An Empirical Examination of the Structure-Process-Outcomes Model Within Head Start Programs

• In Event: Early Care and Education Quality Indicators, Implementation, and Policy

Abstract

Introduction
ECE research has focused on two quality domains, with theory suggesting that structural features of quality predict process quality, which in turn, predicts children’s outcomes (NICHD ECCRN, 2002). The structure-process-outcomes theory has been widely accepted within the field, which has led to the development of ECE policies regulating both structural and process quality.

Hypotheses
Our study seeks to empirically test the structure-process-outcomes model. We test this overall theory in the Head Start context by examining whether quality thresholds as required within Head Start’s Program Performance Standards are predicted by structural features of quality and predictive of children’s outcomes.

We hypothesize that structural features of quality will predict higher levels of process quality, and there will be positive associations, albeit small to moderate, between indicators of process quality and children’s outcomes in general and within the Head Start quality thresholds context.

Study population
Our study relied on data from the Head Start Family and Child Experiences Survey (FACES) 2014 study. The FACES study collected nationally representative data on a sample of Head Start programs, centers, classrooms, and 3- and 4-year old children. A total of 176 programs, 667 classrooms, and 2,462 children were sampled in the FACES 2014 study.

Methods
We conducted OLS regression analyses to examine the association between structural quality, process quality, and child outcomes. Structural quality variables included class size, child-teacher ratio, days per week class meets, teacher education level, teacher field of highest degree, teacher has child development associate credential, and total professional development supports offered. Process quality was measured using the ECERS-R (Harms et al., 2005) and the Pre-K CLASS (Pianta et al., 2008).

For our study, we included the following measures of children’s academic and behavioral development: Peabody Picture Vocabulary Test-Fourth Edition (PPVT-4; Dunn et al., 2006), Expressive One-Word Picture Vocabulary Test-4 (EOWPVT-4; Martin & Brownell, 2010), Woodcock Johnson III Tests of Achievement (WJ-III; Woodcock et al., 2001) Letter-Word ID, Applied Problems, and Spelling subtests, teacher-reported Social Skills Rating System (SSRS; Gresham & Elliott, 1990), and teacher-reported behavior problems.

To examine the competitive CLASS thresholds identified by Head Start program guidelines, we transformed the CLASS variables to become dichotomous, where 1 = above the competitive threshold. The competitive threshold for the Emotional Support and Organization Score domains is 5, and the competitive threshold for the Instructional Support domain is 2.3.

Results
Descriptive statistics are presented in Table 1. We find little evidence to support the structure-process-child outcomes model as defined by the current ECE policy landscape. One structural indicator of quality, days per week class meets, was significantly, positively associated (p < .05) with the ECERS-R. We also found CLASS Instructional Support was significantly, positively associated (p < .05) with children’s Applied Problems score (see Table 2).

We find largely null results for structural features of quality predicting CLASS thresholds. Days per week the class meets was significantly, positively associated (p < .01) with CLASS Emotional Support Threshold scores. We did not find any evidence that CLASS thresholds predicted children’s academic or behavioral outcomes.

Authors

• Cara Lynn Kelly, University of Delaware
  Presenting Author

• Ana Whitaker, University of Delaware

• Margaret Burchinal, University of Virginia - Charlottesville