Introduction

Learning engagement is an evaluative indicator of student growth experiences as well as a predictor of quality in higher education1. University students learn more when they invest their energies in a variety of educationally purposeful activities2. An academic survey based on 311 undergraduate higher education institutions in China concluded that university students’ learning engagement is an important predictor variable of academic success in college3. Many scholars hold similar views4,5. Learning engagement is a function of the time and energy that an individual invests in learning, focusing on the cognitive and affective6as well as behavioral and social factors7that an individual experiences while engaged in the process. It is the sum of the physical and mental energy that a student puts into the academic experience8. University students’ learning engagement consists of three different behaviors: cognitive effort (the amount of time students spend on coursework), affective participation (students’ participation in classroom questions, discussions, and project-based learning), and interactions with instructors (students’ discussions of assignments or career plans with instructors)9. China College Student Survey (CCSS), the National Survey of Student Engagement (NSSE) in the United States, and the National Student Survey (NSS) in the United Kingdom reflect the importance that countries place on university students’ learning engagement. University students’ learning engagement has become a widespread concern in various countries. According to the latest statistics from the Ministry of Education of China, the total number of students enrolled in higher education is more than 58 million, of which 19,656,400 are enrolled in general undergraduate programs10. China has established the most extensive higher education system globally and remains steadfast in its commitment to fostering high-quality development within the sector of higher education11. However, the lack of investment in learning by Chinese university students has become a significant constraint on the transformation of China’s higher education from “big” to “strong”12. In China, many parents attach great importance to their children’s education. Education expenditure is an important part of household expenditure. Before students go to university, their studies are arranged tightly by parents and teachers, especially in senior high school. When it comes to university, students have more time at their disposal, but they may face the awkward dilemma of not knowing how to use their time wisely and how to learn properly. Therefore, this study focused on a group of Chinese university students and explored their learning engagement as China’s higher education entered a phase of universalization. An experimental intervention study has shown that enhancing students’ positive self-concept can significantly improve their learning engagement13. Self-concept constitutes a suite of attitudes that individuals possesses regarding themselves, encompassing cognitive and affective dimensions of self-perception, including an awareness and comprehension of one’s own values, interests, competencies, and personality traits. It has emerged as a pivotal construct in elucidating human behavioral patterns14. Professional identity is the embodiment of self-concept in the professional domain. Based on the theory of professional identity, it has been suggested that students’ cognition, feeling and commitment to their majors will affect their learning engagement behavior15. It has been confirmed that professional identity has a significant positive predictive effect on learning engagement16,17,18. In addition, according to Bandura’s self-efficacy theory, students’ self-efficacy has a significant impact on their classroom participation and learning persistence19. Existing researches have shown that academic self-efficacy significantly and positively predicts learning engagement20,21,22. Moreover, self-efficacy acts as an active precursor of self-concept development23. There may also be correlation between professional identity and academic self-efficacy. While prior research has frequently examined the relationship among these three dimensions without delving into their respective sub-dimensions, this study aims to investigate the interrelationships between the distinct sub-dimensions of professional identity and learning engagement, as well as to elucidate the mediating role of academic self-efficacy sub-dimensions within this dynamic process. Therefore, our research facilitates a more comprehensive, specific, and in-depth understanding of the influence mechanisms of Chinese university students’ learning engagement.

The relationship between professional identity and learning engagement

Professional identity is a subjective feeling within an individual balanced with the profession24. It is a gradual, positive, and dynamic development process25. Professional identity is regarded as an emotional acceptance and recognition of learners’ cognitive understanding of the profession they are studying, accompanied by positive external behaviors and an internal sense of fitness. It is divided into four dimensions: fit, behavioral, affective and cognitive. Fit professional identity reflects the degree of match between the individual and the profession; Behavioral professional identity reflects the individual’s behavioral performance of professional learning; Affective professional identity reflects the extent to which an individual holds emotional preference for the profession, while cognitive professional identity indicates the degree to which they understand its basic aspects26. A number of studies have confirmed a significant positive correlation between professional identity and learning engagement27,28,29,30. A study of medical students has found that a higher sense of professional identity could foster confidence in their future roles as physicians and positively shapes their interactions with colleagues, professional groups and patients. It helps to promote the integration of students’ selves with their career path and increases their learning engagement in their field, thus subsequently facilitating mature career decisions31. An analysis of the literature from 2019 to 2023 further reveals that students with a stronger alignment in professional identity are more likely to adopt tailored and personalized learning strategies to address academic challenges. These students also exhibit a greater tendency to pursue elevated standards of professional performance and engage in extracurricular activities that foster professional growth, thereby enhancing their overall level of academic engagement32. Cognitive professional identity and fit professional identity are positively correlated with learning engagement. Besides, students develop their professional values through interaction and collaboration with other students in the classroom. When the interaction process is good, it facilitates students’ affective professional identity, promotes their self-development and increases their learning engagement33. In a large-scale study of 10,901 medical students across 11 universities in China, it has found that the higher the students’ affective professional identity, the better it is for stimulating positive qualities such as optimism, resilience, sense of meaning and creativity, ultimately facilitating better performance34. These students are able to solve problems in the learning process with a relatively positive, fulfilling, and full mindset. They have a high level of concentration and are willing to put a corresponding amount of energy into learning and overcome difficulties35. A correlation also exists between affective professional identity and learning engagement. In addition, students with higher levels of professional identity tend to perceive tasks as variable, challenging, meaningful, and interesting. This positive perception fosters greater receptivity towards instructors, peers, and the academic environment. Consequently, these students demonstrate a greater propensity to employ deep learning strategies, investing the necessary cognitive effort to comprehend complex concepts, ultimately enhancing their focus and learning engagement36. Behavioral professional identity is also significantly and positively related to learning engagement. In summary, hypothesis H1 is proposed:

H1: Professional identity has a significant effect on learning engagement.

