Introduction

Over the past two decades, many East Asian countries have implemented policies aimed at attracting international students, establishing these countries as educational hubs, and increasing their global recognition [1]. In Taiwan, the government has also actively promoted the internationalization of higher education over the past twenty years by encouraging foreign students to pursue studies in Taiwan. This strategy aimed to enhance Taiwan's visibility in the global higher education market and to foster international relations through multicultural exchanges, thereby influencing the learning experiences of international students [2].

According to Taiwan's Ministry of Education, 21,386 international students were enrolled in undergraduate programs during the 2019–2020 academic year; this figure included 10,727 males and 10,659 females [3]. The learning motivation of these international students is crucial, as it influences both their academic performance and their future employment outcomes. In social psychology and education, motivation is a key factor influencing human behavior [4, 5].

Motivation, as a complex psychological construct, explains behavior and the effort invested in goal-directed activities [6,7,8]. It is a crucial factor in enhancing students’ academic achievement, as high levels of learning motivation significantly influence their success in education and the attainment of academic goals [9,10,11]. Understanding the various dimensions of motivation can provide deeper insights into how it drives academic success and enables educators to effectively support students in their learning process.

Most social‒cognitive models assume that students'motivation is influenced by classroom interactions, activities, reinforcement practices, and cultural factors [12]. Motivation beliefs can be categorized into several dimensions, including intrinsic motivation, extrinsic motivation [13], task value, control beliefs, self-efficacy, and test anxiety [13]. Intrinsic motivation refers to situations in which individuals engage in an activity due to the inherent satisfaction that it offers, in which context individuals are driven by internal rewards and view the activity itself as enjoyable. Conversely, extrinsic motivation refers to situations in which individuals engage in an activity with the goal of achieving an external outcome [8, 13].

In this study, the Motivated Strategies for Learning Questionnaire (MSLQ) was utilized, categorizing motivation into three components. The Value Component includes intrinsic goal orientation (motivation driven by curiosity or mastery), extrinsic goal orientation (focus on external rewards such as grades), and task value (evaluation of a task's interest and utility). The Expectancy Component encompasses control of learning beliefs (the perception that effort influences outcomes) and self-efficacy (confidence in completing tasks successfully). The Affective Component focuses on test anxiety, which negatively impacts performance and includes worry (negative thoughts) and emotionality (physiological arousal) [14].

Students who have learning goals are typically driven by intrinsic motivation, as they seek to learn and understand the content of their education, whereas students who have performance goals are more likely to be motivated extrinsically due to their desire to achieve the highest level of academic performance. Intrinsic motivation drives essential growth-oriented behaviors, such as embracing challenges, honing skills, and actively exploring personal interests [13]. According to [15], intrinsic and extrinsic motivations are both crucial for engagement in various activities, including learning. Extensive research on these two types of motivation has provided valuable insights into developmental and educational practices.

Among the various motivational variables, self-efficacy consistently emerged as the strongest and most reliable predictor of academic achievement, highlighting motivation as a key determinant of students'success in school [16, 17]. Self-efficacy refers to an individual’s confidence in their ability to organize and execute the actions necessary to achieve desired outcomes [16, 17]. In contrast, test anxiety presents a more complex dynamic: while moderate levels of anxiety can enhance learning motivation, excessive anxiety is detrimental, impeding students'ability to learn effectively and achieve optimal performance [18].

Oxford and Shearin's [19]study on language learning identified six key factors that influence motivation in the context of language acquisition. These factors include attitudes, which encompass perceptions of the learning community and the target language; beliefs regarding oneself, which include attitudes toward success, self-efficacy expectations, and anxiety; goals, which refer to the clarity and relevance of learning objectives as well as the student’s reasons for learning; involvement, which measures the extent to which the student participants actively and consciously in the learning process; environmental support, which includes support from teachers and peers as well as the integration of cultural and extracurricular elements into the learning experience; and personal attributes, such as ability, age, gender, and previous language learning experience [19].

