2025 Preliminary Program

This preliminary program does not include all sessions and will be updated as additional program information is received. Please check back for the latest program updates. You may also view program information in the Program-at-a-glance.

28 Sessions

Subject Area: Methodology

Clear Filters

Workshop

Thursday May 22, 2025 9:00 AM - 11:50 AM ET (13:00 - 15:50 UTC)

Subject Area: Methodology

Abstract

This workshop aims to develop theoretical knowledge and practical skills (primarily in R) with multilevel models, focusing on considerations for real data analysis. These models are commonly used with nested data structures, including children nested within classrooms or families, or with longitudinal data. Frequentist and Bayesian perspectives will be addressed.

This workshop aims to develop theoretical knowledge and practical skills (primarily in R) with multilevel models, focusing on considerations for real data analysis. These models are commonly used with nested data structures, including children nested within classrooms or families, or with longitudinal data. Frequentist and Bayesian perspectives will be addressed.

Workshop

Thursday May 22, 2025 9:00 AM - 11:50 AM ET (13:00 - 15:50 UTC)

Subject Area: Methodology

Abstract

We’ll begin this workshop by exploring longitudinal data and learning how to characterize and format it appropriately for analysis. Then, we’ll fit multilevel models to our data, focusing on growth curve models. All models will be conceptualized, fit, and interpreted.

We’ll begin this workshop by exploring longitudinal data and learning how to characterize and format it appropriately for analysis. Then, we’ll fit multilevel models to our data, focusing on growth curve models. All models will be conceptualized, fit, and interpreted.

Workshop

Thursday May 22, 2025 9:00 AM - 10:50 AM ET (13:00 - 14:50 UTC)

Subject Area: Methodology

Abstract

Variable selection methods are useful when a researcher isn’t sure which predictors to include in their model. In this workshop we will review and compare modern approaches to address this age-old problem, including LASSO and Bayesian variable selection methods.

Variable selection methods are useful when a researcher isn’t sure which predictors to include in their model. In this workshop we will review and compare modern approaches to address this age-old problem, including LASSO and Bayesian variable selection methods.

Workshop

Thursday May 22, 2025 1:00 PM - 4:50 PM ET (17:00 - 20:50 UTC)

Subject Area: Methodology

Abstract

Structural Equation Modeling (SEM) combines common factor analysis with multiple regression to allow researchers to assess true score relations among constructs of theoretical interest. This workshop presents an overview of the logic, implementation, and interpretation of SEMs. Topics covered include: path analysis, confirmatory factor analysis, and structural regression analysis.

Structural Equation Modeling (SEM) combines common factor analysis with multiple regression to allow researchers to assess true score relations among constructs of theoretical interest. This workshop presents an overview of the logic, implementation, and interpretation of SEMs. Topics covered include: path analysis, confirmatory factor analysis, and structural regression analysis.

Workshop

Thursday May 22, 2025 1:00 PM - 2:50 PM ET (17:00 - 18:50 UTC)

Subject Area: Methodology

Abstract

A registered report is a multi-stage publication that is accepted at a journal prior to data collection and analysis, removing the opportunity for publication bias. This workshop will introduce registered reports, help researchers identify journals, find examples, and provide guidance for writing and navigating the review and publication process.

A registered report is a multi-stage publication that is accepted at a journal prior to data collection and analysis, removing the opportunity for publication bias. This workshop will introduce registered reports, help researchers identify journals, find examples, and provide guidance for writing and navigating the review and publication process.

Flash Talk

Thursday May 22, 2025 4:00 PM - 5:45 PM ET (20:00 - 21:45 UTC)

Subject Area: Methodology

Talks:

  • When Will Change: A Nuanced Understanding of Turning Points through Bayesian Piecewise Growth Models
    Ihnwhi Heo (Presenting Author), Fan Jia (Author), Sarah Depaoli (Author)
  • Navigating the Pitfalls of MICE Algorithms: A Comparative Evaluation of Missing Data Analytical Methods
    Jiepeng Yuan (Presenting Author), David Freire (Author), Xin Tong (Author), Noelle Hurd (Author)
  • Maximizing Personalized Prediction: A Comparison of Individualized Models, Multitask Learning, and Mixed Effects Methods
    Grant King (Presenting Author), Aidan Wright (Author)
  • Bias of Retrospective Reporting in Longitudinal Research
    Chad Davis (Presenting Author), Xin Tong (Author)
  • Different Methods of Quantifying Age As a Predictor of Rater Disagreements Lead to Diverging Conclusions
    Linn Zapffe (Presenting Author), Veronica Cole (Author)
  • Comparing Implicit Versus Self-Report Measures of Body Image Among Black African Women: Evidence of Social Desirability Bias?
    Oyenike Balogun (Presenting Author)
  • Critical Thinking: Change in Concept and Measurement over the Last 100 Years
    Kai Cortina (Presenting Author), Blake Ebright (Author)

