Psychological Assessment

doi: 10.1037/pas0001424pmid: 41129391

Intensive longitudinal designs are being used with increased frequency in clinical psychology research and assessment, due to their ability to assess within-person, dynamic processes in naturalistic contexts in near real time. However, the complexity inherent in these designs means there is a need for more empirical data to guide decision making for specific research or clinical practice applications. As such, this special issue presents 15 studies (published across two associated journal issues) with innovative findings and methods that can guide clinical psychology researchers and practitioners as they design, conduct, interpret, and analyze intensive longitudinal designs. The articles cover topics including considerations for power/sample size planning and predicting attrition; practices for optimal sampling designs and methods; insights from participants’ experiences in intensive longitudinal design studies; and poststudy procedures regarding assessment of data quality, psychometrics, and conducting analyses. Each study is briefly reviewed, and implications for clinical research and practice are discussed.

doi: 10.1037/pas0001408pmid: 41129392

Ambulatory assessment is popular in research settings for its ability to assess real-world functioning. It is useful for estimating an individual’s typical level of a behavior (individual mean), how (un)stable that behavior is (individual standard deviation), how behaviors associate with others or specific contexts (within-person correlation), and shifts in those statistics that might signal an important change in functioning (e.g., early warning signal). However, many of the methodological advances have not made the jump from the lab to clinical practice. Effective use of ambulatory assessment in applied settings to understand functioning and guide potential interventions requires development and application of psychometric standards for N = 1 assessments. We conducted a simulation study to determine how many assessments are necessary to achieve sufficiently reliable (i.e., precise and stable) estimates of an individual’s mean and standard deviation on a single variable as well as the correlation between two variables. To ensure the ecological validity of the simulation conditions, we used real time series data from a large sample that included psychiatric patients and nonpatients (capturing realistic levels of autocorrelation and skewness). We found that the minimum number of assessments depends on the statistic of interest and the temporal characteristics of the variable of interest. Individual means can be estimated reliably with a reasonably small number of observations under most conditions, but adequately precise and stable individual correlations require more assessments than may be achievable in many applied settings. Implications of these results for the potential of applied ambulatory assessment in clinical practice are discussed.

doi: 10.1037/pas0001409pmid: 41129393

The experience sampling method (ESM) plays a pivotal role in investigating the dynamics of psychopathological processes in daily life. A crucial question when designing ESM studies concerns the sample size needed, defined by the number of participants (N) and the number of measurement occasions per participant (T). Higher N and T increase power, but also increase researcher and participant burden, and study cost. Current approaches for sample size planning rarely account for these feasibility and financial constraints explicitly, despite significant variations in ESM studies’ design, operational expenses, participant incentives, and compliance rates. This oversight can lead to suboptimal or unrealistic sample size planning. In this article, we extend the traditional power analysis framework to integrate budget constraints into sample size decisions. In particular, we demonstrate how to formalize budget considerations into cost functions for ESM studies and how to use these to optimally select N and T values. Through an illustrative example, we showcase how optimal sample size decisions strongly differ across ESM designs and associated cost functions, even when focusing on the same research questions.

Petersen, Fridtjof; Bringmann, Laura F.; Rizopoulos, Dimitris

doi: 10.1037/pas0001397pmid: 41129394

Intensive longitudinal data from ecological momentary assessment (EMA) are widely used in clinical research but often suffer from dropout, leading to reduced statistical power, invalid results, and poor treatment outcomes. Predicting dropout could help with its prevention. While existing methods utilize baseline covariates, few studies account for the temporal dynamics of EMA data or identify the exact timing of dropout. Joint models (JM) enable simultaneous modeling of longitudinal processes and time-to-event data, offering dynamic predictions. However, conventional JMs assume limited measurement occasions and do not account for the autocorrelation inherent in EMA data. We extended the standard JM by incorporating an autoregressive submodel, capturing temporal dependencies in EMA measurements. We validated our approach through simulation studies, demonstrating good parameter recovery across different missingness mechanisms (missing completely at random, missing at random, missing not at random) and high dropout prediction accuracy. We applied the JM to an existing empirical EMA data set, using baseline (e.g., depression) and time-varying (affect, intermittent missingness) predictors of dropout. The extended JM outperformed a baseline-only survival model in predicting dropout. The sensitivity analysis of the missingness mechanism revealed that fixed-effect estimates remained stable across different missing data mechanisms, whereas random-effect estimates for autocorrelation were sensitive to these assumptions. By integrating autoregressive components, the extended JM accommodates temporal dependencies and dynamically updates predictions of dropout risk. This approach improves dropout prediction in EMA studies and highlights the importance of utilizing JMs for predicting clinically relevant outcomes while integrating EMA data.