H1a: Fit professional identity has a significant effect on learning engagement.

H1b: Behavioral professional identity has a significant effect on learning engagement.

H1c: Affective professional identity has a significant effect on learning engagement.

H1d: Cognitive professional identity has a significant effect on learning engagement.

Mediating effects of academic self-efficacy

Academic self-efficacy refers to an individual’s judgment and confidence in his or her ability to successfully complete academic tasks, and it is divided into two dimensions: academic competence self-efficacy and academic behavior self-efficacy. Academic competence self-efficacy refers to an individual’s judgment and confidence in his or her ability to successfully complete school, achieve good grades, and avoid academic failure. Academic behavior self-efficacy, on the other hand, refers to an individual’s judgment and confidence that he or she is able to adopt certain learning methods to achieve learning goals37. It has been found that students with a high sense of professional identity derive intrinsic motivation from the learning process, bolstering their confidence in confronting academic challenges and obstacles. Such students are able to positively predict individual behaviors and make changes to improve their overall competence38. Research conducted on tourism students has further demonstrated that enhancing students’ awareness and comprehension of their professional prospects, coupled with guiding them to establish specific, actionable, and phased goals, can significantly improve their academic self-efficacy39. Professional identity is positively related to academic self-efficacy40. Besides, academic self-efficacy affects students’ learning engagement41,42,43. A survey of second language learning has found that learning engagement is an important factor influencing students’ second language proficiency, as well as one of the key outcomes of academic self-efficacy. As an important factor affecting students’ psychological belonging and academic motivation, professional identity can further promote learning engagement by stimulating academic self-efficacy44. Academic self-efficacy plays a mediating role between professional identity and learning engagement. A study of 1,162 rural tuition-free medical students has found that students with a low sense of professional identity believed that “it is the same whether they study or do not study, and whether they study well or not”. They tend to be pessimistic and disillusioned about the future, and are not willing to spend time and effort to complete their studies in their major at a high level, or even fulfill to their duties as interns during the internship period45. This group of students are less goal-oriented, tend to focus on things that could eventually go wrong, and often imagine failure scenarios that hinder their actual ability and undermine their performance, leading to poorer performance in their studies and lower levels of their learning engagement46. Academic competence self-efficacy may mediate the relationship between professional identity and learning engagement. Another study on students’ self-directed learning has found that students with a low sense of professional identity exhibit diminished self-directed learning capabilities, struggling to personalize learning plans or effectively utilize resources. These students have low self-efficacy and lack the ability to effectively identify, analyze and synthesize information related to their learning needs. Consequently, they may find it difficult to derive pleasure and enrichment of taking in knowledge in the process of learning, which in turn leads to their avoidance of learning behaviors and their low level of learning engagement47. However, students with high professional identity can effectively grasp professional opportunities, identify and utilize objective development resources to effectively cope with learning tasks and academic challenges. They are willing to make efforts to improve their overall development skills48. For this reason, academic behavior self-efficacy may also play a mediating role between professional identity and learning engagement. In summary, hypothesis H2 is proposed:

H2: Academic self-efficacy mediates between professional identity and learning engagement.

H2a: Academic behavior self-efficacy mediates between the sub-dimensions of professional identity and learning engagement.

H2b: Academic competence self-efficacy mediates between the sub-dimensions of professional identity and learning engagement.

Chain-mediated effects of academic competence self-efficacy and academic behavior self-efficacy

Collaborative learning experiences and participation in professional learning communities can strengthen students’ understanding of their chosen major, foster a stronger sense of professional identity, and sustainably stimulate their interest in learning, thus making them more willing to invest more time and effort in their majors49. Another study has a similar view. Students with a high sense of professional identity are able to take advantage of relevant opportunities, which helps them to have clearer and more feasible steps to implement in their future plans. These students are more engaged in their studies50. Conversely, practicing nurses with a low sense of professional identity may experience a compromise in their beliefs regarding professional development due to persistent exposure to stressors, work overload, or high job demands. This can diminish their academic self-efficacy, subsequently leading to reduced learning engagement51. It has been shown that professional identity is positively related to learning engagement and that learning self-efficacy plays a mediating role. Academic self-efficacy is a strong predictor of students’ academic performance. Students with high academic self-efficacy have an active learning state. They are able to appreciate the physiological and psychological energy invested in the learning experience, and they are willing to put in the mental energy and effort to achieve the expected performance and continuously improve their overall competence52. There may also be a correlation between academic competence self-efficacy and academic behavior self-efficacy as two sub-dimensions of academic self-efficacy. A study of Romanian health care students during the COVID-19 pandemic has revealed that uncertainty regarding the pandemic’s duration, fear of infection, and unfamiliarity with the nuances of online education and assessments contributed to a decline in students’ self-confidence in their independent learning abilities, which resulted in a decline in all of their study skills owing to negligence in practicing to consolidate them. Their overall academic self-efficacy was low53. In a study of on-campus or online learning, it has found that students with high academic self-efficacy believe that they are able to monitor their own progress, control their learning resources (including the learning environment and study time), as well as regulate their own level of effort. These students exhibit a strong belief in their capacity to intentionally implement positive behavioral interventions to enhance their learning. This involves actively monitoring, regulating, and controlling their cognitive, motivational, and learning behaviors, thereby promoting progress in their academic achievement to a greater extent54. In summary, hypothesis H3 is proposed:

H3: Academic competence self-efficacy and academic behavior self-efficacy play a chain mediating role between the sub-dimensions of professional identity and learning engagement.

