Friday 24 July 11:25
- 12:55
Hall: 16 - Room 13 SA
Chair and Discussant:
Zumbo Bruno
Division: Division 2: Psychological Assessment and Evaluation
This symposium expands the conversation in psychometrics and assessment by bringing
together international leaders whose research demonstrates how innovation—including
theoretical and conceptual analysis, mixed-methods validation, computational modeling, and AI
integration—can help transform psychological measurement into a more responsive,
contextualized, and dynamic science. Rather than promoting unidirectional discourse, the session
fosters dialogue among researchers and practitioners on how measurement science can better
reflect the complexity of human experience. The focus is on advancing approaches that move
beyond static, in vitro models of psychological traits and constructs.
This symposium presents four papers that challenge static, decontextualized models of
psychological measurement and, in response, promote an in vivo, process-oriented perspective,capturing situated, dynamic, and context-dependent phenomena. Each presentation describes
innovative approaches to capturing the complex interplay of person, context, and process that
characterizes psychological phenomena.
The first paper examines the often-overlooked category of situational constructs, highlighting the
mismatch between conceptual definitions rooted in context and measurement practices that
ignore temporal and situational variability. It calls for psychometric theories and practices that
better reflect the dynamic nature of such constructs.
The second paper addresses response process validation by integrating traditional cognitive
interviews with psychophysiological measures (e.g., eye-tracking, electrodermal activity) and
psycholinguistic item analysis. The study demonstrates how converging sources of qualitative
and quantitative data enrich our understanding of how respondents interpret and engage with test
items.Thethirdpapersituatesmeasurementwithin dynamicsystems theory,emphasizing the
limitationsofstatictraitmodelsandthepromiseofintensivelongitudinaldata(e.g., EMA/ESM),
multilevel and time-series modeling, and idiographic trajectories.Italsoraisesphilosophicaland
methodological challenges associated with shiftingfromtraitstoprocesses.
The fourth paper explores the role of generative AI in response process validation via web
probing. It presents an AI-based solution for coding and analyzing open-ended responses,
comparing multiple generative models. This approach aims to enhance both efficiency and
validity in online assessments by enabling automated, context-sensitive interpretation of textual
data.Thediscussantwillprovideintegrativecommentary andengagetheaudienceinaforward-looking
dialogueaboutreimaginingtestsandmeasuresforasituationaldynamicpsychologicalscience.