doi: 10.1037/pas0001394pmid: 41129395

The future of technology-mediated just-in-time interventions requires detecting moments when skills would be most useful. For example, affect regulation skills could be provided during emotional episodes. But how do researchers (and clinicians) operationalize “emotional” episodes? In this study, we use secondary data from an 8-day ecological momentary assessment study (n = 197) where participants rated emotional adjectives of positive (e.g., joyful, calm, relaxed) and negative (e.g., sad, angry, anxious) feelings on a 0–100 scale and a categorical subjective determination of emotion 5×/day. We compared three different ways of classifying whether a moment was “better” (i.e., more positive), “worse” (i.e., more distressing), or typical/as usual affect: (Option A) elevated level of affect for positive and/or negative affect (e.g., whether the rating was high or low on the scale itself), (Option B) a 17-point deviation from the person’s own average on positive and a 12-point deviation for negative affect, and (Option C) the participant’s own categorical determination of better, same, or worse. Results revealed that affect level (Option A) and the participant’s own subjective determination (Option C) resulted in more moments classified as emotional than person-centered deviations, especially person-centered deviations on negative affect. In validating all classification methods, we found that “worse” was associated with more problems (e.g., lower thought clarity and willpower, greater experiential avoidance and rash action urges) than “affect as usual” using all options. We discuss implications for how researchers and clinicians can use technology to find “emotional” moments in future studies, with the aim of guidance toward just-in-time momentary interventions.

doi: 10.1037/pas0001404pmid: 41129396

Experience sampling methodology has been widely used to study the links between emotion dynamics and mental health. Most studies rely on time-contingent sampling schemes, with momentary questionnaires being sent at prespecified times, usually multiple hours apart. The present study investigated the added value of episode-contingent (EC) burst designs, which may shed further light on emotion dynamics by triggering a series of closely timed beeps upon detecting emotional episodes. Using data from three EC studies (N = 185), we investigated the effectiveness and feasibility of two types of EC designs: signal-based (EC-Signal; bursts initiated when emotion ratings exceed thresholds) and event-based (EC-Event; bursts initiated by participants). Both EC designs are effective in capturing emotional episodes, but the quantity and intensity of the episodes varied depending on which of these two designs was used and the valence of the episodes. Regarding feasibility, compared to EC-Event, EC-Signal typically led to higher participant burden and lower compliance for both regular and follow-up beeps. Moreover, compliance tended to decrease over time and burden tended to increase for both EC-Signal and time-contingent, but not for EC-Event. In conclusion, both EC approaches showed feasibility but have distinct advantages and drawbacks. To select the best approach, researchers should carefully balance these trade-offs to maximize utility within specific research contexts.

doi: 10.1037/pas0001400pmid: 41129397

To advance the design and use of intensive longitudinal methods in investigations of adolescent depression, we conducted a multimethod and multi-informant study of daily parent–youth interactions, specifically, supportive communication, consisting of (a) naturalistic video observations of parent–youth interactions; (b) passive collection of Bluetooth Low Energy signals to approximate parent–youth proximity; and (c) scheduled, proximity-contingent and self-initiated ecological momentary assessments (EMA). We examined whether these novel and complementary approaches enhanced the assessment of parent–youth interactions, a key source of risk and protection for youth mental health. Specifically, we report participant compliance on the video recording procedures and describe preliminary results from our observational coding of supportive communication. We also report compliance rates on EMAs and examine the frequencies of parent–youth interactions per self-report and Bluetooth Low Energy signals. Participants in the 2-week-long protocol were 12- to 15-year-old adolescents (N = 138; 63.8% female, 42% Center for Epidemiologic Studies Depression ≥ 16) and their parents (95.7% biological mothers, 25% Center for Epidemiologic Studies Depression ≥ 16). Dyads completed mean 122.6 min (SD = 85.6) of video recordings. In 387 min of recordings from three pilot families, we identified 52 supportive communication episodes. The average parent and youth were compliant with EMA procedures, completing the recommended minimum of 40 cumulative surveys each. Parents and youth reported that they interacted with the other member in mean 56%–83% of the EMAs. The study demonstrates innovative ways to leverage technology to conduct multimethod and multi-informant intensive longitudinal assessments of interpersonal interactions, a key source of risk and protection for adolescent mental health.

doi: 10.1037/pas0001391pmid: 41129398

Successful personalized treatment requires a thorough understanding of the complex dynamic processes underlying disorders. Intensive longitudinal methods (e.g., experience sampling) that ask patients to complete multiple-item questionnaires several times a day are ideally suited for this. However, collecting such data entails severe patient burden, especially for those with low energy and little concentration (e.g., patients suffering from chronic cancer-related fatigue and/or psychological disorders such as somatic symptom disorder). This burden is currently predominantly lightened with single-item measures, but these cannot validly capture complex conditions, leading to a catch-22 situation: Capturing complex dynamic processes and effective personalized treatment require intensive longitudinal patient data on multiple-item questionnaires, but patients cannot provide this type of data because it is too taxing. To solve this problem, we developed a personalized missingness design that presents an individualized and time-varying minimal subset of items on each occasion, thereby striking an optimal balance between thoroughly mapping patients’ symptoms and keeping the number of items a person needs to answer to a minimum. The design builds on multilevel factor analyses to determine which sets of items are most informative, which can change over time. Expert-informed simulations validated our new design. While the design can be universally applied to any measurement of (psychological) symptoms (e.g., to inform cognitive behavioral therapy), we tailored our simulations to patients suffering from chronic cancer-related fatigue in collaboration with experts in psycho-oncology. In the near future, the design will be implemented in the widely used experience sampling app m-Path in collaboration with the developers.