Premium
Validation of the Early Development Instrument for children with special health needs
Author(s) -
Janus Magdalena,
Zeraatkar Dena,
Duku Eric,
Bennett Teresa
Publication year - 2019
Publication title -
journal of paediatrics and child health
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.631
H-Index - 76
eISSN - 1440-1754
pISSN - 1034-4810
DOI - 10.1111/jpc.14264
Subject(s) - medicine , confirmatory factor analysis , population , construct validity , special needs , consistency (knowledge bases) , construct (python library) , psychometrics , goodness of fit , clinical psychology , structural equation modeling , environmental health , statistics , psychiatry , computer science , programming language , artificial intelligence , mathematics
Aim Population‐level data on the development of children with special health needs attending primary school can assist special education, health and service delivery strategies. This study aims to determine whether results of the Early Development Instrument (EDI) can be interpreted and used with the same confidence for children with special health needs as for those who are typically developing. Psychometric properties of the EDI for kindergarteners with identified special health needs are examined. Methods Data for children with a special health need designation come from a database of all Canadian provincial implementations of the EDI between 2004 and 2014. Item and domain characteristics (distributions of scores and measures of internal consistency) of the EDI were examined. Construct validation was assessed by correlating EDI scores with gender and age. The fit of the original factor structure of the EDI was evaluated using confirmatory factor analysis. Results Data for 29 692 children (69.8% male) were analysed. The performance of items and domains was similar to that of typically developing children. As expected, boys scored lower on all EDI domains (Cohen's d : 0.03–0.46) and a general, but non‐significant, positive trend was observed between age and EDI scores. The factor structure of the EDI in this population yielded similar goodness‐of‐fit statistics as those reported in studies with typically developing children. Conclusion Results of this investigation support the validity of the EDI in children with special health needs, paving the way for a more extensive use of EDI data for this vulnerable, yet often neglected, population.