Conference Agenda

Some scholars suggest that teachers act in a caregiving role and can serve as a secure base for young students (Spilt & Koomen, 2022). Consistent with attachment theorists’ claims that early attachments set the stage for later caregiving relationships, some studies have found significant links between early attachment security and later teacher-student relationships (Howes & Matheson, 1992), though others report conditional or no associations (Howes & Hamilton, 1992; DeMulder et al., 2000), demonstrating the need for systematic synthesis. The present meta-analysis examined the association between early attachment to primary caregivers and subsequent teacher–student relationships, and explored whether this association varied by timing between assessments, methodological features, or operationalization of attachment and relationship domains. Studies were identified using the Child Attachment Studies Catalogue and Data Exchange (CASCADE; Madigan, 2020), a comprehensive repository of studies assessing caregiver–child attachment using gold-standard observational measures. Eligible studies assessed parent–child attachment using the Strange Situation Procedure or Attachment Q-Sort and measured teacher–student relationship quality using the Strange Situation, Attachment Q-Sort, or Student–Teacher Relationship Scale. Studies were included if teacher–student relationships were assessed concurrently with or after parent–child attachment. Effect sizes (correlations, Cohen’s d, or odds ratios) were extracted and converted to Fisher’s z. Multilevel random-effects meta-analytic models with robust variance estimation were conducted in R to estimate the overall association and test moderation. Results indicated a very small but statistically significant association (r = .07), suggesting that early attachment and teacher–student relationship quality represent largely distinct interpersonal domains. The association was not moderated by measurement type, child age, or timing between assessments. Compared to prior meta-analytic findings on early attachment sequelae (e.g., Groh et al., 2017), effects were modest. Notable gaps in the literature included limited racial and socioeconomic diversity and sparse reporting of teacher characteristics.

Attachment theory is often applied to teacher-child relationships (TCR), framing them as secondary attachment relationships playing a crucial role in children’s development, including executive functions (EF) (Verschueren & Koomen, 2012). Notably, teachers have greater difficulty forming high-quality TCRs with children with EF problems (Sankalaite et al., 2023). To better understand these challenges, this multi-method study explores teachers’ perspectives on TCRs with these children.

Most research relies on direct measures of TCRs via questionnaires (e.g., STRS; Pianta, 2001), capturing explicit perceptions but overlooking implicit processes. These implicit, semi-conscious beliefs and feelings are automatically activated in interactions but only indirectly accessible via interview-based methods (Spilt et al., 2022). The Five-Minute Speech Sample (FMSS; Magaña et al., 1986), widely used in parent-child research (Sher‑Censor, 2015; Sher‑Censor et al., 2016), offers a brief way to elicit caregiver narratives to assess implicit processes and has been adapted for educational settings (McGrath & Van Bergen, 2019).

We examined teachers’ relationships with three low‑EF children (randomly selected using the BRIEF‑P; van der Heijden et al., 2013; N=72) using explicit and implicit measures: the STRS (closeness, α=.74; conflict, α=.83; dependency, α=.83) and the FMSS. Using the FMSS, teachers spoke five minutes about the child. Transcripts were coded on focus (ICC=.67), elaboration (ICC=.87), separateness (ICC=.49), concern (ICC=.82), acceptance (ICC=.81), complexity (ICC=.69) and coherence (ICC=.80).

The STRS indicated similar levels of closeness and conflict but significantly higher dependency (p<.05, d=.53) than the norm population. FMSS scores showed that teachers were accepting of children’s challenges and described problems in detail (elaboration), though narratives sometimes lacked complexity. Elevated concern suggested more worries related to various domains of development for low‑EF children. We will discuss the added value of employing an implicit measure to study TCRs, evaluate the FMSS’s validity for this purpose, and delineate the classroom implications of the findings.

Early childhood educators act in loco parentis, and evidence suggests that, like caregivers at home, they can serve as a secure base for children to explore their environments and support healthy psychosocial adjustment (Verschueren & Koomen, 2012). Children with lower-quality teacher-cihld relationships tend to demonstrate more symptoms of psychopathology and lower social functioning (Lei et al., 2016; Magro et al., 2023). Understanding the origins of healthy teacher-child relationships is thus of great concern. However, relatively little is known about teacher traits that support the formation of positive teacher-child relationships in early childhood (Spilt & Koomen, 2022). Thus, the aim of the present study was to test whether teachers’ attachment representations and recalled caregiving experiences are associated with teacher-student relationship quality.

The present study used data collected from 157 childcare providers (84.7% non-Hispanic White, 98.7% women) on 636 three-to-five-year-old children (79.9% non-Hispanic White, 51.3% girls). Providers completed the Attachment Script Assessment (Waters & Waters, 2006), the Adverse/Benevolent Childhood Experiences Questionnaires (Felitti et al., 1998; Narayan et al., 2018), and the Maternal/Paternal Emotional Availability Scales (Sorhagen et al., 2019). Providers also completed the Student-Teacher Relationship Scale (Pianta, 2001) for up to 5 children in their care.

Multilevel SEM was used to estimate associations with teacher-student conflict and closeness. Unconditional models revealed that 14.9% of the variance in conflict and 23.6% of the variance in closeness was at the teacher level. After controlling for child (age, gender, race, language, disability status) and teacher (gender, race, language, education, years of experience, parental status) demographics, adding the block of teacher attachment and caregiving variables significantly improved model fit for both conflict (Δχ²(45)=282.31, p<.001, ΔR²=.06) and closeness (Δχ²(45)=287.16, p<.001, ΔR²=.09). However, no teacher-level variables emerged as unique predictors of relationship quality, suggesting that teachers’ attachment representations and caregiving histories are collectively associated with teacher-child relationships.

N/A -- Discussant