Psychological processes are highly heterogeneous, even among individuals with the same diagnosis. This variability poses challenges for nomothetic approaches that assume everyone is guided by the same broad psychological principles. In contrast, idiographic approaches focus on within-person variability but are often prone to noise and spurious relations and may not translate easily to clinical use due to limited generalizability. These constraints have motivated integrative approaches designed to model person-specific dynamics while still drawing on patterns that generalize across people. In this article, we review group iterative multiple model estimation (GIMME), one of the most widely used integrative approaches for modeling intensive longitudinal data (ILD) in clinical research. GIMME estimates person-specific dynamics using majority-shared paths as the backbone of individual models. We begin by introducing GIMME's core algorithm and its major extensions. We then review simulation studies evaluating its performance, survey empirical applications in clinical psychology, and outline alternative ILD methods. Finally, we discuss current limitations of GIMME and propose directions for its continued refinement.

Background: Individuals with bipolar disorder (BD) face an elevated suicide risk. While machine learning (ML) has been used to estimate suicide risk in BD, early predictors like demographics, past attempts, and self-reports are limited by their inability to provide individualized risk estimation, overemphasis on past attempters, and susceptibility to personal biases, underscoring the need for effective, objective markers. Event-related potentials (ERPs), widely studied in suicide research, remain unexplored in ML applications for BD. This pilot study applies ML to N200 and P300 ERP components from a response inhibition paradigm to estimate suicide risk in BD. Methods: We collected N200 and P300 peak amplitude and latency data from 57 Type I BD individuals (22 attempters and 35 non-attempters). Our two-stage ML approach employed adaptive Lasso logistic regression for feature selection, followed by deep neural network (DNN) modeling for classification. For post-hoc analysis, we used explainable AI to interpret ERP feature importance in top-performing DNN predictions. Results: Key features were exclusively identified from latency data. Notably, N200 latency DNN models effectively distinguished attempters from non-attempters, achieving AUCs of 78.2-89.3 %. Explainable AI pinpointed a right visual hemifield Go stimuli-induced ERP from the left-parietal site as the most predictive. Conclusion: Our ERP-ML approach showed promising preliminary results, with N200 latency identified as a potential suicide marker in BD. Larger samples are required to validate these results. While findings are samplespecific, the methodological approach may have broader applicability and could inform future research to refine clinical strategies for detecting high-risk BD individuals.

ObjectiveThis is an exploratory study aiming to investigate the contemporaneous and lagged relationships between therapeutic alliance, emotional experience, and the focus on Affect Phobia Therapy objectives, a psychodynamic psychotherapy, while also examining their impact on treatment outcomes.MethodA total of seven dyads with successful treatments were included. Clients aged between 20 and 44 years (M = 29.14; SD = 8.43) rated their therapeutic alliance, emotional experience and symptoms after each session while therapists rated alliance and therapeutic interventions. Data were analyzed with Group Iterative Multiple Model Estimation (GIMME), an analytical procedure to identify patterns in time-series data.ResultsOnly one link was identified at the group-level. The remaining interactions among process variables were divided into two distinct subgroups. In one subgroup, it was possible to identify links in line with theoretical expectations, revealing connections between relational and technical factors, whereas in the other, the connections predominantly focused on the therapist's perspective.ConclusionsOur results indicate that despite overall success, individuals may experience the same therapy differently, with only some processes aligning closely with the theoretical model. This variability seems linked to potential moderators like initial level of functioning, highlighting the need to thoroughly assess client characteristics to tailor treatments effectively.Practical implicationsThis study explores the intricate interplay between therapeutic alliance, emotional experience, and focus on psychodynamic objectives within Affect Phobia Therapy (APT), using an innovative dynamic systems approach (Group Iterative Multiple Model Estimation). The research revealed distinct links within two subgroups, emphasizing the idiographic nature of therapeutic processes. This helps to better tailor the interventions considering the different dynamics of interrelationships between process variables and the specific characteristics of individual clients.

