My current work examines three related areas.
Response Latency in Video-Based Surveys
Focus: Response latency and item- and respondent-related characteristics in video-based noncognitive surveys.
Research questions: How do item length, response format, semantic sufficiency, and respondent characteristics relate to response timing?
Methods: Process-data analysis, multilevel models, item-level indicators, and exploratory visualizations.
Semantic Sufficiency Thresholds
Focus: Identifying the earliest point in a spoken survey item at which sufficient meaning is available to support a response.
Research questions: How consistently do human and AI raters identify sufficient meaning? What implications might this have for identifying premature responses?
Methods: Rater agreement, threshold ratings, and timing-based indicators.
Response Processes in Digital Assessments
Focus: Understanding timing and action patterns in computer-administered tasks.
Research questions: What can time to first action, total response time, and post-first-action timing reveal about task interaction?
Methods: Descriptive process analytics and regression modeling.