Studies have reported that students'motivation differs by gender and major, particularly in the context of mathematics and English learning [20]. Motivation is a key affective variable that influences learning success, especially when combined with factors such as language acquisition [21, 22]. Numerous researchers have reported that in the affective domain, certain variables, including motivation, have strong impacts on student learning. Therefore, a great deal of previous educational research has focused on student motivation [23]. Thus, motivation is a critical factor in both mathematics and language learning. Tüysüz et al. [9] reported that university students exhibit higher levels of motivation for science than do high school students, as university students generally exhibit a high level of positive motivation toward science. This situation may be due to the fact that students choose their field of study at the university level and often achieve good scores in science courses, thus leading them to adopt positive attitudes toward science, whereas high school students may exhibit lower levels of motivation in this context.

Furthermore, gender is also an important factor in learning, as females tend to exhibit more positive attitudes and higher levels of motivation toward language learning [24]. Studies that have investigated gender and English learning have revealed that students'motivation for learning English differs across different genders and majors [20]. Baker and MacIntyre [25] and Sung and Padilla [26] reported that females exhibit significantly higher levels of motivation than do males. Cigan [27] also indicated that respondents generally exhibit high levels of learning motivation, particularly with regard to intrinsic motivation. Statistically significant differences have been observed in terms of gender, the average number of middle school credits, the student’s number of years of study, and the types and intensities of motivation. However, some studies have reported conflicting results. Abu-Rabia [28] did not observe gender differences in this context, and other studies on gender differences in English learning motivation have indicated the absence of statistically significant differences in attitudes and motivation between males and females.

Accordingly, it is evident that some studies have reported significant gender differences, whereas others have not. Most previous studies on this topic have focused on gender differences in learning motivation rather than on the ways in which learning motivation varies among international students across different academic years and colleges.

However, the previously mentioned studies did not specifically examine these demographic variables related to learning motivation among international university students. In particular, international students who pursue higher education abroad are more likely to engage in fields of study that they intentionally choose. Accordingly, international students may exhibit distinct configurations of learning motivation at the university level, which may lead to differences in their motivational profiles.

Moreover, research on the learning motivation of international students underscores its pivotal role in shaping their academic experiences and achievements. For example, a study conducted at a private educational institution in Singapore found that international students are more inclined to rely on external regulation and intrinsic motivation compared to their local peers. This observation highlights the nuanced interplay between cultural contexts and motivational orientations within diverse educational settings [29]. Similarly, research conducted in mainland China revealed a significant correlation between learning motivation and the academic outcomes of international students [30]. These findings collectively emphasize the critical importance of understanding and addressing the dynamic nature of motivation among international students. Therefore, the objectives of this study are as follows:

1. To investigate the current state of learning motivation among international university students and to analyze the corresponding differences in learning motivation in light of various demographic variables.

2. To analyze the predictive power of different demographic variables with regard to the learning motivation of international students.

Methodology

Research framework and hypotheses

This study involved a quantitative survey approach that focused on international students at a university in central Taiwan with the goal of investigating their current levels of learning motivation. The study examined differences in learning attitudes among international students across different demographic variables (gender, grade, and college affiliation). The demographic variables of international students were included in as independent variables, whereas learning motivation was included as the dependent variable. The relationships among these variables were analyzed, and the predictive power of the independent variables with regard to the dependent variable was subsequently assessed.

In addition to its role as a descriptive investigation, the study is guided by the following hypotheses:

1. Hypothesis 1: International students exhibit significant differences in learning motivation in light of different demographic variables.

2. Hypothesis 2: Different demographic variables significantly predict the learning motivation of international students.

Research instrument

The Motivated Strategies for Learning Questionnaire (MSLQ) was employed as the research instrument in this study [14]. The MSLQ is a widely recognized, publicly available tool that can be used to assess students'motivational orientations and learning strategies. For the purposes of this study, the motivational items of “Motivation Scales” contained in the MSLQ were selected, which included 31 items that were extended across six dimensions, i.e., intrinsic goal orientation (Cronbach's alpha = 0.74), extrinsic goal orientation (Cronbach's alpha = 0.62), task value (Cronbach's alpha = 0.90), control of learning beliefs (Cronbach's alpha = 0.68), self-efficacy for learning and performance (Cronbach's alpha = 0.93), and test anxiety (Cronbach's alpha = 0.80) [14]. In this study, the Likert scale was defined as follows: 5 = very positive, 4 = positive, 3 = neutral, 2 = negative, and 1 = very negative.

