Predicting Math Performance of Middle Eastern Students: The Role of Dispositions

Abstract

The present research examines the contribution of individual differences in chronotype and self-efficacy to the math performance of male and female students in STEM and no-STEM majors. Questionnaires assessing the selected individual differences were distributed to students of Middle Eastern descent enrolled in math courses of the general education curriculum. Summative assessment indices were used to measure performance comprehensively across the entire semester (course grades) and as a one-time occurrence (final test grades). The contribution of morningness and self-efficacy to both course and test performance of STEM students was sensitive to the interaction of gender and major. Instead, neither factor contributed to no-STEM students’ course and test performance. These findings were used to plan improvements in the instruction and advising of students in STEM majors, thereby complying with a key tenet of action research.

1. Introduction

The present study is a snapshot of the action research that the faculty of a department administering general education courses regularly conduct to assess the effectiveness of the instruction they deliver. Action research is an approach to the process of knowledge discovery that aims not merely to understand, but to guide change to improve the lives of the very people who participated in the research [1,2]. Our study originated from observations made by faculty that math courses in the general education curriculum consistently exhibited high failure rates and withdrawals. Interventions enhancing active learning, reinforcing student-centered pedagogy as well as expanding individualized tutoring were found to alleviate, but not substantially alter, students’ expectations of such courses as being considerably challenging, anxiety-provoking, and too often leading to failures. Such expectations were seen as producing self-fulfilling prophecies of impending doom from which students struggled to recover [3,4]. Thus, a decision was made to focus on identifying individual differences that could ameliorate performance in math courses by altering how students approach such courses. The guiding assumptions were that learning is situated and that individuals’ ability to adapt to course demands can impact performance [5,6]. It was further thought that if such individual differences are identified, they could be cultivated via training to ensure acquisition. As a result of these assumptions, information gathered from the extant literature and informal discussions among faculty and students at the institution selected for the present research were used to identify a set of dispositions that might contribute to performance in math courses. The research described here tested the extent to which two of such dispositions (i.e., self-efficacy and morningness) might relate to math performance. Academic major and gender differences in the relationships of interest were examined as well.

The distinction between STEM (Science, Technology, Engineering, and Mathematics) majors and no-STEM majors was considered relevant as students pursuing the former had a larger portion of their curriculum devoted to math, thereby making math more relevant to STEM majors’ persistence, retention, and academic success. Gender was also considered because of evidence indicating that gender differences in math performance are not attributable to general innate and immutable cognitive differences between females and males, but tend to be the artifact of sociocultural influences, which are by definition changeable [7,8].

Our study included students who were citizens of a country, Saudi Arabia (SA), which has in recent years made vast investments to transition from a rigidly gender-segregated, patriarchal structure to one that is aching to reach gender equity [9]. In such a social structure, young women, who had previously been mostly relegated to the home, are now expected to contribute to the economy of SA along with men. To this end, their access to educational programs has been expanded, previously unavailable professions are now open, and their rights have been promoted through a flurry of decrees and public declarations [10]. Female college students are viewed as key to the plan of transforming an oil-based economy into one that is more diversified [9,11]. The road to equity is now visible, albeit challenges exist. Namely, the dismantling of the old patriarchal structure faces resistance and louder calls to preserve traditions and customs that have defined people’s national identity and heritage since they were born [12,13]. For instance, women have been reported to be still reluctant to approach professions previously only available to men, such as those encompassing the STEM fields of engineering and computer science, many choosing business careers under the allure of financial independence [14,15]. Successful careers particularly in STEM professions entail a solid math education, which has been seen as challenging by women [16], mostly due to the interiorization of traditional gender stereotypes [17,18,19]. According to the theory of stereotype threat, women’s success in math becomes difficult to achieve not because of insufficient talent but because of the possibility that performance may confirm traditional gender stereotypes [20,21]. Thus, under the assumption that the math performance of female and male students is the artifact of the sociocultural context in which performance is exhibited [7,8], our study can be considered a snapshot of the extent to which gender equity efforts have succeeded in leveling the field for young women of college-age. Before presenting the methodology of our study, we offer an outline of the dispositions selected for our study and their purported link to academic performance, and to math performance in particular.

2. Self-Efficacy and Academic Success

General self-efficacy is characterized as a motivational trait that illustrates one’s confidence to be able to overcome task-related challenges and solve problems [22,23,24]. It is believed to be the result of the accumulation over a person’s lifespan of the successes and failures experienced across a variety of circumstances [25,26]. Self-efficacy is assumed to be able to influence the person’s initiation of actions, as well as the persistence with which such actions are exercised when difficulties and challenges arise. It is also assumed to facilitate one’s adaptation to novel situations and events, thereby operating as an effective buffer against the adversities that one may encounter in daily life [22,27,28]. As such, it is not surprising that self-efficacy is often reported to be related to academic performance, albeit the strength and direction of this relationship may vary [29,30]. Yet, studies on the purported positive impact of self-efficacy on performance are intermixed with studies illustrating either a negative relationship (reflecting the impact of over-confidence) or a null one [31,32,33,34,35].

