Poster #49 - Convergent Validity Between the National Developmental Surveillance Guide and the Child Development Inventory.

• In Event: Poster Session 08

Abstract

Early identification and intervention of emerging developmental problems reduce the prevalence of disorders that may cause disability in the child's current or future life (Pallarés, 2007; Wright & Saul, 2013). Timely interventions in development can substantially improve evolution and bring opportunities to reduce inequality, and promote social and educational inclusion (Shonkoff et al., 2016). The design of an inter-agency early development evaluation system for 3 to 6-year-old children in Uruguay would allow integrated strategies for the early detection of children at risk for full development. Uruguay has two validated national instruments for the early detection of developmental risk for children 3 to 6 years of age, both public policies since 2018. On the one hand, pediatricians and family doctors apply the National Developmental Surveillance Guide (GNVD; MSP, 2014) in healthcare centers; on the other hand, preschool teachers administer the School Readiness-Child Development Inventory (INDI; Vásquez-Echeverría, 2022) in public schools. In this study, we aimed to establish the relationship between development dimensions and subscales of INDI and the GNVD, determine concordance between assessment outcomes, and identify adjustments or modifications to improve their complementarity.
Three reviewers found good levels of theoretical correspondence (i.e. 55% of the GNVD items at age 4 have conceptual correspondence in INDI items). According to the developmental domain in the GNVD, there is 40% correspondence for Language development, 50% for Gross Motor skills, 57% for Social development, and 70% for Coordination. According to INDI’s domains, the highest levels of correspondence are present for the Cognitive Development and Motor Development dimensions. At the empirical level, the sample consisted of 308 children (girls = 154) with an age range between 49 and 72 months (M = 62.6; SD = 5.9) from Year 4 and Year 5 classrooms within the public preschool. Results indicate sample profile differences similar to normative in terms of sex, socioeconomic quintile, and educational level, based on INDI data. Positive and statistically significant correlations were found (p < .01) in the range of .20 to .34 among areas of theoretical convergence expected between instruments. Likewise, levels of agreement between the outcome profiles are statistically significant and moderate in magnitude: 81.3% average agreement when both instruments provided the same outcome. On the other hand, we found higher levels of agreement between results when there are no developmental delays: 85% average based on INDI and 93.8% based on GNVD (see Table 1). Some disparities were found according to each assessment outcome when facing developmental difficulties. Inter-observer agreement was moderate and significant for the theoretical convergence domains (see Table 2). We conclude that further studies should address the discrepancies between ways of assessing and criteria for granting the "risk" profile within each instrument. Finally, some improvements should be considered, to contribute to an inter-agency early development evaluation system. This study's contributions have direct implications on the possibilities of response and timely intervention and, consequently, on the rational allocation of public resources.

Authors

• Maite Liz, Universidad de la República
  Presenting Author

• Mercedes Pérez