Keeping an Eye on Looking Measures: Towards More Robust Developmental Methods

Area of Science:

• Developmental Psychology
• Cognitive Science
• Psychometrics

Background:

• Experimental developmental psychology faces challenges in selecting optimal outcome measures for behavioral studies.
• The looking-while-listening paradigm, used for early word comprehension, has over 12 distinct outcome measures lacking empirical validation.
• Limited psychometric information hinders the selection of the most effective measures in developmental research.

Purpose of the Study:

• To evaluate the psychometric properties (validity, robustness, reliability) of 12 outcome measures in the looking-while-listening paradigm.
• To identify the most reliable and valid measures for assessing early word comprehension.
• To determine if the looking-while-listening paradigm captures distinct cognitive processes and assess developmental sensitivity.

Main Methods:

• Archival data from 15 datasets comprising 602 children aged 12-60 months were analyzed.
• Psychometric properties of 12 distinct outcome measures were evaluated.
• Comparison of developmental sensitivity between distractor-initial and target-initial trials was conducted.

Main Results:

• Proportion looking, reaction time, and proportion of trials switching from distractor to target demonstrated the strongest validity, robustness, and reliability.
• The looking-while-listening paradigm effectively captures two distinct cognitive processes: detecting mismatches and confirming matches.
• Distractor-initial trials exhibited greater developmental sensitivity compared to target-initial trials.

Conclusions:

• Proportion looking, reaction time, and switching trials are recommended as the most psychometrically sound measures for the looking-while-listening task.
• The paradigm differentiates between detecting mismatches and confirming matches, with distractor-initial trials being more developmentally sensitive.
• This study provides a framework for evaluating measurement approaches in experimental developmental science.