Poster #79 - Identification of Four Teacher-Child Interaction Styles in Pre-Kindergarten Classrooms

• In Event: 3-053 - Poster Session 10
  In Poster Session: PS 10 Section - Education, Schooling

Integrative Statement

Previous work has begun to apply a parenting styles framework (Baumrind, 1966; Maccoby & Martin, 1983) to teacher-child interactions (Uibu & Kikas, 2014; Walker, 2008; Wentzel, 2002). While many studies focus on teachers of older students (Bassett et al., 2013; Pellerin, 2005; Wentzel, 2002) there is some evidence that an authoritative teacher-child interaction style in elementary school predicts better academic outcomes (e.g., Kiuru et al., 2012). The current study aims to extend this work to pre-kindergarten and identify teacher-child interaction styles using subscales of the Classroom Assessment Scoring System-PreK version (CLASS; Pianta, La Paro, & Hamre, 2004). Validity will be explored by examining teacher characteristics as predictors of styles.

The current sample was from the combined Multi-State Study of Pre-Kindergarten and State-Wide Early Education Programs Study dataset (Early et al., 2013). Participants were 694 lead teachers in pre-kindergarten classrooms (16% Head Start, 61% public school) in 11 states. Teachers were 96% female, 68% white, and 54% had BA and state teaching certificates. Positive climate, teacher sensitivity, behavior management, negative climate, and overcontrol were observed and rated on a scale from one to seven, using the CLASS. Scaffolding was observed using the Emerging Academics Snapshot (Ritchie et al., 2001); higher scores represented more awareness/support of child needs and learning. Teachers reported traditional attitudes toward child rearing (Schaefer & Edgerton, 1985; α = .79); higher scores represented more traditional/authoritarian attitudes.

Data were analyzed using Latent Profile Analysis in Mplus (Version 7.3; Muthén & Muthén, 2014) to identify groups of classrooms characterized by different teacher-child interaction styles. Model fit was evaluated using AIC, BIC, and sample-size adjusted BIC indices, and p-values for VLMR, LMR, and BLRT. Results indicated a four-group solution best fit the data (VLMR = .015, LMR = .016, BLRT< .001, entropy = .84) because it significantly improved model fit beyond the three-group solution, while the five-group model was not significantly better. Group means for the five observed variables are presented in Figure 1. An authoritative interaction style group was identified based on highest ratings of positive climate, sensitivity, and behavior management as well as low negative climate and overcontrol. An uninvolved group was identified based on low positive climate, sensitivity, and behavior management and moderate overcontrol and negative climate. The group with lowest positive climate, sensitivity, and behavior management and highest overcontrol and negative climate was identified as authoritarian. The group with moderate positive climate, sensitivity, and behavior management and low overcontrol and negative climate, was identified as good enough. Finally, teacher beliefs and scaffolding were tested as predictors of profiles using the 3-step procedure (Asparouhov & Muthén, 2014; see Table 1). More traditional beliefs predict greater odds of authoritarian and uninvolved teacher-child interaction style, while greater scaffolding predicts significantly greater odds of good enough and greatest odds of authoritative style.

Findings provide initial support for using the CLASS to extend a broader “teaching styles” framework to teacher-child interactions in early childhood classrooms. Future work should extend findings by exploring the influence of teacher-child interaction styles on children’s academic outcomes.

Authors

• Ashley Kimble, Oklahoma State University
  Presenting Author

• Laura Hubbs-Tait, Oklahoma State University
  Non-Presenting Author

• Isaac J Washburn, Oklahoma State University
  Non-Presenting Author

• Amy C Williamson, Oklahoma State University
  Non-Presenting Author

• Lucia Ciciolla, Oklahoma State University
  Non-Presenting Author