According to the Subjective Well-Being (SWB) Theory proposed by Diener (1984), well-being is a multifaceted psychological construct consisting of two core components: the cognitive evaluation of life satisfaction and the affective experience of positive and negative emotions. This theory posits that individual traits and cognitive processes serve as proximal predictors of happiness, as people’s subjective judgments of life quality are inherently shaped by their personality, self-esteem, and emotional regulation abilities. Previous empirical studies have provided consistent evidence for the theory, confirming that internal factors such as psychological resilience and self-compassion are strongly correlated with higher levels of subjective well-being among college students (Arimitsu & Hofmann, 2015; Baumeister et al., 2003).

Complementing the individual-focused perspective of SWB Theory, Ecological Systems Theory emphasizes that individual development and psychological outcomes are shaped by nested environmental systems (Bronfenbrenner, 1979). This framework categorizes environmental influences into microsystems (e.g., family parenting styles, peer relationships), mesosystems (e.g., interactions between school and family), and macrosystems (e.g., cultural contexts). Drawing on this theory, external factors such as social support and school environment are hypothesized to exert critical impacts on college students’ happiness by shaping their daily experiences and social interactions. Prior research has validated the link between environmental factors and well-being, showing that positive family and social contexts contribute to enhanced psychological and social well-being among young adults (Hou et al., 2020; Wang & Wang, 2018).

Taken together, the Subjective Well-Being (SWB) Theory and Ecological Systems Theory provide a dual explanatory framework for understanding college students’ well-being: the former highlights the role of individual psychological characteristics and cognitive evaluations, whereas the latter emphasizes the influence of family, school, peer, and broader sociocultural environments. However, because meta-analysis requires conceptually comparable and operationally consistent coding categories, it is necessary to further clarify the connotations and boundaries of the core constructs included in this study. Based on classic definitions in general psychology and insights from established theoretical and empirical literature, this study systematically categorizes individual internal factors into distinct dimensions. Specifically, cognition refers to the mental processes through which individuals acquire, process, store, retrieve, and use information (Eysenck & Keane, 2015). In the present study, sensation and perception are treated as two important dimensions of cognition because they represent fundamental components of information processing (Goldstein, 2016; Peng, 2019, p. 83–174). Personality-related internal factors are operationalized into four dimensions: ability (Peng, 2019, p. 407), motivation (Peng, 2019, p. 333; Ryan & Deci, 2000a), character (Peterson & Seligman, 2004; Peng, 2019, p. 452), and values (Schwartz, 2012). To ensure consistency in conceptual operationalization and adherence to research norms, this study explicitly defines the operational definitions for each dimension and the relevant themes included in the meta-analysis, as presented in Table 1.

Goal of the Study. The aim is to examine the relative effects of internal (individual traits and cognitive mechanisms) and external (social support, school factors, family environment) factors on college students’ happiness, and to explore whether cultural context and literature type moderate these relationships. Figure 1 illustrates the proposed analytical framework for our meta-analysis hypotheses.

Hypothesis 1: Internal factors (e.g., personality traits, psychological resilience) have significantly larger effect sizes on college students’ happiness than external factors (e.g., social connection, school structural factors).

Hypothesis 2: The influencing factors of social well-being show greater cultural heterogeneity, whereas those of subjective well-being are more cross-culturally consistent.

Hypothesis 3: Publication type exerts a weak moderating effect on the relationship between influencing factors and college students’ happiness.

This systematic review and meta-analysis was conducted and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) statement (Page et al., 2021). The search for English-language literature of this study covers internationally recognized databases such as Web of Science, ScienceDirect, Springer, and Scopus. The search was conducted using keywords such as “college students’ happiness” and “college students’ well-being,” with search strategies adjusted according to the characteristics of each database. This study conducted a correlation-based quantitative synthesis meta-analysis to examine the association between influencing factors and college students’ happiness. The meta-analysis was performed using a random-effects model strategy determined by heterogeneity test results, with publication bias assessed via multiple complementary methods. We used a diverse search term combination strategy to adapt to the characteristics of different databases. Through the China Knowledge Network Journal Database, Wan Fang Database, and Wei Pu Journal and Dissertation Database, we searched the Chinese literature using the search terms “college students” and “well-being of college students”. The publication period of the studies included in the meta-analysis was limited to January 2020 to April 2024 to ensure that the analyzed literature reflected recent research. In addition, a secondary retrieval was conducted by means of citation backtracking to enhance the comprehensiveness of the literature.

