Self Efficacy


and Academic Performance


Author: Emily Gardner
Date: November 12, 2014


The purpose of this study was to explore the predictive relationship between self-efficacy and academic performance in light of possible third variables (i.e. academic entitlement, ethnicity, gender, class status). Previous studies have found that there is a positive linear relationship between self-efficacy and academic performance and also that self-efficacy positively predicts academic performance. Additionally, the reviewed literature has alluded to the effect that other possible third variables have had on both self-efficacy and academic performance. A total of 58 undergraduate and graduate students from a small, liberal arts university in north Texas completed a survey consisting of demographic questions and a seven-item Likert type measure of self-efficacy. The results revealed no significant predictive relationship between the predictor variables of self-efficacy, ethnicity, or gender and academic performance. Additionally, the sample means for ethnicity, gender, and class status did not differ significantly. The results were discussed in light of the literature review and areas for future research were suggested.

Keywords : self-efficacy, academic performance, academic entitlement, college students

Another Look at Self-Efficacy and Academic Performance

In this paper, I will discuss the current reviewed literature on the relationship between self-efficacy and academic performance among college students in light of possible third variables that may affect this relationship. I will then state my hypotheses and my method in conducting this study. I will present my results and then discuss them, returning to the reviewed literature.

Self-Efficacy and Academic Performance in the Literature



Bandura defined self-efficacy (SE) as the confidence individuals have in their ability to organize and execute courses of action required to attain specific performance outcomes (as cited in Jot, Usher, & Bressoux, 2011; Lane & Lane, 2001; Lent et al., 2003). SE has been extensively studied in the literature in relation to various activities in which a persons performance can be measured. One such area of interest has been students performance in school.

Self-Efficacy and Academic Performance

In the past 15 years, many studies have shown that a relationship between SE and academic performance (AP) exists. Previous studies have found that, among undergraduate as well as postgraduate college students, SE and AP were positively and moderately correlated (Galyon, Blondin, Yaw, Nalls, & Williams, 2012; Klomegah, 2007; Lane & Lane, 2001; Richardson, Bond, & Abraham, 2012) and also that SE was a significant and moderately positive predictor of AP (Choi, 2005; Coutinho, 2008; Lane, Lane, & Kyprianou, 2004). Similarly, Chemers, Hu, and Garcia (2001) found that SE was directly and strongly related to AP among  1st-year college students, and Caprara et al. (2008) found that high SE levels among junior high school students contributed to their AP.

Patterns in the Literature: Possible Third Variables

Academic Entitlement. Though it is a newly proposed psychological construct, and there is little literature on the topic, the concept of academic entitlement (AE) among college students has become an increasingly popular topic in psychological research in the past 10 years. AE is defined as the tendency to possess an expectation of academic success without taking personal responsibility for achieving that success (Chowning & Campbell, 2009, p. 982).

Academically entitled students tend to externalize responsibility for academic outcomes (Chowing & Campbell, 2009); in contrast, SE refers to the confidence one has in ones own ability to obtain such outcomes (Lane & Lane, 2001). Thus it is perhaps unsurprising that SE has been found to inversely and moderately predict AE among undergraduate students (Boswell, 2012). This has implications for the relationship between SE and AP; AE viewed as a possible third variable could change the well-established patterns seen in current literature. However, there is a lack of current studies that explore AE as a phenomenon. Other possible third variables are alluded to throughout the literature.

Age. Meta-analyses have revealed that the strength of correlation between SE and APvaried from weakly to strongly positive depending on level of education, which ranged from elementary to college level (Alivernini & Lucidi, 2011; Multon, Brown, & Lent, 2012). The variation seen in the strength of correlation between different age groups in the reviewed literature suggests that level of education and/or age affects either SE or AP or both.

Gender. Previous studies have found no significant gender differences among SE scores(Choi, 2005; Jonson- Reid, Davis, Saunders, Williams, & Williams, 2005); however, there are also studies that contest this claim. Fast et al. (2010) found that females had lower SE scores than males regarding their ability to perform well on a math test. In the same study, math SE was found to be positively and moderately correlated with AP on a standardized math test. This study has implications for the impact of gender on not only SE but also AP since SE has been shown to predict AP.