It is important to note that gender and academic major may modulate the relationship between self-efficacy and math performance, but the evidence is mixed. For instance, although young women in the general population report lower self-efficacy than young men, gender differences in self-efficacy are found to be minimal or null in STEM fields [36]. Alternatively, they are reported to reflect the pattern found in the general population [37,38]. Gender differences in self-efficacy are said to translate into different interpretations of performance. For instance, a female student may interpret a C on a test as reflecting poor abilities, whereas a male student may interpret the same grade as a passing mark and thus indicative of sound abilities [39,40].

Yet, evidence from understudied student populations, such as college students of Middle Eastern descent, is meager and unclear at best. For instance, in Iran, Zarafshani et al. [41] found that STEM male students displayed higher self-efficacy than STEM female students. Without distinguishing between STEM and no-STEM majors, Rezaei [42] found that the general self-efficacy of Iranian male and female students was not different even though female students’ academic attainment was higher.

3. Morning and Evening Typology and Academic Success

When students are asked about their preferred time for work, they tend to describe themselves as either morning types or evening types. Diurnal and nocturnal preferences reflect differences in not only behavioral patterns but also biological rhythms (e.g., body temperature) [43,44]. Taillard et al. [45] found that people with a preference for evening/night activities (eveningness) exhibited more irregular sleep/wake habits and greater caffeine consumption than people with a preference for morning activities (morningness). The former also flaunted a pronounced sleep debt during weekdays, for which they attempted to compensate during the weekend by spending more time in bed. In their study, although sleep debt was positively associated with daytime somnolence, evening/night types did not recognize themselves as sleepier than morning types. In contrast, Carrier et al. [46] found that after controlling for the age of the participants, morningness was associated with less time spent asleep, and decreased REM activity. The latter is considered a key mechanism for the consolidation/preservation of memory records [47]. Vollmer et al. [48] also reported lower sleep quality for morning types.

Differences in chronotype, which refer to individuals’ preference for morning or evening activities (i.e., morningness and eveningness), are consequential. Preckel et al. [49] found eveningness to be associated with lower academic performance (as measured by grade point average, GPA) in high school students (age range: 14–19). Others found a similar relationship in college students [50,51,52]. Instead, Piffer et al. [53] did not find chronotype to be associated with the GPA scores of both male and female undergraduate students. Akram et al. [54], who examined academic performance as a function of chronotype, reported a pattern dependent on students’ approaches to learning. They distinguished between surface learners who prioritize memorization over understanding and limit learning to the materials required by the course assessment protocol and deep learners who prioritize understanding and go beyond such materials [55]. For students who engaged in deep learning, Akram et al. [54] found that academic performance did not differ between morning and evening/night types. Instead, for students who engaged in surface learning, being a morning type negatively affected performance.

Gender differences in chronotype have also been reported. For instance, Adan and Natale [56], Roenneberg et al. [57], and Mirghani et al. [58] found that women display a stronger inclination towards morningness than men. Evidence that chronotype is related to math attainment is mixed though. For instance, Preckel et al. [59] found eveningness to be a negative predictor of overall grade point average (GPA) in math–science disciplines. Montaruli et al. [60] reported that morningness was linked to higher test performance in STEM disciplines but only for male students. Instead, no correlation between chronotype and academic performance in a STEM field such as medicine was reported by Mirghani et al. [58,61]. To make matters even more unclear, synchrony effects were reported by Carruthers and Young [62], who found that students with a preference for afternoon/evening activities had better math scores in afternoon classes, and by Goldin et al. [63], who found that a preference for morning activities yielded higher scores in math, but only when class timing matched students’ preference. Yet, the available evidence does not clarify the potential interaction between gender and college major, and, as reported earlier, largely neglects understudied student populations, such as Middle Eastern students.