Moreover, the construct validity of the “Motivation Scales” of MSLQ scales was assessed by conducting the confirmatory analyses by Pintrich, Smith, Garcia, and McKeachie [14]. The construct validity of the Motivation Scales of the MSLQ was assessed through confirmatory analyses. The results indicated (X2/df = 3.49, GFI = 0.77, RMR = 0.07, CN = 122).While the goodness-of-fit indices are not ideal, they remain reasonable considering the broad range of courses and subject domains covered by the scales. The models exhibit sound structural properties, supporting the claim of factor validity for the scales. Motivational attitudes and the application of various learning strategies may vary depending on course characteristics, teacher expectations, and individual student attributes [14].

Additionally, the evaluation of Motivation Items (constructs) using standardized factor loadings (LX estimates) revealed varying degrees of correlation and construct stability. Intrinsic Goal Orientation showed moderate correlations, with factor loadings ranging from 0.55 to 0.69. Extrinsic Goal Orientation exhibited similar variability, ranging from 0.44 to 0.71. Task Value demonstrated strong construct stability, with high factor loadings such as 0.88 and 0.86. Control Beliefs about Learning displayed a broader range of factor loadings, from 0.38 to 0.84, reflecting moderate to high correlations. Self-Efficacy for Learning and Performance showed robust construct stability, with factor loadings between 0.63 and 0.89. Lastly, Test Anxiety revealed varied performance across items, with factor loadings spanning from 0.42 to 0.88. These findings affirm the robustness and reliability of the MSLQ scales in assessing motivational constructs across diverse educational settings [14].

Sample

A total of 260 questionnaires were collected (response rate = 89.35%), 210 of which represented valid responses, for an effective sample rate of 80.77%. Descriptive statistical analysis was conducted to investigate the demographic characteristics of the valid sample. The sample included 127 males (60.5%) and 83 females (39.5%). The participants were distributed across different academic years as follows: 50 freshmen (23.8%), 34 sophomores (16.2%), 46 juniors (21.9%), and 80 seniors (38.1%). In terms of college affiliation, 72 students were recruited from the College of Business and Management (34.3%), 40 from the College of Science and Engineering (19.0%), and 98 from the College of Humanities (46.7%). These demographic characteristics provide a solid foundation for analyzing the research questions investigated in this study.

Process

As part of this study, a survey was conducted from March 1 to June 30, 2021. The participants in this study received detailed information regarding the survey, including information indicating that the research findings would be submitted to an international academic journal and that all the data would be anonymized. Participation in this research was voluntary, and participants were allowed to choose whether to complete the questionnaire. After the participants provided consent, the survey was administered in a fully anonymous and deidentified manner to ensure participants’ confidentiality.

Data analysis methods

Statistical methods were employed in the data analysis conducted for this research. Percentages were used to describe participants’ background characteristics, whereas means and standard deviations were used to detail their current state of learning motivation. Independent samples t tests were conducted with the goal of assessing gender differences in learning motivation, and a one-way analysis of variance (ANOVA) was performed to examine differences among students across different academic years and college affiliations. Finally, stepwise multiple regression analysis was employed to evaluate the predictive power of these demographic variables with regard to students'learning motivation.

Results

Results of t-test and ANOVA

Table 1 presents the descriptive statistics for various dimensions of learning motivation among international students. The results indicate that intrinsic goal orientation had the highest mean score (M = 3.768, SD = 0.518), followed by extrinsic goal orientation (M = 3.532, SD = 0.684). In contrast, the dimensions of control of learning beliefs (M = 2.394, SD = 0.581) and test anxiety (M = 2.480, SD = 0.417) showed comparatively lower mean scores. Overall, the students'learning motivation levels were found to be moderate, with an aggregate mean score of M = 2.978 (SD = 0.175). These findings are descriptive in nature and do not imply statistical significance or causal relationships among the dimensions of learning motivation.

