Soo Youn Kim, Paul A Pilkonis, Lan Yu, Carla A Mazefsky
The successful linking of the EDI and EDI-YC supports the interpretation of raw scores on a common metric. These findings demonstrate the feasibility of IRT-based linking as a practical tool for longitudinal research and clinical monitoring of ED, pending further validation of anchor item invariance across the full age range.
OBJECTIVE: The Emotion Dysregulation Inventory (EDI) and its young-child version (EDI-YC) are widely used informant-report measures for assessing emotion dysregulation (ED). However, scores are not directly comparable due to differing age ranges (EDI: 6-17 years; EDI-YC: 2-5 years), limiting longitudinal research. This study used item response theory (IRT) to link EDI and EDI-YC scores, creating a common metric spanning early childhood through adolescence in general and neurodevelopmental populations.
METHOD: Archival data from original calibration and validation studies were analyzed. The EDI sample included 1,755 clinical participants (Age 4-20, 21% female) and 1,000 general population youth (Age 6-17, 49% female). The EDI-YC sample included 1,369 clinical participants (Age 2-5, 27% female) and 768 general population youth (Age 2-5, 51% female). A non-equivalent groups with anchor test (NEAT) design was implemented. Both measures assess Reactivity and Dysphoria. Separate unidimensional IRT co-calibrations were conducted for each subscale using a fixed-parameter approach.
RESULTS: Co-calibration revealed near-perfect correlations between EDI and EDI-YC metrics in both the general population (Reactivity ρ = .999; Dysphoria ρ = .995) and clinical samples (Reactivity ρ = .999; Dysphoria ρ = .997).
CONCLUSION: The successful linking of the EDI and EDI-YC supports the interpretation of raw scores on a common metric. These findings demonstrate the feasibility of IRT-based linking as a practical tool for longitudinal research and clinical monitoring of ED, pending further validation of anchor item invariance across the full age range.