Ethnicity. Additionally, previous studies have explored how ethnicity and immigrantstatus affect SE and AP among undergraduate college students. Jaret and Reitzes (2009) found that GPA scores were higher among whites than blacks or Asians and that blacks had higher SE scores than Asians or whites. Furthermore, it has been shown that SE and AP vary across immigrant status. Aguayo, Ojeda, Herman, and Flores (2011) found that SE significantly and moderately predicted AP among second generation college students and beyond (i.e. third generation, fourth, etc.). In addition, Jaret and Reitzes (2009) found that immigrant college students had higher GPAs than second or third generation college students. Despite the findings of these studies, there is an overall lack of current research that emphasizes the influence of ethnicity and/or immigrant status on SE or AP to further affirm that such differences exist.

Subject or task-specific self-efficacy. In addition to gender, ethnicity, and immigrantstatus, the reviewed literature also alludes to the effect that task and subject-specific SE have on AP. As stated above, math SE was found to be positively and moderately correlated with AP on a standardized math test (Fast et al., 2010). Likewise, Bruning, Dempsey, Kauffman, Zumbrunn, and McKim (2013) found that writing self-efficacy was moderately and positively correlated with self-reported writing performance among middle school students. Furthermore, Choi (2005) found that specific SE was the only significant predictor of term grades among other kinds of SE measured (i.e. general SE, academic SE). These studies suggest that subject or task-specific SE or maybe even major-specific SE among college students may be more reliable predictors of AP than more general SE measures.

In This Study
The present study explores the relationship between SE and AP, which has been studied extensively in the literature. This study tests the following hypotheses:
H1: SE will significantly and positively predict AP among undergraduate and graduate students.
H2: Ethnicity, gender, and year in college will significantly and positively predict AP among undergraduate and graduate students.
H3: The sample means for ethnicity and gender will differ significantly from each other in regards to GPA.
H4: There are differences among group means for year in college in regards to GPA.

Participants were 58 undergraduate and graduate students at a small liberal arts university in north Texas. There were 27 males and 31 females surveyed. Participants were over 18 years of age and were informed that their participation was voluntary and anonymous. The research procedure was approved by the Institutional Review Board.


The survey (see Appendix A) asked participants to report their sex, class status (freshman, sophomore, junior, senior, or graduate), ethnicity (Hispanic or non-Hispanic), and current GPA (open-ended response which I used to operationalize AP). To obtain a measure of the participants self-efficacy, I used the Perlin Mastery Scale, taken from Pearlin and Schooler (1978), which contained seven items. Each item was answered using a Likert type scale ranging from one to seven, one being Strongly Disagree and seven being Strongly Agree. Items one through five needed to be reverse coded, while items six and seven did not. The sum of the seven items was used to operationalize self-efficacy. The categorical variables (i.e. sex, class status, ethnicity) were treated as interval data and assigned numbers (i.e. sex (female=1, male=2), class status (freshman=1, sophomore=2, junior=3, senior=4, graduate=5), ethnicity (non-Hispanic=0, Hispanic=1).


The survey was administered to undergraduate and graduate students on campus during the week between class times. Each survey took less than five minutes to complete and was immediately collected and kept in an envelope.

To test the first and second hypotheses, SE, ethnicity, gender, and year in college were predictor variables (PVs) in a multiple regression design with GPA as the criterion variable (CV). To test the third hypothesis, two two-sample independent t-tests were run. Sample means were compared between men and women and Hispanics and non-Hispanics, which were the independent variables (IVs), with GPA as the dependent variable (DV). To test the fourth hypothesis, a one way ANOVA was conducted with the sample means compared among years in college (IVs) with GPA as the DV. For each test run, alpha was at .05, and all values below this were seen as significant.


The data was entered into SPSS. I first ran a multiple regression analysis in SPSS to determine what proportion of variation in GPA could be accounted for by SE, gender, class status, and ethnicity. The slope of the linear relationship between the PVs and the CV was not significant, with only 6.6% of variance in GPA accounted for by the PVs (R = .066, F(4, 53)=. 940, p = .448). Class status accounted for most of the variance seen in GPA (= .169,t=1.208,p= .232), followed by gender (= .145, t= 1.049, p= .299), then ethnicity (= -.130, t= -.931, p= .356), and finally SE (= .086, t= .594, p= .555), but none of these results were at or below alpha.

I ran two independent sample t-tests to compare the sample means between males and females and also between Hispanic and non-Hispanics with GPA as the DV.Males (x= 3.244) and females (x= 3.355) did not differ significantly on GPA (t(56)= -1.075, p= .287) and, likewise, neither did [Hispanics (x=3.207) and non-Hispanics ( x=3.316)] (t(56)= .826, p= . 413).