Table 1 Summary of the descriptive statistics
Full size table

The results of the independent samples t test, which are presented in Table 2, indicated significant gender differences in certain dimensions of learning motivation. Although no significant differences were observed in intrinsic goal orientation, extrinsic goal orientation, task value, or test anxiety, females obtained significantly higher scores than did males in terms of the control of learning beliefs (M Female = 2.500, SD = 0.677, t = −2.156, p = 0.032, d = −0.304), and self-efficacy for learning and performance (M Female = 2.938, SD = 0.461, t = −2.564, p = 0.011, d = −0.362) overall learning motivation (M Female = 3.014, SD = 0.165, t = −2.484, p = 0.014, d = −0.347).

Table 2 Summary of the results of the t test
Full size table

The results of the one-way ANOVA conducted for this research, which are presented in Table 3, reveal significant differences in certain dimensions of learning motivation across different grade levels. Although no significant differences were observed in intrinsic and extrinsic goal orientation, task value, or control of learning beliefs, significant differences were observed in self-efficacy, test anxiety and overall learning motivation. Importantly, first-year students reported significantly lower levels of self-efficacy (F = 3.426, p = 0.018, ηp2 = 0.048) and lower levels of test anxiety than did their peers (F = 4.025, p = 0.006, ηp2 = 0.058); in particular, fourth-year students exhibited the highest levels of overall motivation (F = 8.261, p < 0.001, ηp2 = 0.017).

Table 3 Summary of the results of the one-way ANOVA by grade
Full size table

The results of the one-way ANOVA, which are presented in Table 4, indicate no significant differences across college affiliations in most dimensions of learning motivation, including intrinsic and extrinsic goal orientation, task value, self-efficacy, test anxiety, and overall learning motivation. However, a significant difference was observed in the control of learning beliefs, in which context students from the College of Science and Engineering obtained significantly higher scores than did students from the College of Humanities (F = 4.064, p = 0.019, ηp2 = 0.038).

Table 4 Summary of the results of the one-way ANOVA by college affiliation
Full size table

Results of the stepwise multiple regression analysis

The results of the stepwise multiple regression analysis are presented in Table 5, 6, and 7. The stepwise multiple regression analysis revealed significant predictors with regard to the dependent variables of control of learning beliefs, self-efficacy for learning and performance, test anxiety, and overall learning motivation. Gender was identified as a significant predictor for both the control of learning beliefs (β = 0.148, p = 0.031, f 2 = 0.023) and self-efficacy (β = 0.175, p = 0.011, f 2 = 0.032), accounting for 2.2% and 3.1% of the variance in these factors, respectively. Additionally, grade level was identified as a significant predictor of test anxiety (β = 0.176, p = 0.011, f 2 = 0.032), accounting for 3.1% of the variance in this factor. The variance inflation factor (VIF) values in all the models were close to 1, thus indicating that the absence of any issues with multicollinearity in the analysis.

Table 5 Control of learning beliefs: Stepwise regression summary
Full size table
Table 6 Self-efficacy for learning and performance: Stepwise regression summary
Full size table
Table 7 Test anxiety: Stepwise regression summary
Full size table

Discussion

The findings of this study provide valuable insights into the factors influencing learning motivation among international students, and several key observations emerge from the data analysis.

The findings of this study indicate the presence of significant gender differences in learning motivation, which is consistent with the conclusions of previous research [20, 24,25,26,27],. Females exhibit higher levels of self-efficacy, control of learning beliefs, and overall learning motivation than do males [31, 32]. The results of this study are consistent with those reported in previous studies. However, those results are different from the findings reported by Saxena, Wright, and Khalil [33]and Abu-Rabia [28], who indicated the absence of any differences between males and females with regard to scales of motivation, attention, and time management, although females obtained significantly lower scores than males. Nonetheless, Gayef, Çaylan, and Temiz [31] highlighted the fact that while Aung et al. reported the absence of statistically significant differences in intrinsic motivation levels between male and female students, Kusurkar et al. revealed that females exhibit intrinsic and extrinsic motivation at higher intensity. Brouse et al. reported that females obtained higher scores than did males with respect to all measures of intrinsic motivation; these results stand in contrast to the findings of the present study, in which context males obtained higher scores. Conversely, Sobral reported that male students obtained higher extrinsic motivation scores, which is in line with our results [31]. Despite the variability in the findings reported in these different studies, the current study highlights the need for targeted interventions aimed at increasing motivation among male students, particularly in the areas of control of learning beliefs, and self-efficacy. These findings reveal the importance of gender-sensitive educational strategies, as gender has a significant effect on learning motivation, as highlighted in previous research.