Workshop

Friday May 23, 2025 10:30 AM - 1:20 PM ET (14:30 - 17:20 UTC)

Subject Area: Methodology

Abstract

This workshop introduces item response theory (IRT), a class of psychometric models describing interactions between people and test/survey questions. Through lectures, discussions, and interactive examples, we will explore the basic tenets of IRT, common dichotomous and polytomous IRT models, and applications in psychological science.

This workshop introduces item response theory (IRT), a class of psychometric models describing interactions between people and test/survey questions. Through lectures, discussions, and interactive examples, we will explore the basic tenets of IRT, common dichotomous and polytomous IRT models, and applications in psychological science.

Symposium

Friday May 23, 2025 1:00 PM - 2:20 PM ET (17:00 - 18:20 UTC)

Subject Area: Methodology

Abstract

Just-in-time adaptive interventions show great promise for intervening on psychological disorders based on inputs collected from ecological momentary assessment studies. A challenge in this arena is successfully predicting these psychological episodes in order to best offer therapeutic intervention. In this symposium, we offer statistical solutions to this issue.

Just-in-time adaptive interventions show great promise for intervening on psychological disorders based on inputs collected from ecological momentary assessment studies. A challenge in this arena is successfully predicting these psychological episodes in order to best offer therapeutic intervention. In this symposium, we offer statistical solutions to this issue.

Talks:

  • The Use of Conditional Variance in Predicting Critical Psychological Events
    Lindley Slipetz (Presenting Author)
  • The Potential of Early Warning Signs in Predicting Panic Attacks
    Donald Robinaugh (Presenting Author)
  • Building Personalized Machine Learning Models Using Real-Time Monitoring Data to Predict Idiographic Suicidal Thoughts
    Shirley Wang (Presenting Author)

Symposium

Friday May 23, 2025 1:00 PM - 2:20 PM ET (17:00 - 18:20 UTC)

Subject Area: Methodology

Abstract

This symposium showcases recent methodological advances and novel applications of Exploratory Graph Analysis (EGA), highlighting its versatility and robustness across different psychological research contexts. Through four interconnected presentations, we demonstrate how EGA provides innovative solutions to common challenges in dimensional assessment and structural analysis of psychological data.

This symposium showcases recent methodological advances and novel applications of Exploratory Graph Analysis (EGA), highlighting its versatility and robustness across different psychological research contexts. Through four interconnected presentations, we demonstrate how EGA provides innovative solutions to common challenges in dimensional assessment and structural analysis of psychological data.

Talks:

  • Evaluating the Impact of Missing Data Mechanisms on Exploratory Graph Analysis (EGA) for Detecting Latent Structure
    Jiepeng Yuan (Presenting Author)
  • Evaluating the Performance of Exploratory Graph Analysis (EGA) and Other Dimensional Assessment Techniques with Skewed Data
    Chad Davis (Presenting Author)
  • Applications of Dynamic Ega to Analyze Brain Data
    Richard Gallagher (Presenting Author)
  • Autonomy and Relatedness in Mother-Teen Interactions: An Investigation Using Exploratory Graph Analysis
    Natasha Bailey (Presenting Author)

Workshop

Friday May 23, 2025 2:00 PM - 4:50 PM ET (18:00 - 20:50 UTC)

Subject Area: Methodology

Abstract

Data Science for Psychologists introduces you to the principles of data science. In this hands-on workshop, you'll gain a strong foundation in R and the tidyverse, including data wrangling, modeling, visualization, and communication.

Data Science for Psychologists introduces you to the principles of data science. In this hands-on workshop, you'll gain a strong foundation in R and the tidyverse, including data wrangling, modeling, visualization, and communication.