This study constructs the parallel mediator model and chain mediator model to examine the influence of professional identity on learning engagement and the mediating role of academic self-efficacy between the two in a group of university students, with a view to providing guidance for improving the level of university students’ learning engagement, and the theoretical conceptual model is shown in Figs. 1.

Fig. 1
Fig. 1
Full size image

Theoretical conceptual model.

A parallel mediation model assumes that each mechanism is independent and separate, accounting for the unique and incremental effects of the independent variables on given dependent variables55. They are conceived as independent and separate pathways leading to the outcome variables. Relative to the parallel mediation model, according to Rosen et al. (2014), the sequential mediation model has the advantage of incorporating multiple theoretical mechanisms into a comprehensive and integrative model, thereby providing a coherent explanation of the effects of given independent variables55.

Therefore, this study independently explained the mechanisms by which each pathway acted individually, analyzed the differences in their effect values through the parallel mediation model, and furthermore analyzed the unified coherent effects of the mediating variables through the chain mediation model. The present study adopted these two mediation models to deeply explore the path relationships among professional identity, academic self-efficacy, and learning engagement.

Research methodology

Participants

This study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of Jimei University(No.12/20240416). Informed consent was obtained from all participants involved in this study. Convenience sampling method was used to select undergraduate students in several universities (Jimei University, Jilin Agricultural University, Wuxi Taihu University, Chengde Medical University, Fujian Medical University, Fuzhou University of International Studies and Trade and so on) in China to conduct the questionnaire survey. The number of questionnaires issued was 4300. After screening invalid samples (reverse questions, trap questions), 4125 valid samples were obtained, and the response rate was about 95.9%. All the participants came from four-year universities students who obtained a bachelor’s degree after meeting the requirements. Higher vocational college and junior college students were excluded from participation. The age of the participants ranged from 17 to 26 years old (M = 19.734 years old, SD = 2.693), and the distribution of the samples was shown in Table 1. The raw data supporting the conclusions of this article will be available from Chunmei Chen (chunmei88@jmu.edu.cn) on reasonable requests.

Table 1 Sample distribution.
Full size table

Instruments

The three sub-questionnaires used in our study are all compiled by domestic scholars according to the characteristics of Chinese university students, and have good reliability and validity respectively. Moreover, in this study, these three questionnaire dimensions also had good reliability and validity regarding KMO value and Cronbach’s α coefficient, which were suitable for use in this research.

The university students’ professional identity questionnaire

The University Students’ Professional Identity Questionnaire in this study was developed by Qin in 200926. The questionnaire consists of 23 questions and is scored on a 5-point scale. The questionnaire consists of four dimensions: fit, behavioral, affective and cognitive. Fit items such as “I have good professional thinking”; behavioral items such as “I actively participate in practical activities related to my major”; affectivel items such as “I am willing to engage in work related to my major”; cognitive items such as “I understand the employment situation of my major. I am aware of the employment situation of my major”, etc. The questionnaire has no reverse scoring questions. The higher the individual’s score on a dimension (dimension mean), the stronger the individual’s professional identity on that dimension, and conversely, the weaker the individual’s professional identity on that dimension. In the total scale measure, the higher the total scale score (total mean score), the stronger the overall professional identity of the individual. The Cronbach’s α coefficient and KMO value of this questionnaire in this study were 0.907 and 0.965 respectively.

The university students’ academic Self-efficacy questionnaire

The University Students’ Academic Self-efficacy Questionnaire in this study was developed by Liang in 200237. The questionnaire consists of 22 questions with a 5-point scale. The questionnaire consists of two dimensions: academic competence self-efficacy and academic behavior self-efficacy. Academic competence self-efficacy has items such as “I believe I have the ability to do well in my studies”; Academic behavior self-efficacy such as “When I review for a test, I am able to review what I have learned before and after in a coherent way”. The scores of all items were averaged by reversing the scores of the reverse questions, and the higher the mean value, the higher the academic self-efficacy of the individual. The Cronbach’s α coefficient and KMO value of this questionnaire in this study were 0.921 and 0.951 respectively.

The university students’ learning engagement questionnaire

The University Students’ Learning Engagement Scale Questionnaire in this study was developed by Wang in 2014 in his PhD thesis56. The questionnaire consists of 22 entries on a five-point scale. Relevant items such as “I will study in advance what the teacher has taught in class”; “I actively participate in group discussions”; “Learning gives me a strong sense of satisfaction”. The questionnaire has no reverse scoring questions. The ratings of all items were averaged, with higher means indicating a higher level of individual learning engagement. The Cronbach’s α coefficient and KMO value of this questionnaire in this study were 0.896 and 0.963 respectively.

Statistical analyses

To achieve the study objectives, we employed a systematic analytical approach consisting of the following steps:

Descriptive statistics

We initially conducted descriptive statistical analyses to examine the fundamental characteristics of the dataset. This included calculating means, standard deviations, and frequency distributions for all key variables.

