In higher education, the pathways through which teacher expectations could contribute to students’ academic achievement are an area of limited empirical investigation. In this study, we investigated the mediating role of the socioemotional classroom environment in the relations between university teachers’ early expectations and their students’ later achievement. The participants were 176 teachers and their 6,506 first-year undergraduate students from public universities in China. Data from teachers’ initial expectations, students’ prior and year-end achievement on standardized tests, and students’ perceptions of socioemotional factors within the classroom were collected. As students were nested in classes, multilevel structural equation modeling was employed for data analysis. The results showed that, at the individual level, with students’ prior achievement controlled, the extent of the teacher’s differential treatment, the quality of the teacher-student relationships, and the quality of the peer relationships in turn were statistically significantly related to the university teachers’ initial expectations, which ultimately was associated with the individual students’ year-end academic achievement. Further, the university teachers’ expectations for their individual students played a role in the quality of the peer relationships which subsequently contributed to students’ later academic achievements. Moreover, the university classes whose teachers tended on average to have higher expectations tended to achieve at higher levels. However, we did not identify mediating paths via socioemotional factors from university teacher expectations to student achievement at the class level.
Differential access to and enrollment in advanced mathematics for historically underrepresented groups is a pervasive problem in education, however, research has primarily focused on achievement rather than access. This necessitates an examination of who is accessing advanced coursework and what differentiates the course trajectories students follow across middle and high school. We utilized data from a large (N = 18,841), majority Latinx (57.6 % Latinx, 35.5 % Black, 6.1 % White/Other, and 0.6 % Asian/Pacific Islander) and low-income (77 % free/reduced-price lunch) sample of students followed longitudinally from middle through high school. Latent class analysis (LCA) categorized students into 6 classes representing commonly followed pathways of advanced math course taking from grade 6 to 12. Multinomial logistic regression was used to connect individual demographics, school readiness skills, and prior achievement variables to the likelihood of being assigned to a particular class. Prior academic performance was most strongly related to advanced math pathway assignment, but even controlling for this, gender, disability status, and cognitive and fine motor skills at age 4 also impacted the math pathway a student was likely to follow in middle and high school. Race/ethnicity was a significant differentiator only when comparing the two most advanced pathways. These findings highlight the importance of early school readiness skills and demonstrate how early opportunity gaps impact later student outcomes. Tailored intervention and supports are necessary to ensure equitable access to coursework which expands a student’s opportunities and chances for postsecondary success.
Using Control-Value Theory, a questionnaire survey and daily diary method were combined to explore the relation between control-value appraisals and achievement emotions in Chinese maths classes. With data from 602 secondary school students (Mage = 13.55, 51.66 % female) who participated in the questionnaire survey and 112 students (Mage = 13.91, 48.21 % female) who completed the daily diary measure over three weeks, the results of multileveled multiple regression confirmed the relation between control-value appraisals and achievement emotions under the two measures. Nonetheless, there were some differences in those relations between the two measures. Lesson-specific emotions were strongly related to internal attribution and intrinsic value, whereas domain-specific emotions were highly related to maths self-concept and intrinsic value. Lastly, the results identified the control-value interaction effect on several commonly reported emotions under the two measures.