Symposium

Friday May 23, 2025 4:00 PM - 5:20 PM ET (20:00 - 21:20 UTC)

Subject Area: Methodology

Abstract

Despite increased research, severe psychological outcomes have not declined in past decades. This is often attributed to psychology’s lack of formalized, theory-driven research. However, some argue theory alone cannot enhance prediction when key aspects of the research process remain underspecified. This symposium explores phenomena-detection and construct validation as paths forward.

Despite increased research, severe psychological outcomes have not declined in past decades. This is often attributed to psychology’s lack of formalized, theory-driven research. However, some argue theory alone cannot enhance prediction when key aspects of the research process remain underspecified. This symposium explores phenomena-detection and construct validation as paths forward.

Talks:

Invited Symposium

Saturday May 24, 2025 10:30 AM - 11:50 AM ET (14:30 - 15:50 UTC)

Subject Area: Methodology

Abstract

Statistical models enable scientists to describe data, explain phenomena, form testable predictions, and (ideally) understand the world. To achieve such aims, a model and its components must undergo thorough evaluation. In this symposium, we confront the widely accepted methods of model evaluation, expose their limitations, and explore more insightful alternatives.

Statistical models enable scientists to describe data, explain phenomena, form testable predictions, and (ideally) understand the world. To achieve such aims, a model and its components must undergo thorough evaluation. In this symposium, we confront the widely accepted methods of model evaluation, expose their limitations, and explore more insightful alternatives.

Talks:

  • How to Lie With Fit Statistics
    Wes Bonifay (Presenting Author)
  • The Shape of Good Model Fit
    Sonja Winter (Presenting Author)
  • Consideration of Indicator Interchangeability Can Strengthen Construct Validation Efforts
    Ashley Watts (Presenting Author)

Workshop

Saturday May 24, 2025 10:30 AM - 11:50 AM ET (14:30 - 15:50 UTC)

Subject Area: Methodology

Abstract

"The elevator pitch. We talk about it, say it is important for networking and sharing our research. And yet when it comes to actually doing it, we find ourselves using jargon, spending too much time on details that aren’t important, or not giving the right context for our audience.

"The elevator pitch. We talk about it, say it is important for networking and sharing our research. And yet when it comes to actually doing it, we find ourselves using jargon, spending too much time on details that aren’t important, or not giving the right context for our audience.

Invited Talk

Saturday May 24, 2025 12:00 PM - 12:25 PM ET (16:00 - 16:25 UTC)

Subject Area: Methodology

Abstract

We examined whether response styles, such as socially desirable response style (SDR), leads to overestimation of classes in growth mixture models (GMMs). Longitudinal data were simulated from a single-class model with SDR, and GMM's were fit to the resulting data. We found spurious extraction of classes when SDR was present but unmodeled.

We examined whether response styles, such as socially desirable response style (SDR), leads to overestimation of classes in growth mixture models (GMMs). Longitudinal data were simulated from a single-class model with SDR, and GMM's were fit to the resulting data. We found spurious extraction of classes when SDR was present but unmodeled.

Invited Symposium

Saturday May 24, 2025 1:00 PM - 2:20 PM ET (17:00 - 18:20 UTC)

Subject Area: Methodology

Abstract

As a discipline psychology has long wrestled with the promise of big data for a precise accounting of human behavior and mental life. In this symposium we will discuss recent work to address these challenges by combining ideas from traditional dynamic modeling with techniques from machine learning and computational statistics.

As a discipline psychology has long wrestled with the promise of big data for a precise accounting of human behavior and mental life. In this symposium we will discuss recent work to address these challenges by combining ideas from traditional dynamic modeling with techniques from machine learning and computational statistics.

Talks:

  • Intra- and Inter-Individual Differences in Affect Forecasting with Machine Learning Methods
    Sy-Miin Chow (Presenting Author)
  • Multilevel Statistical Inference from Machine Learning Tools Using Surrogate Data Analysis
    Timothy Brick (Presenting Author)
  • Fused Approaches for Multiple-Subject Time-Varying Parameter Models
    Zachary Fisher (Presenting Author)

Workshop

Saturday May 24, 2025 1:00 PM - 2:20 PM ET (17:00 - 18:20 UTC)

Subject Area: Methodology

Abstract

In this workshop, participants experience demos of research presentations told with and without storytelling components and then discuss how a story format pushes the audience’s thinking forward about the research.

In this workshop, participants experience demos of research presentations told with and without storytelling components and then discuss how a story format pushes the audience’s thinking forward about the research.

