College student engagement has been a major focus of research. College student engagement can influence learning effectiveness and future societal development. Kuh (2003) determined that engagement is essential for a college student; the more college students study in their field, the more they understand the subject area. Moreover, students perform better when receiving feedback from faculty. Engagement can improve problem solving, cooperation with diverse cohorts, and tolerance of different perspectives. Epistemic cognition refers to how individuals form opinions on the essence, source, and verification of knowledge. Since the study of Perry (1970) on the ethical and intellectual development of students at Harvard, epistemic cognition of college students has been a major field of research. Different from Perry’s qualitative interview method, Schommer (1990) used a quantitative questionnaire survey method to explore the influence of personal “epistemological beliefs” on reading comprehension. The results revealed that epistemological beliefs are closely related to reading comprehension. College life is full of emotions, including joy, hope, pride, anger, fear, and boredom (Pekrun, 2019). Epistemic emotions are felt when processing knowledge and are a key element of the learning process. Studies have indicated that the epistemic cognition of college students can directly influence engagement (DeBacker & Crowson, 2006; Heiskanen & Lonka, 2012; Ravindran et al., 2005). When college students believe that they can construct knowledge, student engagement is promoted. Furthermore, the epistemic cognition of college students influences their engagement indirectly through epistemic emotion (Bahcivan, 2019; Muis et al., 2015; Trevors et al., 2017). When college students believe that they can construct knowledge, they come to possess positive epistemic cognition that enhances their engagement (Denovan et al., 2020; Q. Zhang & J. Zhang, 2013).

On the basis of the literature review, this study proposed the following five hypotheses:

H1: The epistemic cognition of college students has a significant impact on their engagement.

H2: The epistemic cognition of college students has a significant influence on their positive epistemic emotions.

H3: The epistemic cognition of college students has a significant influence on their negative epistemic emotions.

H4: The epistemic cognition of college students has an intermediary effect through their positive epistemic emotions.

H5: The epistemic cognition of college students has an intermediary effect through their negative epistemic emotions.

This study included a pretest in which 314 valid questionnaires were collected. Thereafter, the test was conducted, and 514 valid questionnaires were collected. Four research instruments were used in this study: The Epistemic Cognition Scale, Positive Epistemic Emotion Scale, Negative Epistemic Emotion Scale, and Student Engagement Scale. The items of each of the four scales were rated on a 5-point Likert scale. All the items were reviewed by eight university professors. Before conducting the pretest, we revised the questions on the basis of the advice of the professors. All four research instruments demonstrated good content validity. Exploratory and confirmatory factor analyses were performed using all four scales to test the construct validity, and the reliability of each scale was tested using Cronbach’s α coefficients. The α coefficients obtained for the four subscales of the Epistemic Cognition Scale were between .61 and .85. The α coefficients for the two subscales of the Positive Epistemic Emotion Scale were between .70 and .84. The α coefficients for three subscales of the Negative Epistemic Emotion Scale were between .65 and .84, and those obtained for the three subscales of the Student Engagement Scale were between .68 and .78. Thus, the four scales demonstrated good reliability and validity.

AMOS 21.0 statistical software was used to test the mediating effect model and the structural equation model. The results revealed that H1 was supported, and the standardized regression coefficient was .35 (p < 0.001); H2 was supported, and the standardized regression coefficient was .26 (p < 0.001); H3 was supported, and the standardized regression coefficient was -.11 (p < 0.05); and H4 was supported, and the indirect effect was .40; H5, however, was not supported.

The results of this study revealed that the epistemic cognition of college students has a direct effect on their positive epistemic emotion. Thus, when college students have a diverse range of knowledge and they believe that they can construct their own knowledge, they develop more positive epistemic emotions and fewer negative epistemic emotions. Additionally, their engagement increases. During the learning process of college students, the more positive epistemic emotions students develop, the more learning activities they may engage in.

According to the literature, the indirect effect of epistemic cognition, epistemic emotion, and student engagement is supported by empirical data. Epistemic cognition interacts with positive epistemic emotion and indirectly influences student engagement. If the epistemic cognition of college students has multiple constructs, students may develop positive epistemic emotions that improve engagement. Consequently, positive epistemic emotions play a mediator between epistemic cognition and student engagement. However, the indirect effects of epistemic cognition, negative epistemic emotion, and student engagement were not supported. Therefore, epistemic cognition does not interact with negative epistemic emotions, and student engagement was not directly influenced.

Because epistemic cognition profoundly affects the engagement of college students, we suggest that higher education administrative bodies pay attention to the epistemic cognition of students. Additionally, universities can provide students with relevant resources for developing more diverse opinions among students. Future research can further investigate the mediating effects of epistemic cognition, epistemic emotion, and student engagement, which were inconclusive in the present study.