Moreover, this study identified significant gender differences in self-efficacy and learning beliefs, which may be attributed to several factors. First, according to Bandura’s theory of self-efficacy [17], an individual’s self-efficacy is shaped by past experiences and social comparisons. Gendered societal expectations often influence learning behaviors, with females encouraged to prioritize relational and process-oriented goals, while males are more likely to focus on achievement-oriented objectives [34]. Such societal roles may account for females exhibiting stronger process-oriented learning beliefs and males displaying greater confidence in goal attainment.

The academic environment and field of study may also contribute to these disparities. Research indicates that females in male-dominated disciplines, such as STEM fields, often encounter additional challenges due to their underrepresentation, potentially impacting their self-efficacy [35]. Furthermore, Deci and Ryan [36] suggests that females often derive intrinsic motivation from the perceived value and meaningfulness of learning, whereas males are more likely driven by extrinsic motivators such as performance outcomes or competition. These psychological and contextual differences may help explain the observed gender disparities in self-efficacy and learning beliefs.

The results of the one-way ANOVA conducted as part of this study reveal significant differences in learning motivation across students from different grade levels and college backgrounds. Specifically, first-year university students exhibited lower levels of self-efficacy for learning and performance as well as lower levels of test anxiety than did third- and fourth-year students. However, our findings differ from those reported by Gayef, Çaylan, and Temiz [31], who indicated the absence of any significant differences in intrinsic motivation to accomplish tasks, intrinsic motivation to experience stimulation, extrinsic motivation introjection, and extrinsic motivation external regulation across different grade levels (p > 0.05). Gayef, Çaylan, and Temiz [31] also noted that their findings were inconsistent with those of several previous studies, such as Brouse et al., who reported that both intrinsic and extrinsic motivation decreased as the number of years spent by the student in college increased [31]. Furthermore, these authors reported that high school students obtained both higher average extrinsic motivation scores and higher average motivation scores than did university students, a result that warrants further consideration [31]. Although these various studies have reported different findings, the results obtained by this study highlight the need for grade-specific strategies in this context. This study revealed that grade level significantly influences students'test anxiety. Specifically, throughout students’ academic careers, their test anxiety and self-efficacy tend to increase. Over time, students develop better coping mechanisms and study habits, which can reduce anxiety and enhance their motivation. However, the decline in self-efficacy observed among first-year students highlights the need for ongoing support and engagement strategies that can help these students sustain their motivation throughout their academic journey. Institutions should implement targeted interventions with the goal of preventing declines in first-year students'engagement and performance.

The study revealed notable differences between first-year and fourth-year students in terms of self-efficacy and learning beliefs. These differences can be understood through the lens of accumulated learning experiences and the varying nature of academic pressures at different stages of higher education. Fourth-year students, having adapted to the university environment and completed more specialized coursework, are likely to possess greater self-efficacy and a more developed understanding of their learning goals [17]. In contrast, first-year students may still be acclimating to new learning environments and lack the confidence that comes with academic and experiential growth.

Additionally, the nature of academic pressures changes throughout the university journey. First-year students often grapple with transitional challenges, such as adapting to novel academic and social contexts, while fourth-year students face career-related and graduation pressures [37]. These differences in stressors may shape their learning beliefs and self-efficacy. Curriculum design also plays a significant role. Introductory courses in the first year are typically foundational, offering lower levels of challenge, whereas advanced courses in the final year emphasize applied learning and research, which may foster higher levels of confidence and belief in one’s abilities [36].

Although college affiliation was not observed to have significant impacts on most dimensions of learning motivation in this study, it did lead to statistically significant differences in the control of learning beliefs, particularly among students in the College of Science and Engineering. This finding is in line with that reported by Richardson et al. [38], who indicated that students in science, technology, engineering, and mathematics (STEM) fields often exhibit higher levels of intrinsic motivation due to the structured and outcome-oriented nature of their studies. These findings highlight the importance of considering academic discipline when addressing motivational issues, as students from different disciplines may be influenced by different motivational drivers. Specifically, students at the College of Science and Engineering exhibited stronger control of learning beliefs than did their peers at the College of Humanities. These findings suggest that the nature of the academic discipline in question may impact students’ motivation, in which context certain fields are associated with higher levels of both intrinsic and extrinsic goal orientation.