Invited Talk

Saturday May 24, 2025 2:30 PM - 2:55 PM ET (18:30 - 18:55 UTC)

Subject Area: Methodology

Abstract

Accurately planning sample sizes has been challenging, especially due to uncertainty around nuisance model parameters. This talk introduces an innovative method designed to restore robustness and simplify sample size determination in the structural equation modeling framework, demonstrated through measured variable structural models, aiming to enhance statistical integrity in research.

Accurately planning sample sizes has been challenging, especially due to uncertainty around nuisance model parameters. This talk introduces an innovative method designed to restore robustness and simplify sample size determination in the structural equation modeling framework, demonstrated through measured variable structural models, aiming to enhance statistical integrity in research.

Invited Symposium

Saturday May 24, 2025 3:00 PM - 4:20 PM ET (19:00 - 20:20 UTC)

Subject Area: Methodology

Abstract

A methodological revolution in response to concerns about the credibility and replicability of psychological results has brought attention to concepts such as statistical forensics, power analysis, p-hacking, and accumulating evidence. Four talks in this symposium will touch on these topics, bringing new information on the use of these methods.

A methodological revolution in response to concerns about the credibility and replicability of psychological results has brought attention to concepts such as statistical forensics, power analysis, p-hacking, and accumulating evidence. Four talks in this symposium will touch on these topics, bringing new information on the use of these methods.

Talks:

  • Statistical Power in the Light of Methodological Reform
    Jolynn Pek (Presenting Author)
  • The Poor Statistical Properties of the P-Curve Procedures
    Richard Morey (Presenting Author)
  • Consistent Methods Protect Against False Findings Produced By p-Hacking
    Duane Wegener (Presenting Author)
  • Accumulating Evidence across Studies
    Blakeley McShane (Presenting Author)

Submitted Talk

Saturday May 24, 2025 4:30 PM - 4:45 PM ET (20:30 - 20:45 UTC)

Subject Area: Methodology

Abstract

With increasing access to large datasets, implementing complex multivariate models has become feasible. Yet, identifying important moderators amidst influential variables remains a challenge. Addressing this, we introduce SEM-Boruta, a novel adaptation of the Boruta algorithm, and demonstrate its robustness through application on simulated datasets in a multivariate and longitudinal framework.

With increasing access to large datasets, implementing complex multivariate models has become feasible. Yet, identifying important moderators amidst influential variables remains a challenge. Addressing this, we introduce SEM-Boruta, a novel adaptation of the Boruta algorithm, and demonstrate its robustness through application on simulated datasets in a multivariate and longitudinal framework.

Submitted Talk

Saturday May 24, 2025 4:30 PM - 4:45 PM ET (20:30 - 20:45 UTC)

Subject Area: Methodology

Abstract

Natural language processing is increasingly used to score open-ended text but can often exacerbate adverse impact. I examine the effects of debiasing machine learning admissions essay scores via multi-objective optimization. While reliability and cognitive saturation decrease slightly from optimizing diversity, the main effect is changing what experiences receive high scores.

Natural language processing is increasingly used to score open-ended text but can often exacerbate adverse impact. I examine the effects of debiasing machine learning admissions essay scores via multi-objective optimization. While reliability and cognitive saturation decrease slightly from optimizing diversity, the main effect is changing what experiences receive high scores.

Invited Talk

Saturday May 24, 2025 4:30 PM - 4:55 PM ET (20:30 - 20:55 UTC)

Subject Area: Methodology

Abstract

As datasets grow, identifying key predictors remains a challenge. The LASSO is one method commonly used to perform simultaneous shrinkage and selection, but are you aware of the drawbacks? Bayesian Stochastic Search Variable Selection provides a more robust approach and can easily be applied to applications with incomplete data.

As datasets grow, identifying key predictors remains a challenge. The LASSO is one method commonly used to perform simultaneous shrinkage and selection, but are you aware of the drawbacks? Bayesian Stochastic Search Variable Selection provides a more robust approach and can easily be applied to applications with incomplete data.

Submitted Talk

Saturday May 24, 2025 5:00 PM - 5:15 PM ET (21:00 - 21:15 UTC)

Subject Area: Methodology

Abstract

The presentation introduces a Beta-Bernoulli mixture item factor analytic model (BBM-IFA) for bounded-continuous items from continuous rating scales (CRS) and slider-bar assessments. Via simulation studies and an empirical case study of the Self-Determination Inventory, we show that the BBM-IFA can provide nuanced information about responses process from CRS/slider-bar assessments.