The study screened the literature based on the following inclusion criteria: (1) the research topic focuses on the influencing factors related to the well-being of undergraduate college students; (2) the study population is limited to college students, excluding specialists and graduate students, and excluding the same study published repeatedly; (3) the type of the study is a quantitative study, with a sample size of no less than 30; and the data in the literature are complete; (4) the study contains complete calculations of the effect value required for the data, such as sample size, correlation coefficients, or t-values and F-values that can be transformed into correlation coefficients. These data were converted into correlation coefficients according to the method proposed by Glass et al. (1981). Two researchers independently reviewed the screening and inclusion process of this study. The final screening of controversial literature was carried out by reading the abstracts and full texts. After rigorous screening, a total of 39 papers were included, including 36 academic journal papers and 3 master’s and doctoral dissertations, covering the effect values of 97 independent samples. The 39 studies included in the meta-analysis are presented separately in Supplementary Material. The PRISMA flow diagram of searching literature is detailed in Figure 1.

Two specially trained graduate students independently performed the screening and coding of each document. Any disagreements that arose during the coding process were resolved through discussion and consensus to ensure consistency and accuracy of coding. The final coding results were reviewed and confirmed by the corresponding author. For studies that met the inclusion criteria, we coded key study characteristics, including author and publication date, sample size, correlation coefficient, publication type, and type of subject sample. This coding information provided an exhaustive database for this study, which helped to analyze and compare the characteristics and results of the studies in depth. The publication type and subject sample of the literature were used as potential moderating variables in this study to explore the possible impact of these factors on the results of the study of college students’ well-being. Publication types were categorized into core journals (CJ) and general journals (GJ). Core journals usually refer to journals with high influence and recognition in the academic community, while general journals refer to other academic journals. The types of subject groups were categorized into Chinese university students (CUS) and foreign university students (FUS) based on the country of origin.

Comprehensive Meta Analysis3.0 CMA3.0 software was used as an analytical tool, and the sample size (N) and correlation coefficients (r) from the selected literature were entered into CMA3.0 for calculation. The correlation coefficient r was chosen as the effect size to characterize the extent to which the relevant factors affect the well-being of undergraduate college students. In this study, a meta-analysis of Pearson’s product-difference correlation coefficient r was used to calculate the effect value (Hofmann et al., 2005), standardized the correlation coefficient r through the transformation of Fisher’ Z, and the corresponding weights were calculated according to the sample size. The corresponding weights and 95% confidence intervals were calculated based on the sample size. According to Cohen (1988)’s criteria, effect sizes ≤0.10 were considered small, 0.10 < r < 0.40 medium, and r ≥ 0.40 large.

The heterogeneity test is a key step in meta-analysis, and heterogeneity is inevitable due to the differences in the included literature in terms of study samples, publication time, experimental conditions, etc. (Wang et al., 2021). The heterogeneity test is mainly assessed by Q statistic and I² statistic. Higgins et al. (2003) suggested that a random effects model should be chosen when I² > 75% and a fixed effects model when 0 ≤ I² ≤ 75%. In addition, if the heterogeneity is not significant or the number of studies K is small (K < 5), a fixed-effects model should be used for meta-analysis.

To assess publication bias, we utilized the funnel plot distribution of effect sizes for qualitative tests and the Egger regression coefficient for quantitative tests. The significant Egger regression coefficients indicate possible bias (Begg & Mazumdar, 1994; Sterne & Egger, 2001). In addition, a test was conducted using a fail-safe N coefficient to determine how many unpublished studies would be needed to make the overall effect size insignificant or to make the p-value no longer statistically significant.

