Visualising and understanding human evaluation of house facades: GAN applied to environmental psychology

  Journal of Environmental Psychology · 2025-10-09 · 1 citations

  articleOpen access

  Generative adversarial networks (GANs) are a powerful deep-learning method for creating and manipulating images. In this paper, we investigated the use of GANs in environmental psychology and architecture to analyse human evaluations of architectural facades. We trained StyleGAN2-ADA, a state-of-the-art GAN model, on a dataset of 2000 house images collected for the study (CalHouses). Each house was labelled with rating scores describing how it was perceived. The goal of the first study was to generate labels through an online experiment with 204 participants. Each image was rated by 10 participants on five psychological dimensions: hominess , safety , invitingness , relaxation , and perceived price . The goal of the second study was to evaluate 2000 artificial images, generated by the GAN, by having another sample of 204 participants rate the images on the same psychological dimensions. The statistical analyses showed that participants’ ratings of the GAN-generated images aligned with the targeted characteristics of the psychological dimensions used during generation. A visual analysis of the artificial images indicated that the degree of naturalness, the size and complexity of the house, and the number of openings are potentially relevant features for the evaluation of detached houses on the five investigated psychological dimensions. To the best of our knowledge, this is the first study to utilise GANs to analyse architectural design relative to human evaluations, highlighting its potential as a research method in environmental psychology to investigate architecture. • StyleGAN2-ADA was used to analyse human evaluations of architectural house facades • Artificial images showed houses representing psychological attributes, e.g., safety • Results indicated that GAN could predict participants' ratings on house facades • Visual analysis revealed naturalness, size, and openings as important features

  PublisherDOI

• Idiolectic Models for Diagnostics: A Novel Approach to Understanding Semantic Relationships in Clinical Populations

  2025-03-30

  preprintOpen accessSenior author

  This study presents a novel method, Idiolectic Models for Diagnostics, to analyze and understand unique semantic relationships in clinical populations. The initial paper presents the methodological foundation for the Idiolectic Models for Diagnostics, with subsequent papers focusing on clinical populations. Our method aims to elucidate personal idiolectic connections, providing a deeper understanding of the semantic landscape within and between individuals. This study analyzed group-level differences in the text of forum posts on a popular social media site and individual-level idiolect comparisons in OCD populations compared to more general population. Our results demonstrate significant differences in the semantic associations with the word "attract" between OCD users and general users. Specifically, OCD users exhibited more varied and less consistent associations, reflecting the diverse nature of their obsessions and compulsions, while general users showed more stable and uniform associations. Additionally, group-level comparisons between art and programming subgroups revealed significant differences in semantic distances in associations with the word “abstract,” indicating distinct word usage patterns across communities. These findings highlight the potential of Idiolectic Models for Diagnostics in uncovering meaningful differences in semantic relationships within clinical populations and online communities, both at group- and individual-level.

  PublisherOA PDFDOI

• Idiolectic Models for Diagnostics: A Novel Approach to Understanding Semantic Relationships in Clinical Populations

  2025-01-31

  preprintOpen accessSenior author

  This study presents a novel method, Idiolectic Models for Diagnostics, to analyze and understand unique semantic relationships in clinical populations. The initial paper presents the methodological foundation for the Idiolectic Models for Diagnostics, with subsequent papers focusing on clinical populations. Our method aims to elucidate personal idiolectic connections, providing a deeper understanding of the semantic landscape within and between individuals. This study analyzed group-level differences in the text of forum posts on a popular social media site and individual-level idiolect comparisons in OCD populations compared to more general population. Our results demonstrate significant differences in the semantic associations with the word "attract" between OCD users and general users. Specifically, OCD users exhibited more varied and less consistent associations, reflecting the diverse nature of their obsessions and compulsions, while general users showed more stable and uniform associations. Additionally, group-level comparisons between art and programming subgroups revealed significant differences in semantic distances in associations with the word “abstract,” indicating distinct word usage patterns across communities. These findings highlight the potential of Idiolectic Models for Diagnostics in uncovering meaningful differences in semantic relationships within clinical populations and online communities, both at group- and individual-level.

