Cross-Informant Reliability & Validity of Autism Screening Using the First Year Inventory in Israel

Background: Screening for early signs of autism spectrum disorders (ASD) poses both an empirical and clinical challenge. The First Year Inventory (FYI) offers a norm-referenced risk scoring that takes into consideration typical variations in early development of social and regulatory behaviors at 12-months, however requires investigation of its generalizability to other countries. Most previous screening evidence stems from maternal report. The increasing involvement of fathers in child rearing calls for understanding of both parents report of their child's development.  

Objectives: (1) Examine cross-informant reliability in reporting early ASD markers; (2) Test the construct validity of the FYI relative to the Infant Toddler Social Emotional Assessment (ITSEA).

Methods: A sample of 160 parents (82 mothers, Mage= 32 years, 78 fathers, Mage= 34 years) of infants ages 11-13 months (55% boys) completed the FYI. Mothers of 51% of infants were full-time workers and 33% of infants were cared for at home by a family member. Mothers also completed selected scales from the ITSEA that correspond with FYI items in the sensory-regulatory and social domains. Infants with medical or developmental concerns were excluded.  

Results: The FYI showed high internal consistency among mothers (α=.76) and fathers (α=.73), but there was low cross-informant reliability (ICC=0.47). Repeated measures MANCOVA was used to compare mean FYI social and sensory scores between mothers and fathers controlling for the mother’s employment status and the infant’s care setting. There was a significant informant effect (F(2, 71)=5.54, p=.006), with fathers reporting significantly higher mean FYI social scores than mothers (F(1, 72)=11.08, p=.001). There were no significant interactions between informant and covariate effects. There was a significant effect for mother’s employment status upon mean FYI sensory scores, with full-time working mothers reporting significantly higher frequencies of regulatory behaviors than at-home mothers. The distribution of FYI Total risk scores differed between parents, with mothers showing a 95^th percentile cutoff of 17.39 which is similar to the original US FYI sample cutoff while fathers had a higher risk cutoff of 22.25. On average fathers reported an FYI Total risk score, which was 3.3 points higher than mothers.

Testing construct validity indicated that the mean FYI social score was significantly correlated with the ITSEA social score (r=-.54, p<.001, negative due to the inverse direction of scores). The mean FYI sensory-regulatory score was significantly correlated with the ITSEA dysregulation score (r=0.53, p<.001). There were no significant correlations between social and dysregulation scores across measures. 

Conclusions: The FYI shows good levels of reliability and validity when implemented in Israel. Differences between parents may relate to the nature of fathers' involvement in caring for their infants and to their unique perspective on typical social milestones. Findings call for careful interpretation of elevated social risk scores in fathers’ report and for the need to collect cross-informant screening information.