These insights have valuable implications for educational policymakers and educators, particularly with respect to the design of targeted learning support and interventions that can address the unique needs of students across different genders and grade levels. Female students and students in lower grades exhibit distinct motivational profiles, thus highlighting the need for targeted interventions. This situation emphasizes the complex interactions demographic backgrounds and motivation, thus suggesting that tailored educational strategies could enhance learning outcomes by addressing the specific needs of different student groups.

Moreover, the findings of the stepwise multiple regression analysis highlight the significant roles played by gender and grade level in predicting key academic constructs. Gender was revealed to be a significant predictor of both control of learning beliefs and self-efficacy, thus suggesting that gender differences may influence students'confidence in their learning ability. The significant negative relationship observed between grade level and test anxiety indicates that students in higher grades experience less anxiety, which may be due to their higher levels of academic experience and increased use of coping strategies. Additionally, the combined influence of gender and grade level on overall learning motivation highlights the importance of considering these demographic factors in the process of developing educational interventions. The lack of multicollinearity strengthens the validity of these findings and suggests the existence of robust relationships among the variables.

Although the findings of this study are valuable, they may be limited due to the focus of this research on a single university in central Taiwan. Additionally, the cross-sectional design of this study restricts our ability to observe changes over time.

Moreover, this study acknowledges that the R2 value across the regression models is relatively small, indicating that the predictors included in the analysis explain only a limited proportion of the variance in the dependent variable. This suggests that the predictors analyzed may play marginal roles compared to other potential predictors not considered in this study. Future research should aim to identify and incorporate additional variables that might account for a greater proportion of the variance, thereby enhancing the explanatory power of the models.

To address these limitations, future researchers should consider conducting studies across multiple regions or countries with the goal of increasing the generalizability of the results. Furthermore, longitudinal research designs could be used effectively to track changes in motivation across different academic years. Expanding the scope of the sample and including additional variables could also provide a more comprehensive understanding of learning motivation among diverse student populations. Finally, the findings of this research could inform the development of targeted interventions and educational policies aimed at sustaining student motivation throughout their academic journey.

Conclusion

This study presents an in-depth analysis of the factors influencing learning motivation among international students and reveals several key findings. First, significant gender differences were observed in learning motivation, in which context female students generally exhibited higher levels of self-efficacy, control of learning beliefs, and overall learning motivation. These results are consistent with those of some previous studies but differ from those of other such studies, thus highlighting the variability in research findings as well as the importance of developing gender-sensitive educational strategies.

Second, significant differences in learning motivation were also observed across grade levels and college affiliations. First-year international students exhibited lower levels of self-efficacy than did first- and fourth-year students in general, whereas senior international students exhibited higher levels of overall motivation. Additionally, this study revealed that students in the College of Science and Engineering exhibited significantly higher levels of control of their learning beliefs than did studies in the College of Humanities; this finding primarily reflects the structured and outcome-oriented nature of STEM disciplines. These results highlight the need to take disciplinary differences into account when designing educational strategies, as students from different academic fields may be influenced by distinct motivational drivers.

Furthermore, stepwise multiple regression analysis provided further support for the influence of demographic variables on learning motivation, particularly with respect to the ability of gender, grade level, and college affiliation to predict students'learning motivation. These findings suggest that future educational interventions should consider these demographic factors, especially gender and grade level differences, in their efforts to enhance students'academic performance and psychological well-being. This study revealed that demographic factors such as gender and grade level are crucial predictors of students'learning beliefs, self-efficacy, and test anxiety. Educational strategies should be tailored to address these differences with the goal of enhancing self-efficacy and managing anxiety across students from different grade levels. The findings of this research can deepen our understanding of how demographic characteristics influence academic outcomes, thereby providing a foundation for future research and practical applications in educational settings. This study also provides a foundation based on which future researchers can explore additional factors that may influence students'learning behaviors with the aim of improving the effectiveness of educational practices further.