The demographic characteristics of the 527 participants encompassed various factors such as age, gender, ethnicity, religion, marital status, residential location, academic level, educational background, institution types, academic year and students types (i.e., employed vs pure studetns).
Cut-off scores were established to categorize participants' self-efficacy levels, ranging from 'very low' to 'very high,' providing valuable insights into the distribution of self-efficacy within the sample. In our sample, we found a very low self efficacy (15.56%), low self-efficacy (18.41%), average self efficacy (22.96%), high level of self efficacy (22.58%) and very high level of self efficacy (20.49%). Notably, a substantial proportion of participants exhibited 'high' and 'very high' levels of self-efficacy, indicating a predominantly positive perception of one's abilities among college students in our study context.
In our study, genaral self-efficacy served as a non-significant predictor of GPA, indicating minimal contribution to academic achievements. Consistently, in another study with college students. Fenning and May (2013) found no significant interrelationships between general self-efficacy and college GPA. However, they found that general self-efficacay was associated with high school GPA. Similarly, another study also revealed no significant relationship between self efficacy and GPA in university students (Ramos-sánchez & Nichols, 2007). Conversely, Yip (2012) found the self efficacy as a significant predictor of students’ GPA, indicating that high academic achievers differed significantly from low academic achievers in the level of self-efficacy. Galyon et al. (2012) found that academic self-efficacy was the second strongest factor explaining college GPA following closely behind standardized test scores, while combining predictors. Self-efficacy demonstrated a positive correlation with academic motivation, suggesting that higher self-efficacy is associated with elevated levels of academic motivation (Shrestha et.al, 2021). In a study, motivation was found to be significant predictor of academic performance (Yip, 2012). This suggests that self-efficacy can generate academic motivation which in tern contributes to academic performance. Furthermore, students with elevated levels of self-regulation, self-efficacy, and self-efficacy for self-regulation were less likely to displaying procastination in academic behavior; however, this relationship is very weak (Bhusal, 2023).
We found no significant relationship between age and general self-efficacy. The correlation between age and GPA was also non-significant. However, a significant weak negative relationship was observed between age and percentage. The regression analysis confirmed the significance and identified age as a predictor of percentage. However, the casual factor in this relationship was unknown.
Our study revealed no significant difference in general self-efficacy between male and female participants. In line with our findings, other studies also found the difference was not statistically significant between male female (Lindley & Borgen, 2002; Mahat & Pradhan, 2012). Consistently, D’Lima et al. (2014) also reported that there was no significant gender-by-ethnicity interaction observed for academic self-efficacy.
We observed gender as non-predictor of GPA which is consistent with finding that no significant gender differences were observed in academic performance (Busch, 2006). According to Busch (2006), female students outperform their male counterparts except for statistics. They also reported that the gender differences in self-efficacy, a construct central to the study of business administration, were found to be small. They further explained that female students exhibited significantly lower self-efficacy in computing and marketing, while demonstrating higher self-efficacy in statistics compared to their male counterparts (Busch, 2006). It is noteworthy that possessing higher self-efficacy in statistics does not seem to be correlated with obtaining higher scores compared to males. Inconsistently, a main effect for gender was observed, indicating that female university students had higher GPAs than male students (D’Lima et al., 2014).
We observed ethnicity showing no significant influence on self-efficacy levels. DeFreitas (2012) also found that there was no significant relation between ethnicity and self-efficacy. However, D’Lima et al. (2014) mentioned that African Americans and Caucasians reported significantly higher levels of self-efficacy compared to Asian American students.
Our study revealed Janajati ethnicity as a significant predictor of GPA which is consistent again with the finding that ethnicity had significant effect on GPA (DeFreitas, 2012). Furthermore, they demonstrated that individuals with higher self-efficacy, particularly among European Americans, exhibited higher GPAs (DeFreitas, 2012).