I then ran an ANOVA on class status in order to find any differences among sample means between freshmen, sophomores, juniors, seniors, and graduates. No significant differences were found between years in college in regard to GPA (F(4, 53)= 1.611, p= .185).


The relationship between SE and AP has been studied extensively in the literature. It has already been found by many studies that a relationship between SE and AP exists; however, the interplay of possible third variables with SE and AP is an interesting theme that I wanted to explore and one which is only informally addressed in some of the current literature. There was insufficient support for my hypotheses, and the null hypotheses could not be rejected.

It is interesting that I found no significant predictive relationship between SE and AP since this is inconsistent with several studies within the reviewed literature (Choi, 2005; Coutinho, 2008; Lane, Lane, & Kyprianou, 2004). However, this could be due to my small sample, which was a convenience sample of 58 undergraduate and graduate students. A small sample size means a decrease in the power of my test statistic to detect a significant relationship between variables, and the non-probablistic character of my sample means that it may have been biased. This is a threat to the external validity of my study since the sample tested may not have been representative of the population at large (i.e. college students).

In addition to the non-probablistic character of my sample, AE may have had a role to play as a third variable in the relationship between SE and AP. AE is the tendency to expect academic success regardless of how one performs (Chowing & Campbell, 2009). Jean Twenge (2006) speculates that, since the 1970s, children have been raised to be self-focused and to feel affirmed and entitled in all areas of life, whether it be relationships, sports, or academia, with little or no merit on their part. That would put the current generation of college students among some of the first to be raised with an entitled mindset since birth.

SE has been found to inversely and moderately predict AE among undergraduate students (Boswell, 2012). This could account for the lack of a significant relationship seen in SE and AP. It is possible that students are now relying less on their own ability to succeed and, instead, are externalizing responsibility for their academic performance outcomes, deviating from the typical patterns seen in the current literature. Though I did not include AE in the design of my study, if anyone wishes to explore AE in relation to academic SE or AP, I would suggest testing the linear relationship between these variables to see if they are significantly correlated. This would be an interesting topic for future research.

There was also little ethnic variety among the sample tested. The majority of students that attend UD are of caucasian ethnicity, while the second largest prevalent ethnicity is Hispanic. For this reason, I dichotomized ethnicity to only include Hispanic or non-Hispanic as options, and I looked at differences between sample means in regard to GPA. In addition to UD being ethnically exclusive, I also had an imbalance between number of Hispanics and non-Hispanics in my sample (11:47); however, since UD is an academically rigorous school, it is possible that its admitted students perform similarly academically, regardless of their ethnicity. All of these are threats to the external validity of my study since my sample is not representative of the population at large (i.e. college students). Although I found no significant results, it has been suggested that there are differences among other ethnicities in AP (Jaret & Reitzes, 2009). It would be interesting to see if AP or SE differed among different ethnicities at a larger or more ethnically diverse college campus. For researchers who wish to explore this topic further, I would suggest making ethnicity an IV in two multiple regression designs with SE and AP as CVs.

I also did not take into consideration immigrant status among Hispanics, which has been shown to affect SE and AP among college students (Aguayo, Ojeda, Herman, & Flores, 2011; Jaret & Reitzes, 2009). It would be a worthwhile topic for future research to look at the influence of not only ethnicity but also immigrant status on SE and AP since there is an overall lack of current research that identifies this as a possible third variable. I propose that the best statistical design to test for differences among immigrant statuses would be an ANOVA with immigrant status as the IV consisting of five levels (i.e. immigrant, first generation, second generation, third generation, fourth generation or above) and AP as the DV.

That there were no differences between men and women for GPA is perhaps not a surprising finding since current research has also found no significant gender differences in regard to SE (Choi, 2005; Jonson-Reid, Davis, Saunders, Williams, & Williams, 2005), and SE has been shown to be directly and strongly related to AP among college students (Chemers et al.,2001). However, since SE did vary when a subject, specifically math, was introduced (Fast et al.,2010), this suggests that perhaps AP for a specific subject could also differ between men and women; however, there are studies that contest this and an overall lack of studies in the current literature that emphasize gender as a third variable in the relationship between SE and AP. Although it was not looked at in this study, future researchers who wish to look at gender and AP could see if means for men and women differ significantly from each in regard to GPA.