  PublisherDOI

• A within-person investigation of recovery identity following substance use disorder: examining the impact of recovery-focused social contexts

  Frontiers in Public Health · 2025-02-12 · 2 citations

  articleOpen access

  Background The social identity model of recovery (SIMOR) posits that adopting a recovery identity is vital for achieving favorable recovery outcomes. Until now, no studies have investigated recovery identity as a dynamic construct, although recent findings suggest it fluctuates from one day to the next. The present study examines the within-person association between recovery identity and sense of meaningfulness—an aspect of holistic recovery wellbeing. Because recovery-focused social contexts exist to support individuals’ recovery wellbeing, we assessed the moderating impact of two such contexts (recovery community centers [RCCs] and recovery meetings) as same-day moderators. Methods and materials 91 RCC visitors across Pennsylvania completed daily diary surveys for 10 evenings. Daily measures of recovery identity, meaningfulness, recovery meeting and RCC attendance were analyzed in a multilevel Tobit model (to address right-censoring in the outcome data). Results Results indicated both day-level recovery identity ( b = 0.79, SE = 0.04, p < 0.001) and person-level recovery identity ( b = 0.94, SE = 0.11, p < 0.001) were positively associated with daily meaningfulness. Although the day-level interaction with RCC attendance was not significant ( b = −0.11, SE = 0.14, p = n.s. ), the interaction with recovery meeting attendance was ( b = −0.27, SE = 0.13, p = 0.039), suggesting that meeting attendance buffered the effect of recovery identity on meaningfulness. A simple slopes analysis indicated that the relationship of recovery identity with meaningfulness was still statistically significant and positive in both cases (attended: b = 0.56, SE = 0.08, p < 0.001; not attended: b = 0.87, SE = 0.06, p < 0.001). Conclusion These results suggest that people reporting stronger recovery identity also reported greater day-to-day meaningfulness. Further, on any given day for an individual, meaningfulness was higher on days recovery identity was stronger than usual for that individual, and lower on days when recovery identity was weaker. Meeting attendance reduced this effect, suggesting that meeting attendance may be especially helpful to recovery on days when recovery identity is low.

  PublisherOA PDFDOI

• Energy Density Selected in Immersive Virtual Reality Buffet Meals Is Associated with Both Energy Density Consumed and Energy Intake in Laboratory Meals

  2025-07-09 · 1 citations

  article

  Innovative methods are needed to understand associations between food selection and intake across different contexts. We explored whether the energy density (ED, kilocalories per gram) of meals selected in an immersive virtual reality (iVR) food buffet predicted the ED consumed and overall energy intake in measured laboratory meals. In a secondary analysis, 91 adults (64 female, aged 18 to 71 years) selected foods for a meal in our iVR buffet before consuming a standard laboratory meal once a week for two weeks. The iVR buffet contained 30 foods varying in ED, ranging from 0.3 to 4.9 kcal per gram, including entrees, sides, soups, and desserts. The laboratory meals consisted of pasta, rolls, chicken, broccoli, grapes, and cookies, with ED values ranging from 0.4 to 4.8 kcal per gram. Linear mixed-effect models were used to examine associations between food selections in iVR meals and intake in laboratory meals. We found that the ED selected in iVR significantly predicted the ED consumed in laboratory meals. The ED consumed in laboratory meals and the ED selected in IVR meals were both positively associated with energy intake in laboratory meals. In addition, the carbohydrates, fats, and protein selected in iVR meals were each significantly associated with their respective intake in laboratory meals. The significant associations between food selections in iVR meals and intake in laboratory meals demonstrate the predictive validity of iVR. These findings highlight the utility of iVR as an innovative method to assess associations between food selection and intake across diverse contexts.

  PublisherDOI

• Use of a GLP-1 Receptor Agonist Shows Promise for Reducing Craving in Opioid Use Disorder

  SSRN Electronic Journal · 2025-01-01

  preprintOpen access
  PublisherDOI

• CRAVE: Cross-Reality Analytics for Virtual Environments - Deepening Insights into Human Food Selection and Eating Behavior

  SSRN Electronic Journal · 2025-01-01

  preprintOpen access
  PublisherDOI

Frequent coauthors

• Michael D. Hunter

  22 shared

• Simone Kühn

  University Medical Center Hamburg-Eppendorf

  21 shared

• Steven M. Boker

  University of Virginia

  20 shared

• Zita Oravecz

  Pennsylvania State University

  14 shared

• Michael C. Neale

  Virginia Commonwealth University

  14 shared

• Ryne Estabrook

  University of Illinois Chicago

  12 shared

• Joshua N. Pritikin

  Virginia Commonwealth University

  12 shared

• Hermine H. Maes

  Virginia Commonwealth University

  11 shared

Education

• PhD, Psychology

  University of Virginia

  2011

• MS, Computer Science and Engineering

  University of Notre Dame

  2007

• BA, Psychology

  University of Notre Dame

  2002

• BS, Computer Science and Engineering

  University of Notre Dame

  2001