4. The Present Study

Our study rested on the findings of the extant literature about the selected individual difference variables, which are not uniform, as well as on the acknowledgment that understudied student populations, such as Middle Eastern female students of a society in transition, may reveal idiosyncrasies in the way such variables purportedly contribute to academic success. The practical interest that motivated our research was the examination of the relationship between academic success and selected constructs (e.g., self-efficacy and chronotype), not only to address performance difficulties in math classes but also to gather a glimpse of the impact of top-down societal policies on female students who will soon enter the workforce and will be asked to contribute to the economy of SA side by side with men. Our study exists in the context of conflicting values in the social fabric of a society that is in the process of being re-shaped by top-down policies to diversify its economy to make it sustainable and competitive on the global stage [13]. For instance, reforms of the entire SA educational system [64,65] have embraced Western educational tenets and methods, which include active learning, student-centered pedagogy, inquiry-driven instruction, systematic and objective assessment, etc., but their widespread implementation is a work in progress [66,67]. The women of this society have understood that “education is the passport to the future, for tomorrow belongs only to the people who prepare for it today” [68] (p.43). As a result, women have increased their enrollment in higher education [67] and have entered fields before forbidden to them, even though patriarchal values still largely define gender roles [67]. Of particular interest here are STEM fields, which have experienced a considerable increase in female enrollment, but also have posed serious challenges to women as such fields are perceived as openly defying traditional gender roles [69]. To complicate matters, the collectivistic values of tribal and religious traditions emphasize adaptation to social norms, whereas the individualistic values of the Western world underscore independence. Which values will overshadow the others is in the hands of the women who are now college students. The participants of our study were a sample of such women.

In the present research, we specifically asked whether general self-efficacy and morningness contribute to math performance in courses of the general education curriculum of a university conforming to a US curriculum and Western pedagogical methods. We hypothesized that self-efficacy would be more relevant to the academic performance of female students than male students, particularly in STEM fields, since such fields have only recently ceased being the sole domain of men. We reasoned that females’ confidence in their abilities might become particularly advantageous in fields that are perceived as challenging by females due to their being unfamiliar as well as cluttered with traditional gender stereotypes (e.g., STEM fields). For the same reasons, we also hypothesized that if morningness indeed promotes academic performance, its impact might be more visible in female than male STEM majors due to the greater challenges that female students in such majors encounter. These hypotheses were tested with the methodology described below.

5. Method

5.1. Participants

The participants were 627 students enrolled in a math course of the general education curriculum of a Saudi university conforming to a US curriculum, which was designed by the Texas International Education Consortium, and Western pedagogy, which emphasized student-centered instruction, active learning, and structured assessment. Students’ age ranged from 18 to 28. Females were 73.37% and males were 26.63% of the sample. Their majors encompassed all those offered by the selected university, including business, engineering, computer science, architecture, and law. Among the participants, 9.41% were no-STEM male majors, 17.23% were STEM male majors, 27.11% were no-STEM female majors, and 46.25% were female STEM majors. That is, STEM majors were preferred by students irrespective of their gender [ꭓ²(1, n = 627) = 0.14, ns].

A broad variety of math courses were selected to include both STEM and no-STEM majors and all educational levels (45.61% freshmen, 33.65% sophomores, 14.04% juniors, and 6.70% seniors), as well as to minimize any overlap of students. To this end, the following math courses were included: Introductory Algebra (n = 51), Finite Mathematics for no-STEM majors (n = 85), Calculus for non-STEM majors (n = 81), Statistical Methods (n = 111), Calculus I (n = 129), Calculus II (n = 29), Calculus III (n = 64), Ordinary Differential Equations (n = 52), and Linear Algebra and Differential Equations (n = 25). If a student was enrolled in more than one course, random selection was used to pick one course in which his/her performance data would be collected. All courses were taught by senior faculty (n = 7) with at least 5 years of teaching experience at the college level and graduate degrees in the areas they taught. Instruction, which relied on textbooks imported from the US, was delivered in English, albeit students were Arabic–English bilingual speakers. Each faculty taught at least two different courses. Peer observations, course evaluations, and self-evaluations classified the pedagogy adopted by faculty in their classes as student-centered instruction focused on promoting active learning.

5.2. Materials and Procedure

Two individual difference measures were distributed to students, which included the composite scale of morningness developed by Smith et al. [70] and the self-efficacy scale of Chen et al. [71]. The composite scale of morningness was used to assess individual differences in circadian rhythms. The scale contains 13 items that produce a range of scores from 13 to 55. A score of 22 or less indicates an evening type, a score between 23 and 43 stands for an intermediate type (i.e., no stark preference for being either a lark or an owl), and a score of 44 or above defines a morning type. Cronbach’s alpha, a measure of internal consistency, was estimated to be 0.78.

The self-efficacy scale of Chen et al. [71] was used to measure students’ general confidence in their abilities to complete tasks. The scale entails 8 generic statements of confidence (e.g., “When facing difficult tasks, I am certain that I will accomplish them”) to be rated on a continuum from strongly disagree (1) to strongly agree (5). Cronbach’s alpha was estimated to be 0.95.

At the start of the semester, rosters were collected from each participating faculty. At the end of the semester, records of the course grades and final tests earned by students were also collected. Faculty were not told of the purpose of the study until all records were gathered. The percentage of students who had completed both questionnaires was 73.16% (627 out of a total of 857 students). In this sample, 88.36% completed the course, whereas 11.64% withdrew for poor performance at some point after the middle of the semester.