In this meta-analysis, we included a total of 39 studies covering 97 effect sizes and involving 56,524 subjects. These studies included master’s theses as well as studies in Chinese and English, including 9 studies in Chinese and 30 studies in English. Individual factors included 58 effect sizes (sample size = 33,885), social factors included 28 effect sizes (sample size = 17,315), and school factors included 11 effect sizes (sample size = 5324).

The results of the analysis showed that the overall Q value of 9282.1 and the I² value of 98.9%, which is much more than 75%, indicate that there is significant heterogeneity among the findings. Therefore, it is more reasonable to use a random effects model. In addition, we tested for heterogeneity for each dimension under individual factors, social factors, and school factors. The results showed that the level of heterogeneity for all factors exceeded 75%, further confirming the high degree of heterogeneity in the findings. This result suggests the need to further explore possible moderating variables to explain the source of this heterogeneity. The details are shown in Table 2.

To verify the robustness of the findings, this study implemented a sensitivity analysis. First, we compared the overall effect sizes of the random effects model (r = 0.265) with the fixed effects model (r = 0.278). In addition, a stability test was conducted for each literature by the one-by-one elimination method, which showed that the value of the effect size r fluctuated between 0.257 and 0.278. The results of the combined sensitivity analyses indicated that the overall effect size of this study was relatively stable.

The present meta-analysis indicates that individual-level factors are significantly associated with college students’ well-being, with cognitive and personality-related factors showing particularly important roles. This finding is consistent with previous research suggesting that Personality is one of the most stable and effective predictors of well-being, and it plays a greater role in determining well-being than other influences (Diener et al., 1999). Personality can be understood as a relatively stable configuration of behavioral tendencies and psychological characteristics that shapes how individuals relate to themselves, others, and their surrounding environment during social adaptation. It involves multiple dimensions, including ability, temperament, character, needs, motivation, and values, and represents a distinctive psychosomatic organization gradually formed through socialization (Shi, 1988).

Through the above results, it can be seen that among the individual factors, feeling, motivation, ability, and values play a crucial role in college students’ well-being. Among them, personal feelings especially significantly and positively affect college students’ well-being. This includes the sense of sincerity and the sense of meaning in life; the sense of sincerity refers to an individual’s attitude of accepting past, present, and possible future events. If college students are able to face stressful or negative events in their studies, life, or relationships with an open mind, it will help them to adopt more effective coping strategies (Harlianty et al., 2022), which will enhance their well-being experience. People of different ages pursue different meanings in life, and college students are in the important period of pursuing their dreams and realizing the value of life. If self-actualization is successful, it will correspondingly enhance their happiness experience (Jin et al., 2016). Both the experiences gained in the face of the past, present, or future, as well as holding a relaxed or positive and optimistic attitude in the face of stressful events, can contribute to college students’ well-being.

In addition, the motivational elements of personality explored in this study, such as hope motivation and hedonic motivation, showed moderate correlations with college students’ well-being. Hope is a positive state based on the interaction of the path to success and motivation (Snyder et al., 2000, 2002). Different psychologists have offered different interpretations of well-being, often distinguishing between hedonic well-being, which emphasizes pleasure and life satisfaction (Diener, 1984), and eudaimonic well-being, which emphasizes meaning, self-realization, and personal growth (Ryan & Deci, 2000a). From this perspective, different motivational orientations may be associated with different forms of well-being, thereby leading to complex relationships among motivational factors and college students’ happiness. Indeed, individuals’ pursuit of goals often depends on their life values, with some believing that material abundance is happiness, while others believe that spiritual pleasure can lead to fulfillment, even in the context of material deprivation. The findings of this study support the idea that higher levels of hope are associated with higher levels of self-satisfaction (Bailey et al., 2007). When students face negative emotions or stressful events, emotion regulation ability (Chen et al., 2022), psychological resilience, and empathy (Sanchez et al., 2018; Tran et al., 2022) help them maintain well-being.