In this study, we extend the dynamic fit index (DFI) developed by McNeish and Wolf to the context of time series analysis. DFI is a simulation-based method for deriving fit index cutoff values tailored to the specific model and data characteristics. Through simulations, we show that DFI cutoffs for detecting an omitted path in time series network models tend to be closer to exact fit than the popular benchmark values developed by Hu and Bentler. Moreover, cutoff values vary by number of variables, network density, number of time points, and form of misspecification. Notably, using 10% as the upper limit of Type I and Type II error rates, the original DFI approach fails to identify cutoffs for detecting an omitted path when effect size and/or sample size is small. To address this problem, we propose two alternatives that allow for the derivation of cutoffs using more lenient criteria. DFIA extends the original DFI approach by removing the upper limit of Type I and Type II error rates, whereas DFIB aims at maximizing classification quality measured by the Matthews correlation coefficient. We demonstrate the utility of these approaches using simulation and empirical data and discuss their implications in practice.

The increasing use of learning management systems (LMSs) generates vast amounts of clickstream data, opening new avenues for predicting learner performance. Traditionally, LMS predictive analytics have relied on either supervised machine learning or Markov models to classify learners based on predicted learning outcomes. Machine learning excels at pattern recognition but often overlooks temporal learning dynamics and obscures the reasoning behind predictions due to the black-box nature of many algorithms. Alternatively, Markov models provide an effective solution by capturing temporal learning dynamics for prediction, uncovering distinctive learning patterns between high and low performers. Despite these advantages, Markov model classification struggles with the heterogeneity of learning sequences, limiting its broad applicability. To address these limitations and bridge the gap between the two dominant approaches, we propose a hybrid framework: sequence-based Markov machine learning classification (seqMAC). Leveraging early-stage clickstream data, seqMAC provides an interpretable sequence classification method that captures critical behavioural transitions and identifies distinct learning patterns across performance groups. Tested on six LMS samples, seqMAC effectively identified at-risk students despite sequence heterogeneity, uncovering key predictive learning dynamics that differentiate performance groups. It also demonstrated promising generalizability, accurately identifying future at-risk students based on historical clickstream data.

The increasing use of learning management systems (LMSs) generates vast amounts of clickstream data, opening new avenues for predicting learner performance. Traditionally, LMS predictive analytics have relied on either supervised machine learning or Markov models to classify learners based on predicted learning outcomes. Machine learning excels at pattern recognition but often overlooks temporal learning dynamics and obscures the reasoning behind predictions due to the black-box nature of many algorithms. Alternatively, Markov models provide an effective solution by capturing temporal learning dynamics for prediction, uncovering distinctive learning patterns between high and low performers. Despite these advantages, Markov model classification struggles with the heterogeneity of learning sequences, limiting its broad applicability. To address these limitations and bridge the gap between the two dominant approaches, we propose a hybrid framework: sequence-based Markov machine learning classification (seqMAC). Leveraging early-stage clickstream data, seqMAC provides an interpretable sequence classification method that captures critical behavioural transitions and identifies distinct learning patterns across performance groups. Tested on six LMS samples, seqMAC effectively identified at-risk students despite sequence heterogeneity, uncovering key predictive learning dynamics that differentiate performance groups. It also demonstrated promising generalizability, accurately identifying future at-risk students based on historical clickstream data.

Objective: Alcohol use offers social benefits for young adults, but also carries risk of significant negative consequences. Better understanding of processes driving alcohol use for those who experience negative consequences can prevent these harms. These at-risk young adults likely have drinking patterns in common and patterns unique to each individual. Evidence for these processes have been limited by methods that fail to capture the complex, heterogeneous, multivariate nature of drinking. We overcome these limitations with idiographic computational models. Method: We studied a sample of 97 young adults who regularly binge drink and experience negative drinking consequences. Participants completed daily surveys for 120 days. We estimated temporal networks of each person's drinking patterns by searching all possible dynamic relations among self-reported alcohol consumption and various cognitive, motivational, and emotional constructs. This method allowed us to identify common and uncommon drinking processes in a data-driven manner. Results: We found clear patterns of drinking characteristic of this population (i.e., shared by 60%-100% of the sample) in which young adults drink more per occasion, when they expect positive outcomes and are motivated to get drunk and enhance social experiences, which leads to positive and negative consequences. We also identified subsets of participants with uncommon (i.e., shared by <51% of the sample) drinking patterns. Conclusions: Most young adults may continue to drink despite experiencing negative drinking consequences, because it also satisfies their desire for fun and social connection. Additionally, subsets of young adults have relatively uncommon drinking patterns that may reflect risk or resilience factors.

