Psychometric Properties of a Measure of Behavioral and Emotional Engagement in Academic Activities in Middle School Students

Academic engagement is a key contributor to students’ school success and it typically refers to student behaviors such as attending to the teacher and time on task; conversely, off-task or disruptive behaviors have routinely been classified as disengagement (Bishop & Pflaum, 2005).In the short term, engagement predicts student's learning, grades, and achievement test scores; over the long term, it predicts patterns of attendance, retention, graduation, and academic resilience (Skinner et al., 2008). Studies have also suggested that academic engagement serves as a protective factor against risky activities (O’Farrell & Morrison, 2003) such as substance abuse, risky sexual behavior, and delinquency (Skinner et al., 2008). Thus, students who are engaged in school are both more suc-cessful academically and more likely to avoid the pitfalls of adolescence (Skinner et al., 2008).Scholars recently argue that academic engagement involves three interconnected dimensions, behavioral engagement, emotional engagement, and cognitive engagement. Behavioral engagement encompasses student’s effort, persistence, participation, and compliance with school/classroom rules and structures. Emotional engagement includes student’s positive emotions related to school activities. Cognitive engagement is a student’s completion of academic tasks and monitoring of his or her own learning habits.Academic engagement correlates strongly with academic achievement. One of the teacher's aims is to connect their students to school and to learning because they know that engagement is crucial to school success.Since there is no evidence on the performance of Iranian participants on the behavioral and emotional engagement in academic activities, the purpose of this study was to investigate behavioral and emotional engagement in academic activities of Iranian middle school students using the Skinner et al.’s questionnaire (Skinner et al., 2009), to assess the psychometric properties (reliability and validity) of the measure, and to examine gender and grade level differences in behavioral and emotional engagement in academic activities. The participants of the present study were 354 middle school students (183 girls and 171 boys) from grades one to three who were selected using multi-state cluster sampling method. They completed the Skinner et al.’s behavioral and emotional engagement in academic activities questionnaire and Hubner’s life satisfaction questionnaire in their regular classes and in random ordering.The questionnaire of children’s behavioral and emotional engagement in academic activities in the classroom (Skinner et al., 2009) was used to measure behavioral and emotional engagement and disengagement in academic activities. It is a 27-item self-report scale measuring student’s behavioral and emotional engagement/ disengagement in academic activities on a 4-point Likert scale, ranging from 0 to 4 (0=“never true”, 4=”almost always true”). The questionnaire is appropriate for students aged 9-18 years. Previous studies have reported adequate internal consistency, test-retest reliability, and convergent validity for this questionnaire (Furrer et al., 2006; Skinner et al., 2009).Other Instruments: The Life Satisfaction Questionnaire (Hubner, 2001) was used to establish the convergent validity of the Skinner et al.’s questionnaire. It is a 20-item self-report inventory. Each item is rated on a 1 to 5 scale. The life satisfaction questionnaire has adequate internal consistency reliability and has been well-validated (Latifian & Sheikholeslami, 2004). To determine the reliability, Cronbach’s coefficient alpha was utilized and the obtained coefficient for factor1 scores, factor2 scores, and total scores were. 79,. 87, and. 89, respectively, The correlation coefficient between total scores of life satisfaction and the two obtained factors (engagement and disengagement) were. 30 and -.39, respectively (p≤.01). These correlation coefficients could be considered a measure of convergent validity.The results of confirmatory factor analysis through AMOS Software revealed two factors: behavioral and emotional engagement and behavioral and emotional disengagement in academic activities.The fit indices suggested a good fitting data (RMSEA=.048, RMR=.050, IFI=.933, TLT=.914, GFI=.911, CFI=.932, X2=469.903, df=258, X2/df=1.821).The data were analyzed using a 2×3 two-way analysis of variance design with gender (2 levels) and grade level (3 levels) as independent variables. The dependent variables were behavioral and emotional engagement (or disengagement) in academic activities. The results revealed that gender had no significant main effect on behavioral and emotional engagement in academic activities, but the main effect of grade level (F=8.76, p=.0001) and the interaction effect of gender and grade level (F=3.21, p=.042) on the dependent variable (behavioral and emotional engagement) was statistically significant. The results of Scheffe post hot test indicated that the performance of students in grade one was significantly better than the performance of students in grade two. The difference between the mean scores of behavioral and emotional engagement of the two grade levels was not statistically significant. The findings also showed that the main effect of gender (F=3.98, p=.047) and interaction effect of gender and grade (F=3.68, p=.026) on disengagement in academic activities were statistically significant, but the main effect of grade (F=.51, p=.602) on the dependent variable was not statistically significant. This study was conducted to investigate the psychometric properties of the Skinner et al.’s behavioral and emotional engagement in academic activities questionnaire and to study the effect of gender and grade on the performance on this questionnaire. The results indicated that the reliability and validity of the questionnaire were at a satisfactory level and therefore the questionnaire can be used in an Iranian context.