Poster #118 - Child Care Experiences and Reading Trajectories: What can we learn from multiple-group growth modeling?

• In Event: 3-106 - Poster Session 11
  In Poster Session: PS 11 Section - Methods, History, Theory

Integrative Statement

Formal child care experiences prior to kindergarten are known contributors to how prepared children are when they start kindergarten (Duncan & Magnuson, 2013). For example, children who had a formal care experience at a high-quality center-based program (v.s. no nonparental care or formal care with lower quality) tend to experience higher starting points and growth rates in various cognitive and non-cognitive skills at the onset of kindergarten and beyond (Peisner‐Feinberg et al., 2001). To empirically test this hypothesis with a cognitive outcome such as reading, researchers have often utilized a growth model that estimates the reading trajectories of all children and then introduces the time-invariant covariate (TIC) of interest (i.e., early care experience prior to kindergarten). Subsequently, they examine the extent to which this TIC explains whether prior experiences are predictive of higher or lower starting points in reading trajectories and/or steeper or flatter growth trajectories. This conditional growth model is typically fit as a single-group model where the underlying assumption is that all parameters defining this model are the same for the different groups being defined by the TIC (Curran, Obeidat, & Losardo, 2010; Grimm, Ram, Estabrook, 2016). If looking at reading trajectories, this will imply that those children with prior early care experiences and those with no nonparental care only differ in their conditional means of the growth factors (i.e., intercept and slope for reading trajectories) but not in the variances and covariances that describe these factors. Given what we know about the effects of formal center-based programs (Duncan et al., 2013) it is sensible to expect that center-based programs decrease the variability in either the intercept and/or slope of the reading trajectories. This in turn will potentially make the variance estimates distinctly different across those children with early care experiences and those with no nonparental care. The purpose of this methodological illustration is to empirically test the hypothesis of multiple reading growth trajectories on children with some formal child care experience prior to kindergarten versus those with no nonparental care experience. This particular question will be examined using a multiple-groups growth framework (McArdle, 1989; McArdle & Hamagami, 1996). Explicitly, a structural equation modeling approach (Jöreskog, 1971) will be implemented to simultaneously estimate the growth models across those children who had a formal care experience the year prior to kindergarten versus those children who did not. To illustrate this process, we will use data from the Early Childhood Longitudinal Study - Kindergarten (ECLS-K: 2011). Specifically, reading scores from 6 time points will be used (fall & spring K, spring G1 to G4) as well as an item that asked caregivers about the care arrangements the year before kindergarten. The two main groups of interest are those children with a formal center-based program experience (n = 5, 751) versus those with no nonparental care (n = 2,080). Results from this analysis will inform modeling strategies for researchers interested in the longitudinal association of preschool experience on various outcomes.

Author

• Gloria Yeomans-Maldonado, The University of Michigan
  Presenting Author