Common method Bias assessment

To identify and evaluate potential standard method bias, we performed Harman’s single-factor test following established procedures57. All measurement items were entered into an exploratory factor analysis, assuming that a single factor would emerge from the unrotated factor solution if substantial standard method variance were present.

Correlational analysis

Pearson correlation coefficients were calculated to examine the bivariate relationships among the study variables. This preliminary analysis provided initial insights into the strength and direction of associations between variables.

Multicollinearity diagnostics

To ensure the robustness of our regression-based analyses, we conducted variance inflation factor (VIF) diagnostics. A threshold of VIF < 5 was used to indicate the absence of problematic multicollinearity among predictor variables.

Mediation analysis

We employed Hayes’ (2017) PROCESS macro (Version 3.5)58 to test our mediation hypotheses. Specifically, Model 4 was used to examine parallel mediation effects; Model 6 was applied to investigate serial mediation pathways. The significance of mediation effects was tested using bias-corrected percentile bootstrap methods with 5,000 resamples. A 99% confidence interval was adopted, with effects considered statistically significant when the interval did not include zero59.

All statistical analyses were conducted using SPSS 26.0 with the PROCESS macro extension.

Results

Common method bias test

In this study, data were collected using a self-report method. While this approach is straightforward to implement, it carries the risk of introducing common method bias, which may compromise the accuracy of the study’s findings. To evaluate the extent of such bias, the Harman single-factor test was employed. This method assesses the presence of common method bias by conducting a principal component analysis to determine whether a single factor accounts for an excessive proportion of the variance in the data. The results of the test showed that there were a total of 13 principal components with eigenvalues greater than 1, and that the first principal component explained only 34.654% of the total variation, which did not exceed the standard threshold of 40%60. Based on this result, it can be concluded that the data of this study were not affected by serious common methodological bias. This finding enhanced the validity and credibility of the results of the study and showed that even if the self-reporting method was used to collect the data, the accuracy and reliability of the study was still guaranteed when appropriate statistical tests were put in place.

Descriptive statistics and correlation analysis of the variables

In this study, we used Pearson’s correlation analysis61 to investigate the relationship between standard deviation, professional identity, academic self-efficacy, and learning engagement. The results ( as shown in Table 2) of the study revealed that the correlation coefficient between professional identity and academic self-efficacy was 0.643, indicating a significant moderate-strength positive correlation between the two (p < 0.01). This finding implied that students’ identification with their major may be related to their academic self-efficacy. Further, the correlation coefficient between professional identification and learning engagement was 0.759, indicating a strong positive relationship (p < 0.01). This implied that there is a positive relationship between students’ identification with their major and their learning engagement. Finally, the correlation coefficient between academic self-efficacy and learning engagement was 0.761, which also showed a strong positive correlation (p < 0.01). This result suggested that the stronger a student’s academic self-efficacy, the deeper his/her learning engagement is likely to be. This finding was consistent with the existing literature on the positive correlation between academic self-efficacy and learning engagement.

Table 2 Descriptive statistics and correlation matrix for each variable.
Full size table

Further, the relationship between the sub-dimensions of professional identity (fit, behavioral, affective, and cognitive) and the sub-dimensions of academic self-efficacy (academic behavioral efficacy, and academic competence self-efficacy) was analyzed (as shown in Table 3). The results showed that there were generally significant positive correlations between these variables, all of which reached statistical significance (p < 0.01). Among the sub-dimensions of professional identity, the correlation coefficients of behavioral with the other three dimensions (fit, affective, and cognitive) exceeded 0.7, showing a strong correlation. Similarly, among the sub-dimensions of academic self-efficacy, the correlation coefficient between academic behavior self-efficacy and academic competence self-efficacy was 0.768, which also indicated a strong positive correlation between them. In addition, the correlation between these sub-dimensions and learning engagement was also of interest. Learning engagement had the highest correlation coefficient with behavioral (0.753), followed by fit (0.688), suggesting that learning engagement was more strongly associated with these dimensions. Precisely, the behavioral dimension reflected students’ actual actions and efforts in learning, which was highly consistent with the direct manifestation of learning engagement. In contrast, the adaptive dimension reflected students’ ability to adapt and adjust to the learning environment, which was also closely related to learning engagement. In addition, the strong correlations between learning engagement and academic behavior self-efficacy (0.697) and academic competence self-efficacy (0.731) further indicated that academic self-efficacy played an important role in promoting students’ learning engagement.

Table 3 Correlations between sub-dimensions.
Full size table

The relationship between professional identity and learning engagement: a parallel mediation model

After conducting the initial analyses of this study, we found a significant correlation between the variables examined, which triggered the attention of this study to the issue of potential covariance. In order to ensure the accuracy and reliability of further effect tests, the predictor variables were standardized in this study prior to the formal test, and the data set was also diagnosed for covariance. The results of the covariance diagnostics indicated that the Variance Inflation Factor (VIF) values for all the predictor variables ranged from 2.056 to 3.742, which were significantly below the thresholds 5 that is commonly used to indicate serious covariance problems. This indicated that there were no serious covariance problems in the data-set used in the present study, ensuring that the data were suitable for further mediation effect tests62.

Next, this study utilized the Process plug-in developed by Hayes to assess the path of impact of professional identification on learning engagement and the parallel mediating effects of academic competence self-efficacy and academic behavior self-efficacy in this process. This process used a bootstrap method (sample size of 5000) to determine 95% confidence intervals (CI), aiming to accurately assess the statistical significance of the mediating effect. Through the developed chain mediation model, we provided insights into how the sense of academic competence self-efficacy and the sense of academic behavior self-efficacy mediate the relationship between professional identity and learning engagement. The results of the model were presented in detail in Table 4.