Much of cognitive neuroscience research is focused on group-averages and interindividual brain-behavior associations. However, many theories core to the goal of cognitive neuroscience, such as hypothesized neural mechanisms for a behavior, are inherently based on intraindividual processes. To accommodate this mismatch between study design and theory, research frequently relies on an implicit assumption that group-level, between-person inferences extend to individual-level, within-person processes. The assumption of group-to-individual generalizability, formally referred to as ergodicity, requires that a process be both homogenous within a population and stationary within individuals over time. Our goal in this review is to assess this assumption and provide an accessible introduction to idiographic science (study of the individual) for the cognitive neuroscientist, ultimately laying a foundation for increased focus on the study of intraindividual processes. We first review the history of idiographic science in psychology to connect this longstanding literature with recent individual-level research goals in cognitive neuroscience. We then consider two requirements of group-to-individual generalizability, pattern homogeneity and stationarity, and suggest that most processes in cognitive neuroscience do not meet these assumptions. Consequently, interindividual findings are inappropriate for the intraindividual inferences that many theories are based on. To address this challenge, we suggest precision imaging as an ideal path forward for intraindividual study and present a research framework for complementary interindividual and intraindividual study.

Machine learning (ML) has extended the scope of psychological research by enabling data-driven discovery of patterns in complex datasets, complementing traditional hypothesis-driven approaches and enriching individual-level prediction. As a principal subfield, supervised ML has advanced mental health diagnostics and behavior prediction through classification and regression tasks. However, the complexity of ML methodologies and the absence of established norms and standardized pipelines often limit its adoption among psychologists. Furthermore, the black-box nature of advanced ML algorithms obscures how decisions are made, making it difficult to identify the most influential variables. Automated ML (AutoML) addresses these challenges by automating key steps such as model selection and hyperparameter optimization, while enhancing interpretability through explainable artificial intelligence (XAI). By streamlining workflows and improving efficiency, AutoML empowers users of all technical levels to implement advanced ML methods effectively. Despite its transformative potential, AutoML remains underutilized in psychological research, with no dedicated educational material available. This tutorial aims to bridge the gap by introducing AutoML to psychologists. We cover advanced AutoML methods, including combined algorithm selection and hyperparameter optimization (CASH), stacked ensemble generalization, and XAI. The utility of AutoML is demonstrated using the “H2O AutoML” R package with publicly available psychological datasets, performing regression on multi-individual cross-sectional data and classification on single-individual time-series data. We also provide practical workarounds for ML methods not currently supported in the package, so researchers can adopt alternative solutions when needed. These examples illustrate how AutoML democratizes ML, making it more accessible while providing advanced methodologies for psychological research.

Idiographic measurement models such as p-technique and dynamic factor analysis (DFA) assess latent constructs at the individual level. These person-specific methods may provide more accurate models than models obtained from aggregated data when individuals are heterogeneous in their processes. Developing clustering methods for the grouping of individuals with similar measurement models would enable researchers to identify if measurement model subtypes exist across individuals as well as assess if the different models correspond to the same latent concept or not. In this paper, methods for clustering individuals based on similarity in measurement model loadings obtained from time series data are proposed. We review literature on idiographic factor modeling and measurement invariance, as well as clustering for time series analysis. Through two studies, we explore the utility and effectiveness of these measures. In Study 1, a simulation study is conducted, demonstrating the recovery of groups generated to have differing factor loadings using the proposed clustering method. In Study 2, an extension of Study 1 to DFA is presented with a simulation study. Overall, we found good recovery of simulated clusters and provide an example demonstrating the method with empirical data.