So, are college students happy or not? In higher education, a student’s self-identity has both inhibition and promotion functions (Liao & Xiao, 2022). Self-identity refers to an individual’s experience after subjectively evaluating his or her own inner ability, cognitive world, and life development state (Oyserman et al., 2012). When students strive for self-actualization, those with higher self-assessments usually set stricter requirements. If these goals are successfully achieved, their overall well-being will subsequently increase; conversely, this may lead to mental health problems, which may reduce their overall satisfaction.

In addition to internal factors, external conditions may also shape individual well-being This study examined social and school-level factors, among which social factors emerged as important contributors to students’ well-being, including social connection, social support, and social media use. Of these, social connection had the greatest impact on college students, followed by social support and social media use.

From a social-relational perspective, social relationships involve interactions among individuals, groups, and broader social contexts. Social connection, social support, and social media embody just about every aspect of social relationships. Individuals who form strong bonds are more likely to be aware of their proximity to others. They are also more willing to engage in group activities. As a result, their interpersonal trust, self-esteem, psychological health, and fulfillment tend to increase (Baumeister & Leary, 1995). The college years are a time of transition when a range of skills are needed to manage interpersonal relationships, cope with complex environments, and participate in a variety of activities. Social support can alleviate the negative consequences of stress while meeting basic social needs (Huang & Pu, 2024). Furthermore, social support provides a sense of belonging, security, and self-worth (Cobb, 1976), ultimately contributing to the enhancement of well-being. Currently, there are two hypotheses about the relationship between Internet use and college students’ well-being. One suggests that online interactions undermine subjective satisfaction because users invest time in virtual space in lieu of real-life communication. The other suggests that online interactions are beneficial in enhancing real communication and compensating for the barriers of real-life face-to-face interactions, which in turn enhances well-being (Valkenburg & Peter, 2007). Compared with social connection and social support, social media use showed a weaker but still significant positive correlation with college students’ well-being. The reason for this connection might be that online media provides a platform and opportunity for stressful students to interact. Through social media, students may obtain positive interactional experiences, which can contribute to their well-being.

In this study, the analysis of the moderating effects on college students’ well-being revealed that publication type had significant moderating effects on individual factors (including ability, perception, motivation, and values), social factors (especially social connection), and school factors (e.g., academic performance). Among these, the moderating effects on ability and perception were particularly noteworthy. This phenomenon may stem from the limited number of the master’s thesis included in the analyses, resulting in significant differences between publication types. General and core journals both demonstrated significant moderating effects on the personality factors of motivation and values. However, general journals had higher combined effect values than core journals. This difference suggests that there may be some instability in the research conclusions drawn from different journal types. This instability may be related to sample size or the choice of research methodology. The literature included in this study covered the last five years and was affected by the COVID-19 pandemic. As a result, the literature was likely predominantly cross-sectional in nature. These studies may limit the extent to which the findings can be generalized to broader contexts.

The moderating effects of different subject groups on individual factors (including perception, motivation, and personality) and social factors (especially interpersonal interactions) were analyzed in depth. The analysis results show that different subject groups have significant moderating effects on individual factors such as perception, motivation, and personality. Specifically, in terms of perceptual factors, studies involving foreign college students mainly focused on perceived stress, whereas studies involving Chinese college students mainly examined perceived control. As a result, the combined effect values of the two showed significant differences: foreign college students’ perceptions were dominated by negative effects, whereas Chinese college students’ perceptions were dominated by positive effects. This finding is consistent with previous evidence that higher perceived stress is associated with lower life satisfaction (Matheny et al., 2008; Rezaei & Mousanezhad Jeddi, 2020). Conversely, stronger perceived control is associated with higher levels of well-being (Matheny et al., 2008). Subject group also significantly moderated the association between motivation and well-being. This significance may be related to the limited sample size of foreign university students involved in motivation research, as is the case with personality factors. Social factors, particularly interpersonal interactions, demonstrated significant differences. This significance is largely attributed to the focus on negative interpersonal relationships within the Chinese college student sample, which negatively impacted well-being. Conversely, the foreign college student sample was mainly investigated from the perspective of positive interpersonal relationships, and thus the two showed significant differences among different groups of subjects.