Providing Daily Remote Learning Services During the COVID-19 Pandemic: Examining Disparities Between Pre-Kindergarten Teachers Across Urbanicities in North Carolina

• In Event: Poster Session 10
  In Poster Session: PS 10 Topic - School Readiness/Childcare

Abstract

During the spring of 2020, early childhood educators in North Carolina began to navigate the transition to remote learning (RL) as a result of the COVID-19 pandemic and statewide quarantine. The current study examined the experiences of 1,918 teachers who administered North Carolina’s NC Pre-K program and provided RL services during this period. Two research questions were address: (1) Did the frequency at which NC Pre-K teachers provide RL services to children and families on a daily basis vary across urbanicities? (2) Did teachers’ lack of access to reliable internet and technology mediate the association between frequency and urbanicity?

In this study, I defined RL services as the services that teachers provided to children and families to facilitate learning that would otherwise take place in the classroom. I further defined RL services as those services provided through video-based communication (e.g., Zoom), phone calls, emails, text messages, or sending home learning or activity packets. I calculated the percent of children in each classroom enrollment to whom teachers reported providing RL services on a daily basis during the spring of 2020. Teachers were also asked to indicate if they needed access to reliable internet and/or technology in order to better provide RL services to children and families (1 = yes, 0 = no). I indexed the level of urbanicity in the county where the teacher was employed based on population density (i.e., the number of people residing in the county per square mile). Other characteristics of teachers and counties were also considered (see Table 1).

I estimated multi-level regression models using the GLIMMIX procedure in SAS 9.4 to model (a) the percent of children served daily as an outcome using the “beta” distribution, which is appropriate for modeling continuous proportions, and (b) teachers’ need for access to technology/internet as outcomes using the “binary” distribution. These multi-level models accounted for the nesting of teachers (Level-1) within counties (Level-2). I found that population density was positively associated with the percent of children served daily (b=0.25, p=0.49; see Table 2), such that a 1 standard deviation unit increase in population density resulted in a 25% increase in the percent of children that teachers served on a daily basis (see Table 2). Based on Sobel tests of mediation, I found no evidence to indicate that teachers’ need for access to reliable internet (b=0.15, SE=0.00, p=0.88) nor reliable technology (b=0.39, SE=0.02, p=0.70) mediated the association between population density and the percent of children served daily. I also examined the non-linear association between population density and the percent of children served daily by including a quadratic term for population density in the main regression model, but found no evidence to indicate that the association was enhanced/diminished at higher/lower levels of population density (b=0.03, SE=0.09, p=0.77). These findings suggest that teachers experienced greater challenges in providing RL services to children and families in the more rural compared to the more urban regions of NC. Therefore, children in rural communicates may have experienced greater learning loss during the COVID-19 pandemic.

Author

• Robert Carr, Duke Univeristy
  Presenting Author