We discovered that religion was as potential predictors of both general self-efficacy and GPA. Religion exhibited a significant association with both general self-efficacy and GPA, implying a potential link between religious affiliation, self-efficacy, and academic achievement.
Marital status did not significantly influence either general self-efficacy or GPA, suggesting its limited role in determining self-efficacy or academic performance among college students.
Residential location comparison between valley and non-valley residents revealed no significant difference in the level of general self-efficacy or GPA scores, indicating geographical location's negligible impact on both general self-efficacy and academic performance.
Analysis of education levels showed no significant differences in general self-efficacy level. However, a significant predictive relationship was found between education level and overall GPA, suggesting that academic performance in higher education level were associated with lower overall GPAs. A study in college student found that the high school GPA was best predicted by general self-efficacy, while college GPA was most strongly associated with self-efficacy for learning, but not with general self efficacy (Fenning & May, 2013). The correlation results revealed a significant positive correlation between general self-efficacy and high school GPA. However, no significant interrelationships were observed between general self-efficacy and college GPA (Fenning & May, 2013).
Furthermore, no significant differences were observed in the level of general self-efficacy among students in various educational faculties, although faculty of management showed a significant association with higher GPA.
Lastly, no significant differences in the level of general self-efficacy was observed among students attending different types of institutions, while students attending private colleges tended to have higher GPAs compared to those in government colleges.
We found statistically non-significant difference among year of studies in the level of general self-efficacy. In line with our findings, a study reported that each ethnic group showed consistent self-efficacy across the semester (D’Lima et al., 2014). Similarly, we also found that academic year distribution did not significantly predict GPA. Consistently, a study reported that self-efficacy did not significantly change over time (Ramos-sánchez & Nichols, 2007).
We observed being employed or a pure student did not influence on general self-efficacy. A study found no significant differences in self-efficacy levels based on the income (Ramos-sánchez & Nichols, 2007). We also found income as a non-predictor of GPA outcomes significantly. However, a study found that individuals with higher income levels exhibited higher SAT math scores (DeFreitas, 2012).
The reliability analysis demonstrated a robust internal consistency with a high Cronbach's alpha coefficient (α = 0.78). Schwarzer et al. (1995) reported the internal consistency ratings for each of the five samples examined indicated high reliability, with alpha values ranging from 0.82 to 0.93. In a sample of 991 migrants from what was then Germany, the retest reliability over a two-year period was 0.47 for men and 0.63 for women (Schwarzer et al., 1995). The generalised self efficacy scale demonstrated strong internal consistency with a Cronbach's alpha of 0.83 in a Colombian Sample (Juarez & Torres, 2008).
Additionally, the correlation of individual items with the total scale ranged from r = 0.53 to r = 0.67, indicating a moderate level of convergent validity. Schwarzer et al. (1995) assert that concurrent validity is supported by significant correlations with other tests. Positive correlations were observed with self-esteem (0.52), internal control belief (0.40), and optimism (0.49). Conversely, negative correlations were identified with general anxiety (-0.54), performance anxiety (− 0.42), shyness (− 0.58), and pessimism (− 0.28). Predictive validity was assessed in a one-year follow-up of East German migrants. For women, self-efficacy positively correlated with later self-esteem (0.40) and optimism (0.56). However, men showed less impressive correlations (0.20 and 0.30) over a two-year period (Schwarzer et al., 1995). Additionally, a study reported correlations between items and the total scale ranging from 0.3 to 0.66 (Juarez & Torres, 2008).
Limitations
The survey research, while insightful, has limitations. The sample size of 527 participants and the specific context may limit generalizability. The cross-sectional design restricts causal inference, and reliance on self-report measures introduces potential bias. The authors focus on GPA as the sole academic measure oversimplifies student success. Cultural nuances and unknown moderating factors could impact the observed relationships. Opportunity sampling may potentially introduce selection bias. Clarification of causal links and consideration of additional variables could enhance the study's practical implications.