We present the R package MIIVefa, designed to implement the MIIV-EFA algorithm. This algorithm explores and identifies the underlying factor structure within a set of variables. The resulting model is not a typical exploratory factor analysis (EFA) model because some loadings are fixed to zero and it allows users to include hypothesized correlated errors such as might occur with longitudinal data. As such, it resembles a confirmatory factor analysis (CFA) model. But, unlike CFA, the MIIV-EFA algorithm determines the number of factors and the items that load on these factors directly from the data. We provide both simulation and empirical examples to illustrate the application of MIIVefa and discuss its benefits and limitations.

Objective: Behavioral Activation (BA) is a brief intervention for depression encouraging gradual and systematic re-engagement with rewarding activities and behaviors. Given this treatment focus, BA may be particularly beneficial for adolescents with prominent anhedonia, a predictor of poor treatment response and common residual symptom. We applied group iterative multiple model estimation (GIMME) to ecological momentary assessment (EMA) treatment data to investigate common and person-specific processes of change during BA for anhedonic adolescents.Method: Thirty-nine adolescents (Mage = 15.7 years old, 67% female, 81% White) with elevated anhedonia (Snaith Hamilton Pleasure Scale) were enrolled in a 12-week BA trial, with weekly anhedonia assessments. EMA surveys were triggered every other week (2-3 surveys per day) throughout treatment assessing current positive affect (PA) and negative affect (NA), engagement in pleasurable activities and social interactions, anticipatory pleasure, rumination, and recent pleasurable and stressful experiences. Results: Multilevel models revealed significant decreases in anhedonia (t(25.5) = -4.65, p < 0.001) over the 12-week trial. GIMME results indicated substantial heterogeneity in symptom networks across patients. Within-patient change in PA was the variable with the greatest number (21% of all paths vs. 11% for NA) of predictive paths to other symptoms (i.e., out-degree). Higher PA (but not NA) out-degree was associated with greater anhedonia improvement (t(25.8) = -2.22, p = 0.035).Conclusions: Results revealed substantial heterogeneity in symptom relations across patients, which may obscure the search for common processes of change in BA. PA may be a particularly important treatment target for anhedonic adolescents in BA.

The efficacy of well-designed instructional videos for STEM learning is largely reliant on how actively students cognitively engage with them. Students' ability to actively engage with videos likely depends upon individual characteristics like their prior knowledge. In this study, we investigated how digital trace data could be used as indicators of students' cognitive engagement with instructional videos, how such engagement predicted learning, and how prior knowledge moderated that relationship. One hundred twenty-eight biology undergraduate students learned with a series of instructional videos and took a biology unit exam one week later. We conducted sequence mining on the digital events of students' video-watching behaviors to capture the most commonly occurring sequences. Twenty-six sequences emerged and were aggregated into four groups indicative of cognitive engagement: repeated scrubbing, speed watching, extended scrubbing, and rewinding. Results indicated more active engagement via speed watching and rewinding behaviors positively predicted unit exam scores, but only for students with lower prior knowledge. These findings suggest that the ways students cognitively engage with videos predict how they will learn from them, that these relations are dependent upon their prior knowledge, and that researchers can measure students’ cognitive engagement with instructional videos via mining digital log data. This research emphasizes the importance of active cognitive engagement with video interface tools and the need for students to accurately calibrate their learning behaviors in relation to their prior knowledge when learning from videos.

To develop effective and personalized interventions, it is essential to identify the most critical processes or psychological drivers that impact an individual’s well-being. Some processes may be universally beneficial to well-being across many contexts and people, while others may only be beneficial to certain individuals in specific contexts. We conducted three intensive daily diary studies, each with more than 50 within-person measurement occasions, across three data sets (n1 = 44; n2 = 37; n3 = 141). We aimed to investigate individual differences in the strength of within-person associations between three distinct process measures and a variety of outcomes. We utilized a unique idiographic algorithm, known as i-ARIMAX (Autoregressive Integrated Moving Average), to determine the strength of the relationship (Beta) between each process and outcome within individuals (“i”). All of the computed betas were then subjected to meta-analyses, with individuals treated as the “study”. The results revealed that the process-outcome links varied significantly between individuals, surpassing the homogeneity typically seen in meta-analyses of studies. Although several processes showed group-level effects, no process was found to be universally beneficial when considered individually. For instance, processes involving social behavior, like being assertive, did not demonstrate any group-level links to loneliness but still had significant individual-level effects that varied from positive to negative. Using i-ARIMAX might help reduce the number of candidate variables for complex within-person analyses. Additionally, the size and pattern of i-ARIMAX betas could prove useful in guiding personalized interventions.

