Poster #66 - Contexts Matter: A Multilevel Modeling Approach to Examine the Perceived Academic Stress Among Chinese Adolescents

• In Event: Poster Session 14

Abstract

Background: While academic stress is more prevalent and therefore mostly studied among college and graduate students in the United States (e.g., Beiter et al., 2015; Misra & McKean, 2000), it is a primary source of everyday stress faced by Chinese adolescents as a result of the examination-oriented educational system in China (Liu & Tein, 2005; Wang & Chen, 2001). The associations between academic stress and other psychological constructs such as psychological wellbeing, depression, and suicidal ideation and attempts have been well established among this population (e.g., Liu & Lu, 2011; Ying et al., 2020), but less is known about how academic stress varies at different levels of context in which adolescents are situated (Bronfenbrenner & Morris, 2006). Using a multilevel modeling approach, the current study conducted a preliminary examination of the variations in different dimensions of Chinese adolescents’ perceived academic stress (PAS) at the individual, classroom, and school levels.

Method: The sample included 598 Chinese adolescents (53.2% female) from 28 classrooms at two public high schools in China. Perceived academic stress (PAS) was measured by the Academic Stress Scale for Adolescents (ASSA; Huang, 2022), which consists of 16 items measuring three dimensions related to PAS: (1) Academic burden, (2) meeting self-expectations, and (3) meeting external expectations from parents and teachers. Each item was rated on a 1-5 Likert scale, with higher score representing greater stress. Factor scores were computed by averaging corresponding item scores. Following confirmatory factor analysis (CFA), a series of multilevel models, with students nested within classrooms, were conducted to examine how individual, classroom, and school factors are associated with each dimension of PAS.

Results: CFA supported the three-factor structure of the ASSA among the current sample, χ² (98) = 367.81, p < .001, SRMR = .04, RMSEA = .07, CFI = .96, TLI = .95. Results of multilevel modeling suggested that PAS related to academic burden (ICC=.06) and meeting self-expectations (ICC=.02) varied significantly across classrooms. At the individual level, girls experienced greater PAS related to academic burden (β=0.15, p=.05) and self-expectations (β =.15, p=.05) than boys, after controlling for school, grade level, and self-reported family income. Meanwhile, 12th graders reported higher level of PAS from these two dimensions than their younger counterparts (β =0.23, p=.02 and β =0.17, p=.06, respectively). In addition, moderation analysis suggested significant cross-level interaction between school and grade level on PAS related to academic burden and self-expectations. More specifically, while students from one school experienced PAS from both dimensions similarly throughout high school, the stress level elevated significantly in grade 12 for students from the other school (see figures below). PAS related to external expectations did not vary by gender, grade level, classroom, or school.

Discussion and Implications: The study expanded the understanding of perceived academic stress among Chinese adolescents. First, it is a multidimensional construct. Second, the overall stress level is median-high. Third, it varies across different levels of contexts: individual, classroom, and school. Future studies should include more indicators of classroom and school characteristics to disentangle why and how contexts matter.

Authors

• Yawei Huang, Boston University
  Presenting Author

• Jonathan F Zaff, Boston University