Authentic Assessment for Infants and Toddlers: Exploring the Reliability and Validity of the Ounce Scale

Abstract

This study used a mixed methods methodology to investigate the reliability and validity of the Ounce Scale, an authentic, observational assessment of infants' and toddlers' development from birth through 42 months of age. Quantitative cross-sectional data were collected from 287 children and 124 teachers in seven urban Early Head Start programs; qualitative data were derived from interviews with 21 teachers and seven supervisors. Data were collected across eight age groups. Results showed moderate reliability of the Ounce Scale and provided evidence of agreement with criterion measures for concurrent validity. Receiver operating characteristic curve (ROC) analyses demonstrated very good levels of accuracy in predicting which children were at-risk or not at-risk. Hierarchical regression analyses indicated that, after controlling for child and family variables, the Ounce Scale contributed significantly to explaining the variance in children's performance on the criterion measures. Analysis of qualitative interview data elaborates on these findings in terms of the strength-based philosophy of the caregivers, the binary structure of the scale, the cultural context in which the scale was used, and the need for additional professional development. Discussion also centers on the relationship between norm-referenced and performance-based assessments in early childhood.

Funding was provided for this study by the Head Start-University Partnerships for Measurement Development and Pearson Early Learning, the publisher of the Ounce Scale™, although the authors alone are responsible for the findings and interpretations presented herein.

We thank Ilona Helin and Lisa Michels for their assistance in data collection, as well as the teachers, administrators, parents, and children who participated in this study. We are also grateful to Sally Atkins-Burnett, Yange Xue, and Dot Marsden for their comments on an earlier version of this article.

Notes

¹See Meisels (1996b), from which this discussion is derived, for a more complete review of these types of assessments.

²The Bayley-III was not yet available when this study began.

Note: ASQ:SE has different versions across ages, hence scores for each age group are reported separately. MDI = Mental Development Index; PDI = Psychomotor Development Index; PLS-4 = Preschool Language Scale-4; ASQ:SE = Ages and Stages Questionnaire: Social-Emotional.

Note: The overall reliability of the developmental profile based on 11 shared items across ages is .65.

Note: Ounce developmental profile is represented by the number of needs development ratings.

MDI = Mental Development Index; PDI = Psychomotor Development Index; PLS-4 = Preschool Language Scale-4; ASQ:SE = Ages and Stages Questionnaire: Social-Emotional.

∗p < .05. ∗∗p < .01.

Note: Ounce developmental profile is represented by the number of needs development ratings.

MDI = Mental Development Index; PDI = Psychomotor Development Index; PLS-4 = Preschool Language Scale-4; ASQ:SE = Ages and Stages Questionnaire: Social-Emotional.

Low education = HS/CDA; less experience with infants/toddlers = <5 years; shorter duration of using Ounce = <1 year.

∗p < .05. ∗∗p < .01.

Note: ASQ: SE scores are not comparable across ages, hence they are not included in this analysis. MDI = Mental Development Index; PDI = Psychomotor Development Index; PLS-4 = Preschool Language Scale-4; ASQ:SE = Ages and Stages Questionnaire: Social-Emotional.

∗∗p < .01.

Note: MDI = Mental Development Index; PDI = Psychomotor Development Index; PLS-4 = Preschool Language Scale-4; ASQ:SE = Ages and Stages Questionnaire: Social-Emotional; Ounce = Ounce Scale.

Note: The number of needs development ratings is used to represent the developmental profile. MDI = Mental Development Index; PDI = Psychomotor Development Index; PLS-4 = Preschool Language Scale-4; ASQ:SE = Ages and Stages Questionnaire: Social-Emotional.

^a AR² = .05.

^b AR² = .07.

^c AR² = .06.

^d AR² = .18.

^e Child gender, 0 = boy.

^f child ethnicity, 0 = minority.

∗p < .05. ∗∗p < .01.