Analysis of regression model coefficients

Four multiple regression linear models, corresponding to Models 1 through Models 4, were demonstrated in Table 4.

  1. 1)

    Model 1: i.e., the effect of the four sub-dimensions of professional identity on learning engagement. Except for affective professional identity, the other three sub-dimensions of professional identity showed a significant positive influence relationship on learning engagement, with the highest influence coefficient of behavioral professional identity (B = 0.395,p < 0.001), and the lowest influence coefficient of fit professional identity (B = 0.174,p < 0.001), and the hypotheses H1, H1a, H1b, and H1d were all Validated;

  2. 2)

    Model 2: i.e., the effect of the four sub-dimensions of professional identity on academic behavior self-efficacy. A similar pattern to that of Model 1 was presented, except for affective professional identity, the remaining three sub-dimensions of professional identity also showed a significant positive influence relationship on the academic behavior self-efficacy, and with the highest level of influence being behavioral professional identity (B = 0.253,p < 0.001), and the lowest level of influence being fit professional identity (B = 0.087,p < 0.001);

  3. 3)

    Model 3: i.e., the effect of the four sub-dimensions of professional identity on academic competence self-efficacy. All four sub-dimensions of professional identity showed a significant relationship of influence on the academic behavior self-efficacy, with affective professional identity showing a significant negative relationship (B = −0.047,p < 0.001), and the highest and lowest coefficients of the other three positive relationships being fit professional identity (B = 0.317,p < 0.001) and cognitive professional identity respectively (B = 0.148,p < 0.001);

  4. 4)

    Model 4: i.e., the effects of the four sub-dimensions of professional identity and the two sub-dimensions of academic self-efficacy on learning engagement. With the addition of academic behavior self-efficacy and academic competence self-efficacy, fit professional identity, behavior identity, and cognitive professional identity still showed a significant positive influence relationship, but the effect values were all reduced. Whereas, unlike in Model 1, affective professional identity showed a positive significance (B = 0.040,p < 0.001). The difference between the results of Model 4 and Model 1 predicted the existence of possible mediating effects that affected the path coefficients between the independent and dependent variables.

Analysis of mediating effects

Further, we conducted bootstrap analysis of mediating effects and impact coefficients to determine whether there were significant mediating effects in each path. As shown in Tables 5 and 6, Path 5: affective = > academic behavior self-efficacy = > learning engagement showed a non-significant mediating effect. And Path 6: affective = > academic competence self-efficacy = > learning engagement showed a masking effect (i.e., the direct effect positively predicted learning engagement, but negatively reduced the total effect through the mediating effect). Except for Path 5 and Path 6, the remaining six mediating paths showed significant partial mediation effects, and the direct, indirect, and total effects were all significant positive influence relationships. Among them, path 2 had the highest mediating effect as a proportion of the total effect, at 46.3%. The highest total effect was in the influence path of behavioral-learning engagement (B = 0.395,p < 0.001) and the lowest total effect was in the influence path of affective-learning engagement (B = 0.022,p > 0.01). In Table 5, except for the non-significant path 5: affective = > academic behavior self-efficacy = > learning engagement, the highest percentage of mediating effect was in path 2:

fit = > academic behavior self-efficacy = > learning engagement, and the lowest percentage of mediating effect was in path 1: fit = > academic competence self-efficacy = > learning engagement. At this point, hypotheses H2, H2a, and H2b were all validated.

Table 4 Parallel mediation model: regression coefficients for each variable.
Full size table
Table 5 Mediated effect percentage.
Full size table
Table 6 Summary of inter-mediation test results.
Full size table

The relationship between professional identity and learning engagement: a chain mediation model

Based on the parallel mediation in Sect. 3.3, we further investigated the relationship between the two mediating variables. Using Plug-in 6 in Process, we conducted chained mediated effects analysis through four multiple regression linear models, with the independent variables being the four sub-dimensions of professional identity, the mediator variables being academic competence self-efficacy and academic behavior self-efficacy, and the dependent variable being learning engagement. The results were shown in Table 7. Since Models 1, 2, and 4 were all the same as in Table 4, only Model 2 added academic competence self-efficacy as a dependent variable, so only Model 2 was shown in the table below.

Table 7 Chained mediation model: partial regression coefficients.
Full size table

From Table 7, we could see that there was a significant positive relationship between academic competence self-efficacy and academic behavior self-efficacy (B = 0.435,p < 0.001), revealing a possible chain mediation effect. Further, we performed bootstrap analysis on the 4 chained mediated effects and the results were shown in Table 8. Among the 4 chained mediated effects, BootLLCI and BootULCI did not include 0 between them, indicating that chained mediated effects were all present63. At this point, hypothesis H3 was verified.

Table 8 Bootstrap analysis of the mediation effect test.
Full size table

Discussion

The influence of professional identity on learning engagement

The findings of this study revealed that, with the exception of affective identity, the remaining three sub-dimensions of professional identity exerted a significant positive influence on learning engagement. Among these, behavioral professional identity demonstrated the highest influence coefficient, while fit professional identity exhibited the lowest influence coefficient. This was overall similar to the conclusions reached by existing studies, but there were slight differences.