Although there were no differences found between years in college for GPA, this could be due to the uneven samples of participants from different classes. Freshmen, sophomores, and graduate students were all under 10 participants. Juniors and seniors were larger sizes but still not representative of University of Dallas students or the population at large, which threatens the external validity of my study. The strength of correlation between SE and AP has been found to vary from weakly to strongly positive depending on level of education (Alivernini & Lucidi, 2011; Multon, Brown, & Lent, 2012). This suggests that either different ages or classes could differ significantly in regard to SE or AP or both; this is a worthwhile potential third variable in the relationship between these two variables. I would suggest, for anyone who wishes to replicate this study, that he/she use an equal and substantial amount of participants (i.e. 30 or more) from separate classes using stratified random sampling from a representative college student sample and use an ANOVA design to explore differences between years in college in regard to GPA.

There is a lack of current literature that takes age into account among college students exclusively.

Another limitation of this study was that I did not ask for the major of the participants. It has been found that subject or task-specific SE was moderately and positively correlated with the specific subject or task evaluated among a variety of age groups (Bruning et al., 2013; Choi, 2005; Fast et al., 2010). This suggests that it would be of value to also consider differences between majors in college in regard to SE and GPA if one was interested in studying the effects this would have for subject-specific tests or measures of performance. This is a threat to the internal validity of my study since I am not sure that I am most effectively measuring SE by having participants take a general SE survey as opposed to a subject-specific SE survey according to different majors and in relation to subject-specific AP. For researchers who wish to explore this topic further, I would suggest using stratified random sampling in order to collect data from students of a variety of different majors, pulling from a representative college student sample and looking at differences among group means in two ANOVA designs with subject-specific SE and GPA as DVs.

Though I was not able to reject the null hypothesis with regard to any of my alternate hypotheses, one significant result was found that was not expected. I ran a series of seven two-sample independent t-tests with GPA as the IV, dichotomized between high and low using a mean split, to compare group means on each measure of the SE survey. I found that, for item four ofthe SE measure (i.e. You often feel helpless in dealing with the problems of life), the GPA means differed significantly (t(56)= 3.062,p= .003). Those with higher GPAs (m= 5.11) scored higher on the SE measure for this statement than those with lower GPAs (m= 3.64). This suggests that those students who have a higher measure of AP feel a higher sense of responsibility when it comes to dealing with problems in their lives. This further suggests that having confidence in ones own abilities in difficult situations may lead to higher levels of performance. This would be an interesting topic for future research.

In conclusion, the relationship between SE and AP is a dynamic one, and it is important to take into consideration variables that may affect both (e.g. academic entitlement, the size of the sample tested, ethnic variation, gender, year in college, and many more). Past studies have explored the relationship between SE and AP extensively but taken certain third variables into account rather sparingly. My results show that there is much left to be explored in these areas; especially in the college world where academics are privileged, knowledge of performance-related variables can become a valuable topic of interest.


Aguayo, D., Herman, K., Ojeda, L., & Flores, L. Y. (2011). Culture predicts Mexican Americans' college self-efficacy and college performance. Journal Of Diversity In Higher Education4(2), 79-89. doi:10.1037/a0022504

Alivernini, F., & Lucidi, F. (2011). Relationship between social context, self-efficacy, motivation, academic achievement, and intention to drop out of high school: A longitudinal study.

The Journal Of Educational Research, 104(4), 241-252. doi:10.1080/00220671003728062

Boswell, S. S. (2012). 'I deserve success': Academic entitlement attitudes and their relationships with course self-efficacy, social networking, and demographic variables. SocialPsychology Of Education, 15(3), 353-365. doi:10.1007/s11218-012-9184-4

Bruning, R., Dempsey, M., Kauffman, D. F., McKim, C., & Zumbrunn, S. (2013). Examining dimensions of self-efficacy for writing. Journal Of Educational Psychology, 105(1), 25-38. doi:10.1037/a0029692

Caprara, G., Fida, R., Vecchione, M., Del Bove, G., Vecchio, G., Barbaranelli, C., & Bandura, A. (2008). Longitudinal analysis of the role of perceived self-efficacy for self-regulated learning in academic continuance and achievement.Journal Of Educational Psychology100(3), 525-534. doi:10.1037/0022-0663.100.3.525

