Situational and Relational Engagement in Urban Science Classrooms: A Mixed-Methods Study

• In Event: Student Engagement Across Contexts

Abstract

Objective and Perspective: Student engagement is crucial for science learning (Authors, 2020; Wang et al., 2016). This study takes a situative perspective of student engagement in complex science classrooms (Hilpert & Marchand, 2018). An explanatory mixed methods approach (Creswell & Plano Clarke, 2011) was used to 1) establishes a bifactor model of engagement and 2) examine the predictive strength of antecedents related to culturally responsive environments (belongingness, Gray et al., 2018; teacher relationships, Varelas et al., 2002; discourse opportunities, Brown & Spang, 2008) on student engagement, and 3) qualitatively illustrate relational processes (Calabrese Barton et al., 2008) that support engagement in urban science classrooms.

Method and materials. Data was collected from a total of 101 middle school students. The sample was 45.5% male (n = 45) and 52.5% female (n = 52), with M age of 12.66 (SD = 0.92); White/15.8%, Black/38.6%; LatinX/17.8%; Asian/3.0%; Other/Mixed Race/22.8%.
Student questionnaire (adapted from Anderman, 2002; Wang, et al., 2019). The student questionnaire consisted of five subscales including engagement (cognitive, behavioral, affective, and social, sense of belonging, teacher relationships, and discourse opportunities (Cronbach’s alpha ranged from .65 to .86).
Classroom videos/audios. Seven classroom videos were collected. One camera followed the teacher and the other captured group interactions. The teacher wore a microphone and audio-recorders were used to capture small-group discussions.

Analyses. Mplus8 (Muthén & Muthén, 1998–2017) was used to conduct the analyses, with MLR estimation. Model fit was assessed using the cut-off criteria recommended by Hu and Bentler (1998, 1999; CFI ≥ 0.90, TLI ≥ 0.90, RMSEA ≤0.06, and SRMR ≤ 0.06). A structural equation model was then estimated with the latent variables sense of belongingness, teacher relationship, and discourse opportunities predicting the global and four engagement dimensions (Figure 1). Videos were segmented into episodes of student-teacher and student-student interactions, and coded for the markers of engagement.

Results. The quantitative results showed that the fit of the bESEM showed best fit (Table 1). In the SEM, results showed that sense of belonging was a significant predictor of social engagement (ß = .85), whereas relationship with teacher predicted affectieve engagement (ß = .60). Qualitative analyses illustrated the situation-specific nature of student engagement that parallels some of the quantitative relationships identified, but also identifies new links. For example, opportunities for discourse through open-ended questions (e.g., What TV shows or movies remind you of DNA?) triggered students’ interest (affect) and idea generation (cog). Close-ended questions (e.g., What is located in DNA?) triggered students’ recall of science content (e.g., adenine, guanine; cog engagement). Teacher check-ins with students (e.g., personalized encouragement) was associated with affective and social responses (e.g., students sharing about their academic performance in other classes).

Significance. This mixed methods study illustrates how situation- and relationship-specific factors relate to a global and four dimensions of engagement. Findings suggest equitable science learning is achieved through culturally responsive (relational) practices that differentially impact moment-to-moment engagement in urban science classrooms.

Authors

• Christine Lee Bae, Virginia Commonwealth University

• Daphne C. Mills

• Martinique A. Sealy, Virginia Commonwealth University