The presentation introduces a Beta-Bernoulli mixture item factor analytic model (BBM-IFA) for bounded-continuous items from continuous rating scales (CRS) and slider-bar assessments. Via simulation studies and an empirical case study of the Self-Determination Inventory, we show that the BBM-IFA can provide nuanced information about responses process from CRS/slider-bar assessments.

Submitted Talk

Saturday May 24, 2025 5:15 PM - 5:30 PM ET (21:15 - 21:30 UTC)

Subject Area: Methodology

Abstract

Missing data are ubiquitous in psychological research, particularly at the item level. There are not easily implemented methods for estimating effect sizes and relative importance in regression models using modern missing data methods. We propose and evaluate a two-stage method for computing these parameters with item level missing data.

Missing data are ubiquitous in psychological research, particularly at the item level. There are not easily implemented methods for estimating effect sizes and relative importance in regression models using modern missing data methods. We propose and evaluate a two-stage method for computing these parameters with item level missing data.

Submitted Talk

Sunday May 25, 2025 9:00 AM - 9:15 AM ET (13:00 - 13:15 UTC)

Subject Area: Methodology

Abstract

The abbreviated version of the Norse Feedback is reliable and provides clinically equivalent information to the full measure in a clinical sample (N = 2923). This psychometrically valid shortened measure is more feasible for use in routine care, particularly for screening and triage, where time or resources may be limited.

The abbreviated version of the Norse Feedback is reliable and provides clinically equivalent information to the full measure in a clinical sample (N = 2923). This psychometrically valid shortened measure is more feasible for use in routine care, particularly for screening and triage, where time or resources may be limited.

Symposium

Sunday May 25, 2025 9:00 AM - 10:20 AM ET (13:00 - 14:20 UTC)

Subject Area: Methodology

Abstract

This symposium examines the application of artificial intelligence tools that analyze multimodal data (video, audio, text) to track fluctuations in mental states. Presentations will showcase how integrating these complementary data streams provides valuable insights to inform therapist decision-making and personalize digital interventions.

This symposium examines the application of artificial intelligence tools that analyze multimodal data (video, audio, text) to track fluctuations in mental states. Presentations will showcase how integrating these complementary data streams provides valuable insights to inform therapist decision-making and personalize digital interventions.

Talks:

Submitted Talk

Sunday May 25, 2025 9:15 AM - 9:30 AM ET (13:15 - 13:30 UTC)

Subject Area: Methodology

Abstract

We investigated the optimal number of response categories for Likert and Visual Analog Scales (VAS) using simulations. Reliability plateaus after 4–10 categories; however, when measurement error increases with more response options, reliability declines, and converting Likert items to VAS can drastically decrease reliability.

We investigated the optimal number of response categories for Likert and Visual Analog Scales (VAS) using simulations. Reliability plateaus after 4–10 categories; however, when measurement error increases with more response options, reliability declines, and converting Likert items to VAS can drastically decrease reliability.

Submitted Talk

Sunday May 25, 2025 9:30 AM - 9:45 AM ET (13:30 - 13:45 UTC)

Subject Area: Methodology

Abstract

This study confirms the cross-cultural validity of the Multidimensional Self-Control Scale across the USA, Romania, and Norway. Using integrated MCFA and IRT methods, results demonstrate full measurement invariance and robust psychometric properties, highlighting the scale’s universal applicability in assessing self-control dynamics.

This study confirms the cross-cultural validity of the Multidimensional Self-Control Scale across the USA, Romania, and Norway. Using integrated MCFA and IRT methods, results demonstrate full measurement invariance and robust psychometric properties, highlighting the scale’s universal applicability in assessing self-control dynamics.

Submitted Talk

Sunday May 25, 2025 9:30 AM - 9:45 AM ET (13:30 - 13:45 UTC)

Subject Area: Methodology

Abstract

We introduce a mathematical theory of optimal experimental manipulation that enables maximal change in psychological traits of various stimuli using generative networks. Across three studies (N = 417) we demonstrate superior performance over existing approaches for face image manipulation, particularly for previously unattainable increase in valence-laden impressions.

We introduce a mathematical theory of optimal experimental manipulation that enables maximal change in psychological traits of various stimuli using generative networks. Across three studies (N = 417) we demonstrate superior performance over existing approaches for face image manipulation, particularly for previously unattainable increase in valence-laden impressions.