Firstly, many studies have confirmed that fit professional identity is positively correlated with university students’ learning engagement. A study of trainee nurses at Fujian Medical University in China has found a significant positive correlation between learning engagement, resilience, and professional identity. Nurse trainees with stronger professional identities who perceive themselves to be matched to their specialty are more likely to remain in nursing jobs in their future career plans. These students exhibit a greater inherent interest in their chosen field and demonstrate an enhanced capacity to transform stress and negative experiences into opportunities for personal growth and self-improvement when faced with challenges. Consequently, they have a higher level of learning engagement to their studies64. Another study focusing on undergraduate students majoring in tourism at 10 large public universities in Jiangsu Province, China, further corroborated that students with a strong sense of professional identity are more inclined to perceive a positive alignment between their personal growth and their chosen field of study. This alignment fosters positive learning emotions and behaviors, ultimately enhancing their level of learning engagement65. These students have a stronger sense of belonging to and reliance on their majors, who are willing to work hard to complete their well-developed study plans66. They are more interested in the development of their personal professionalism and have a higher opinion of themselves in terms of communication. High level of professionalism makes them involuntarily value their profession67. It can be seen that the higher the university students’ fit professional identity, the more conducive to stimulate their positive learning emotions, and thus improve their learning engagement.

Secondly, it has also been established that behavioral professional identity is positively related to students’ learning engagement. Nursing students who possess a robust professional identity are better equipped to coordinate the resources and competencies of an inter-professional team to address patient needs, while engaging in active reflection on their individual roles within the team. This collaborative dynamic enables the student body to learn from one another, thereby enhancing their academic performance68. Similarly, normal university students with higher levels of professional identity are able to use teamwork purposefully and are more likely to pay attention to whether the internships and programs offered by their schools could actually contribute to their professional growth and practice, as well as to reflect on their behaviors in their professional learning. This group of students is willing to invest more time and resources in finding an improvement bill for the next enhancement that promotes more effective learning69. Whether they are nurses or normal students, the higher their behavioral professional identity, the more inclined they are to practice and reflect in teamwork and promote deep learning.

Finally, relevant studies also have confirmed that cognitive professional identity is positively correlated with students’ learning engagement. A survey on normal physical education students has figured out that strengthening and enhancing students’ implicit cognition of their major is conducive to mobilizing their interest in study and work, so that they can maintain a relatively full state of knowledge reception in day-to-day professional learning and internship, and thus improving their learning engagement level70. Students who possess a deeper understanding of the current circumstances surrounding normal university students and the teaching profession are more inclined to dedicate time and effort to acquiring professional knowledge and honing their teaching skills71. These students are able to understand their own professional skills and have a more detailed plan for their future development, which promotes their enthusiasm and interest in professional knowledge learning72. The higher the cognitive professional identity of university students, the more inclined they are to plan and prepare for future development.

However, in our study, the effect of the affective professional identity sub-dimension of professional identity on learning engagement is not significant. This is somewhat different from the conclusions drawn from existing studies33. This may be related to our research participants. Many Chinese students, even if they lack familiarity with or interest in a particular major, strive to fulfill the corresponding academic requirements to achieve high grades, secure scholarships, and ensure successful graduation. Nevertheless, for the sake of university students’ mental well-being, their affective professional identity during the learning process must not be overlooked.

Parallel mediating effects of academic competence self-efficacy and academic behavior self-efficacy

The results of this study indicated that, with the exception of Path 5 and Path 6, the remaining six mediating paths demonstrated a significant partial mediation effect. Furthermore, the direct, indirect, and total effects all exhibited significant positive relationships. This suggests that academic competence self-efficacy and academic behavior self-efficacy function as independent partial mediators between the three sub-dimensions of professional identity (excluding affective professional identity) and learning engagement. It was worth noting that in Path 5 and Path 6, affective professional identity, while positively and significantly influencing learning engagement through the direct effect, did not produce a positive mediated influence effect through the mediating variable (with a non-significant mediating effect in Path 5 and a negative masked mediating effect in Path 6). These findings were consistent with the basic ideas of self-concept theory and self-efficacy theory.

Firstly, the two sub-dimensions of academic self-efficacy play parallel mediating roles between fit professional identity and learning engagement. Studies have confirmed that, students with low fit professional identity are prone to indulge in a depressive state of learning. They tend to believe that all behaviors are beyond individual control and they lack concrete actions to change the status for the better, and their academic self-efficacy is reduced73. These students tend to believe that they are unable to perform their daily academic tasks and fulfill their required student obligations. They have a lower likelihood of completing their learning program. Their learning adaptability in various learning environments is reduced, leading to their lower level of learning engagement74. In addition, students with lower fit professional identity tend to believe that the effort they put into their majors is unnecessary and unworthy. They advocate negative self-regulation strategies, lack the regulatory thinking of “plan first, then act”, have lower levels of action control, and have lower academic behavior self-efficacy, which in turn affects their level of learning engagement75.