Chemers, M. M., Hu, L., & Garcia, B. F. (2001). Academic self-efficacy and first year college student performance and adjustment. Journal Of Educational Psychology, 93(1), 55-64. doi:10.1037/0022-0663.93.1.55

Choi, N. (2005). Self-Efficacy and Self-Concept as Predictors of College Students' Academic Performance. Psychology In The Schools, 42(2), 197-205. doi:10.1002/pits.20048

Chowning, K., & Campbell, N. (2009). Development and validation of a measure of academic entitlement: Individual differences in students externalized responsibility and entitled expectations. Journal Of Educational Psychology, 101(4), 982-997. doi:10.1037/ a0016351

Coutinho, S. (2008). Self-efficacy, metacognition, and performance. North American Journal OfPsychology, 10(1), 165-172.

Fast, L. A., Lewis, J. L., Bryant, M. J., Bocian, K. A., Cardullo, R. A., Rettig, M., & Hammond, K. A. (2010). Does math self-efficacy mediate the effect of the perceived classroom environment on standardized math test performance?. Journal Of EducationalPsychology, 102(3), 729-740. doi:10.1037/a0018863

Galyon, C. E., Blondin, C. A., Yaw, J. S., Nalls, M. L., & Williams, R. L. (2012). The relationship of academic self-efficacy to class participation and exam performance. SocialPsychology Of Education, 15(2), 233-249. doi:10.1007/s11218-011-9175-x

Jaret, C., & Reitzes, D. C. (2009). Currents in a stream: College student identities and ethnic identities and their relationship with self-esteem, efficacy, and grade point average in an urban university. Social Science Quarterly, 90(2), 345-367. doi:10.1111/j. 1540-6237.2009.00621.x

Jot, G., Usher, E. L., & Bressoux, P. (2011). Sources of self-efficacy: An investigation of elementary school students in France. Journal Of Educational Psychology, 103(3), 649-663. doi:10.1037/a0024048

Jonson-Reid, M., Davis, L., Saunders, J., Williams, T., & Williams, J. (2005). Academic self-efficacy among African American youths: Implications for school social work practice.

Children & Schools, 27(1), 5-14. doi:10.1093/cs/27.1.5

Klomegah, R. (2007). Predictors of academic performance of university students: An application of the goal efficacy model.College Student Journal, 41(2), 407-415.

Lane, J., & Lane, A. (2001). Self-efficacy and academic performance. Social Behavior AndPersonality, 29(7), 687-693. doi:10.2224/sbp.2001.29.7.687

Lane, J., Lane, A. M., & Kyprianou, A. (2004). Self-efficacy, self-esteem and their impact on academic performance. Social Behavior And Personality, 32(3), 247-256. doi:10.2224/ sbp.2004.32.3.247

Lent, R. W., Brown, S. D., Schmidt, J., Brenner, B., Lyons, H., & Treistman, D. (2003). Relation of contextual supports and barriers to choice behavior in engineering majors: Test of alternative social cognitive models. Journal Of Counseling Psychology, 50(4), 458-465. doi:10.1037/0022-0167.50.4.458

Multon, K. D., Brown, S. D., & Lent, R. W. (1991). Relation of self-efficacy beliefs to academic outcomes: A meta-analytic investigation. Journal Of Counseling Psychology, 38(1), 30-38. doi:10.1037/0022-0167.38.1.30

Richardson, M., Abraham, C., & Bond, R. (2012). Psychological correlates of university students' academic performance: A systematic review and meta-analysis. PsychologicalBulletin, 138(2), 353-387. doi:10.1037/a0026838

Twenge, J. M. (2006). Generation me: Why today's young Americans are more confident,assertive, entitled--and more miserable than ever before. Simon and Schuster.


New Scholarships Forge Connections, Open Opportunities

At UD, it is now possible to create a nonendowed, named scholarship with a minimum commitment of $20,000, which may be payable over up to four years. Recently, three families have done just this in order to forge a connection and provide needed aid to current students.

+ Read More

Weekly Campus Monitor: Early Testing Success

As in-person classes continue on campus, President Thomas S. Hibbs, Ph.D., BA '82 MA '83, spent time walking the UD Mall, gathering vital student feedback. He reported a joyful response to initial testing success, “especially among the seniors — how much they hope we can keep this going.” There have been only four positive tested cases out of nearly 1,000 tests performed on campus since June, with no active COVID-19 cases as of Friday, Sept. 4.

+ Read More