Spearman (Am J Psychol 15(1):201–293, 1904. https://doi.org/10.2307/1412107 ) marks the birth of factor analysis. Many articles and books have extended his landmark paper in permitting multiple factors and determining the number of factors, developing ideas about simple structure and factor rotation, and distinguishing between confirmatory and exploratory factor analysis (CFA and EFA). We propose a new model implied instrumental variable (MIIV) approach to EFA that allows intercepts for the measurement equations, correlated common factors, correlated errors, standard errors of factor loadings and measurement intercepts, overidentification tests of equations, and a procedure for determining the number of factors. We also permit simpler structures by removing nonsignificant loadings. Simulations of factor analysis models with and without cross-loadings demonstrate the impressive performance of the MIIV-EFA procedure in recovering the correct number of factors and in recovering the primary and secondary loadings. For example, in nearly all replications MIIV-EFA finds the correct number of factors when N is 100 or more. Even the primary and secondary loadings of the most complex models were recovered when the sample sizes were at least 500. We discuss limitations and future research areas. Two appendices describe alternative MIIV-EFA algorithms and the sensitivity of the algorithm to cross-loadings.

Even highly motivated undergraduates drift off their STEM career pathways. In large introductory STEM classes, instructors struggle to identify and support these students. To address these issues, we developed co-redesign methods in partnership with disciplinary experts to create high-structure STEM courses that better support students and produce informative digital event data. To those data, we applied theory- and context-relevant labels to reflect active and self-regulated learning processes involving LMS-hosted course materials, formative assessments, and help-seeking tools. We illustrate the predictive benefits of this process across two cycles of model creation and reapplication. In cycle 1, we used theory-relevant features from 3 weeks of data to inform a prediction model that accurately identified struggling students and sustained its accuracy when reapplied in future semesters. In cycle 2, we refit a model with temporally contextualized features that achieved superior accuracy using data from just two class meetings. This modelling approach can produce durable learning analytics solutions that afford scaled and sustained prediction and intervention opportunities that involve explainable artificial intelligence products. Those same products that inform prediction can also guide intervention approaches and inform future instructional design and delivery.Practitioner notesWhat is already known about this topicWhat this paper addsImplications for practice and/or policy Learning analytics includes an evolving collection of methods for tracing and understanding student learning through their engagements with learning technologies. Prediction models based on demographic data can perpetuate systemic biases. Prediction models based on behavioural event data can produce accurate predictions of academic success, and validation efforts can enrich those data to reflect students' self-regulated learning processes within learning tasks. Learning analytics can be successfully applied to predict performance in an authentic postsecondary STEM context, and the use of context and theory as guides for feature engineering can ensure sustained predictive accuracy upon reapplication. The consistent types of learning resources and cyclical nature of their provisioning from lesson to lesson are hallmarks of high-structure active learning designs that are known to benefit learners. These designs also provide opportunities for observing and modelling contextually grounded, theory-aligned and temporally positioned learning events that informed prediction models that accurately classified students upon initial and later reapplications in subsequent semesters. Co-design relationships where researchers and instructors work together toward pedagogical implementation and course instrumentation are essential to developing unique insights for feature engineering and producing explainable artificial intelligence approaches to predictive modelling. High-structure course designs can scaffold student engagement with course materials to make learning more effective and products of feature engineering more explainable. Learning analytics initiatives can avoid perpetuation of systemic biases when methods prioritize theory-informed behavioural data that reflect learning processes, sensitivity to instructional context and development of explainable predictors of success rather than relying on students' demographic characteristics as predictors. Prioritizing behaviours as predictors improves explainability in ways that can inform the redesign of courses and design of learning supports, which further informs the refinement of learning theories and their applications.