Secondly, the two sub-dimensions of academic self-efficacy play parallel mediating roles in the relationship between behavioral professional identity and learning engagement. It has been established that students with a high level of behavioral professional identity demonstrate an enhanced ability to adapt more quickly to their surrounding learning environment. These students are able to consciously take ownership of their learning and regulate their behavior to solve problems in a purposeful and strategic manner. They perceive themselves as capable and experienced in mastering the uncertainties of learning and life, and improve their academic competence self-efficacy. They spend more time on academic tasks, which makes it easier for them to achieve academic success76. In addition, students with a higher behavioral professional identity are willing to participate in activities held in their major and are able to look at their personal future plans with a developmental perspective. In the course of their professional studies, they tend to help themselves become more aware and clear about the resources and strengths available to them. This adaptive capacity is accompanied by enhanced academic behavior self-efficacy, characterized by greater autonomy over their professional development and an increased willingness to assume responsibility for their own learning77. They are able to incorporate critical thinking to refine and categorize their learning content, and actively perceive and master the management of resources or rules in their learning environments, which in turn improves their engagement in learning78. Conversely, students with low academic behaviors self-efficacy do not like to participate in school academic and social activities such as engaging in online discussions and asking questions to peers and teachers. They have low levels of learning engagement79.

Finally, the two sub-dimensions of academic self-efficacy act as parallel mediators of cognitive professional identity and learning engagement. Students with a high sense of professional identity have a more comprehensive and clear understanding of the nature of their major as well as clearer goals for their future personal work plans. They tend to believe that they can accomplish challenging, difficult and complex tasks and their sense of belonging to the profession is stronger80. These students are more confident in accomplishing their learning tasks. They are also able to allocate their personal resources more effectively when faced with challenging learning situations, which in turn increases their level of learning engagement81. Although many studies have explored the mediating role of academic self-efficacy in the process of professional identity and learning engagement, their analysis and elaboration are relatively broad. Our research presents a more detailed and comprehensive understanding of the mechanism of this impact. University workers and related personnel can better promote university students’ learning engagement by improving their academic competence self-efficacy and academic behavior self-efficacy.

Chain mediation effect of academic competence self-efficacy and academic behavior self-efficacy

This study found that all four chain mediation paths showed significant mediation effects, indicating that all four sub-dimensions of professional identity influenced learning engagement through the chain mediation effects of academic competence self-efficacy and academic behavior self-efficacy (in which the chain mediation path of affective professional identity-learning engagement had a negative effect close to 0, and the rest of the sub-dimensions suggest a significant positive effect). Related studies have confirmed that academic self-efficacy plays a mediating role in the influence mechanism of professional identity and learning engagement. Students with low professional identity tend to perceive themselves as incapable of accomplishing tasks and are unable to regulate their state and have less self-control. They do not adapt well to professional learning, which in turn makes them reluctant to put in the appropriate effort82. When faced with academic burnout, this group is more likely to indulge in it, not believing that they can complete challenging tasks and not interested in exploring optimal learning modes. Their academic self-efficacy is reduced and they are easily disturbed by negative emotions such as anxiety and depression, which in turn affects the development and implementation of their learning decisions83. However, few studies have clearly divided academic self-efficacy as a mediating variable into two sub-dimensions and explored its influence mechanism on professional identity and academic engagement. This is a more comprehensive exploration done in our study, and it is also consistent with the view of existing research.

It has been confirmed that there is a correlation between the two sub-dimensions of academic self-efficacy. Students with high levels of academic self-efficacy are more likely to perceive themselves as capable of organizing and executing a range of learning behaviors necessary to achieve specific learning goals, and to be able to actively use independent learning strategies and experience gained from collaborative work to overcome academic challenges. They support their own self-confidence and autonomy by striving to implement positive beliefs about their ability to succeed in a course, from cognition to action. These qualities profoundly influence their approaches to learning, empowering them to take ownership of their learning goals and proactively pursue academic achievement84. This group of students believe in their ability to organize and execute the actions required by the established study plan, and show more active participation as well as a hardworking outlook. As a result, they are also able to demonstrate greater perseverance and endurance in the face of difficulties, and are not willing to give in easily to difficulties and challenges. They are able to actively use self-regulated learning strategies (including time and learning environment management), actively seek available learning and social resources, and further enhance their own sense of belief in learning and intrinsic motivation. Inadvertently, they increase the amount of time and energy they devote to learning85. University workers and related personnel can improve students’ academic competence self-efficacy through relevant measures, help them establish self-confidence, believe that they are capable to perform learning tasks, and thereby improve their academic behavior self-efficacy. Students believe that they can adopt appropriate methods to achieve their learning goals, and then stimulate their enthusiasm and initiative in learning, thus promoting their learning engagement.

Contributions, limitations and prospects

Contributions

University students’ learning engagement is still a hot topic that has attracted extensive attention from scholars in various countries. This study explored the effects of the dimensions of professional identity on university students’ learning engagement and the mediating role of academic self-efficacy in this process. It was found that (1) except for affective professional identity, all the sub-dimensions of professional identity showed a significant positive influence on learning engagement; (2) all sub-dimensions of professional identity except the 2 pathways of affective professional identity had positive partial mediating effects through the sub-dimension of academic self-efficacy; and (3) the sub-dimension of academic self-efficacy acted as a chain mediator in the effects of the sub-dimensions of professional identity on learning engagement.

There are a lot of studies on university students’ learning engagement, such as the discussion of factors affecting university students’ learning engagement86,87, and many of them analyze the influence mechanism of university students’ learning engagement in online teaching88,89,90. In addition, many studies have discussed the influence of professional identity on academic self-efficacy91,92and university students’ learning engagement16,93. However, there are fewer studies on academic self-efficacy and university students’ learning engagement94,95. Even fewer studies have examined the relationship between the three simultaneously. In this study, we also explore the relationship between different dimensions in professional identity and different dimensions of academic self-efficacy and university students’ learning engagement. Specifically, compared with previous studies, we discussed the mechanism of university students’ learning engagement from different sub-dimensions of professional identity and academic self-efficacy. Therefore, this study facilitates university workers and related personnel to understand the influence mechanism of Chinese university students’ learning engagement more comprehensively and deeply.

On the one hand, this study enriches the theoretical research on university students’ professional identity, university students’ academic self-efficacy and university students’ learning engagement. On the other hand, this study can provide valuable insights for university workers and related personnel to improve learning engagement among Chinese university students, particularly when considering the unique characteristics of this student population. Over the years, China’s university entrance examination has played an important role in selecting talents and has been widely concerned by the society. As China’s higher education system has transitioned into a stage of massification, coupled with the limited availability of high-quality educational resources, the competition in the university entrance examination remains intensely fierce. Consequently, despite recent advocacy for quality-oriented education, exam-centric education practices are likely to persist for an extended period. Many senior high school students dedicate immense effort to preparing for the university entrance examination, often at the expense of developing their comprehensive competencies. Due to their limited participation in extracurricular activities, they have little knowledge of themselves and the careers they want to pursue in the future. Most of these students fill in the major of the university entrance examination are suggested by their parents or others, and some choose to fill in the major with the majority with the primary goal of gaining admission to prestigious “985”, “211” good university. However, they don’t have a clear understanding and preparation for what major they want to learn and how to plan their university life after they go to university. Combined with the above analysis and the findings of this study, we can promote the learning engagement of Chinese college students from the following aspects.

First, to improve the professional identity of university students. Career counseling for university students should be strengthened before they fill out their volunteer forms. It has been found that candidates who fill out their volunteer positions have significantly higher levels of professional identity and academic self-efficacy than those who apply for admission or transfer under the advice of others96. Students are unfamiliar with and do not understand their majors and are confused about the prospects of their majors; they have a low sense of identity with their majors, which in turn affects their learning engagement. Therefore, it is particularly important for the relevant departments and parents and others to strengthen counseling for students to fill in their university volunteers in the university entrance examination.Secondly, to enhance the prestige of the profession, universities should prioritize the advancement of academic discipline construction through rigorous research and elevate the professional development of faculty members. By emphasizing scholarly research and maintaining high teaching standards, universities can bolster their academic reputation, thereby fostering students’ sense of identification with their chosen fields. Lastly, it is imperative to provide comprehensive guidance for students’ career planning. At the beginning of the school year, Universities can help students to have a preliminary plan for their professional studies by introducing them to the basic information of their majors, professional training programs, and the scope and direction of employment. Universities can also help students gain a comprehensive understanding of their professional studies by offering lectures and career guidance classes, so that they can have a clearer picture of their career development71,97. At the same time, schools should also strengthen students’ career guidance. Relevant staff of universities and parents should help students to understand the employment prospects of their majors, difficulties in the employment process and employment paths, etc., so as to eliminate students’ confusion about the future of their majors and thus enhance their sense of professional identity.

The second is to improve university students’ academic self-efficacy. Firstly, pay attention to individual differences and provide targeted guidance. Due to different family environments, own learning experiences and mastery of learning methods, there are differences in the learning levels of different students98. Teachers can set different learning goals so that students at different levels can accomplish them within their abilities, thereby increasing their academic self-efficacy. Secondly, teachers, parents and other significant others should provide timely and positive feedback on students’ learning as much as possible, and give students more encouragement and praise, thus to help students gain a sense of accomplishment and satisfaction in the learning process. Thirdly, Universities can reduce students’ anxiety about learning by conducting psychological lectures and academic counseling, guiding them to positive psychological construction, improving their academic self-efficacy, and thus increasing their level of learning engagement.

In addition, in our study, although affective professional identity is not significantly correlated with learning engagement, stakeholders should still try to improve students’ affective professional identity as much as possible through counseling, lectures, and improving the teaching level of professional teachers. Although students are able to complete their learning tasks as much as possible even if they don’t like or are not interested in their majors, they are prone to negative emotions such as learning burnout, anxiety and even depression, which is not conducive to the healthy development of their body and mind in the long run.

Limitations and prospects

First, this study relied on cross-sectional data and convenience sampling methods, which may have limited the representativeness of the sample and thus the general applicability of the findings. As a result, the correlations revealed in the results may not be sufficiently reflective of what is actually happening in the wider population. In addition, the nature of cross-sectional studies limits the possibility of a deeper understanding of causality, as it captures data only at a single point in time and does not allow for tracking the dynamics between variables over time. Second, there is a risk of selection bias due to the limitations of the study design. Participants may have specific characteristics or tendencies (for example, our samples are mainly from Chinese university students), which may affect the genera ability and reliability of the findings. And this study fails to use an experimental or intervention design, which limits the exploration of possible causal and mediating effects between variables. Finally, due to time and effort constraints, whether the mechanisms influencing professional identity, academic self-efficacy, and learning engagement differed across demographic variables such as gender, grade level, and major are not explored in this study. To overcome these limitations, a longitudinal design could be considered for subsequent studies to track and analyze variable changes over time. At the same time, implementating of a randomized controlled trial will help to identify and validate the causal relationship between these variables more accurately. In addition, expanding the sample size and using randomized sampling methods will enhance the study’s representativeness and external validity. Follow-up studies can also conduct differential analysis of the influence mechanism. These improvements will provide a stronger foundation for a deeper understanding and explanation of the relationships among the study variables.