All information was coded to protect participants’ identities (i.e., any identifying information was deleted after matching students’ responses and their performance). Data from the participating students were aggregated. Participation complied with the guidelines for educational research of the Office for Human Research Protections of the US Department of Health and Human Services and with the American Psychological Association’s ethical standards. The study was conducted under the purview of the Deanship of Research.

6. Results of Analyses

The results of inferential statistics presented below are considered significant at the 0.05 level. A Bonferroni correction was applied to all pairwise comparisons following significant effects [72]. Analyses are organized by the question they answer. These analyses are preceded by a description of the properties of the sample.

6.1. Descriptors of the Sample

Responses to the composite scale of morningness [70] yielded three separate but unequal categories of students: evening-type = 7.01%, intermediate = 88.36%, and morning-type = 4.63%. Namely, most students classified themselves as between an owl and a lark, thereby illustrating adaptation to both day and evening schedules. The average score on the self-efficacy scale [71] was 3.82 (SEM = 0.041). If 3 is considered the middle score on this scale (i.e., an indicator of neutrality), then scores can be organized into low, intermediate, and high self-efficacy. In our data set, 16.11% of the students qualified as low self-efficacy, 19.46% qualified as medium self-efficacy, and 64.43% as high self-efficacy.

These patterns were examined to determine whether differences existed as defined by gender and major.

Table 1. Students’ mean morningness and self-efficacy responses as a function of gender and major (SEM are in parentheses).

Major	Morningness		Self-Efficacy
	Male	Female	Male	Female
No-STEM Majors	29.05 (0.89)	32.46 (0.45)	4.00 (0.12)	3.82 (0.08)
STEM Majors	33.30 (0.63)	32.50 (0.38)	3.85 (0.10)	3.78 (0.06)

displays the means and standard error of the mean (SEM) of the participants as a function of gender and major.

A 2 (gender: female and male) X 2 (major: no-STEM and STEM) between-subjects ANOVA on morningness scores, yielded a main effect of gender [F(1, 623) = 4.82, MSE = 39.97, p = 0.028, ηp²= 0.008], and a main effect of major [F(1, 623) = 12.90, MSE = 39.97, p < 0.001, ηp²= 0.020]. A significant interaction [F(1, 623) = 12.48, MSE = 39.97, p < 0.001, ηp²= 0.020], indicated that males in STEM majors had a greater predilection for morning activities than males in no-STEM majors [t(165) = 3.95, p < 0.001], whereas no differences existed in females [t < 1, ns]. Another interesting pattern was that there were no gender differences in morningness among STEM students [t = 1.10, ns], whereas, among no-STEM students, females had greater predilection for morningness than males [t(227) = 3.71, p < 0.001]. The same ANOVA on self-efficacy scores yielded no effects or interaction [ Fs ≤ 1.73, ns].

Age was positively correlated with educational level (coded as 1 = freshmen, 2 = sophomores, 3 = juniors, and 4 = seniors) [r_s = +0.555, n = 627, p < 0.001]. However, educational level was not related to either morningness scores [r_s = +0.041, ns] nor self-efficacy scores [r_s = +0.056, ns]. Morningness and self-efficacy were also unrelated [r_s = +0.022, ns].

In our study, both course grades and final test grades were treated as summative assessment indices of performance. However, course grades represented the cumulative effort of students across an extended timeframe (a semester), whereas final test grades were representative of a one-time effort. The first two columns of

Table 2. Female and male no-STEM and STEM students’ performance as measured by course grades and final test grades.

	Course Grades		Final Test Grades
No-STEM Majors	Male	Female	Male	Female
Pass (≥66%)	44.06%	72.35%	57.41%	38.82%
Fail (<66%)	47.46%	17.06%	42.59%	61.18%
Mean	61.52 (3.46)	70.03 (2.18)	58.05 (4.14)	52.36 (2.22)
Withdrawal	8.48%	10.59%
STEM Majors
Pass (≥66%)	50.00%	68.97%	44.79%	37.30%
Fail (<66%)	38.89%	17.93%	55.21%	62.70%
Mean	60.48 (2.54)	67.34 (1.77)	53.39 (2.82)	50.01 (1.81)
Withdrawal	11.11%	13.10%

display the percentage of female and male students in no-STEM and STEM majors who passed the course in which they were enrolled with a cumulative grade of 66% or higher, the percentage who failed (a cumulative grade below 66%), and the percentage of students who withdrew before the final exam (usually in the second half of the semester). The last two columns display the percentage of female and male students who passed the final test (i.e., obtained a grade of 66% or higher), and the percentage who failed (i.e., obtained a grade below 66%). The mean of the grades received and the standard error of the mean (in parentheses) are also reported.

When course grades, measured as pass or fail, were submitted to a chi-square test, females emerged as more likely to pass math courses than males in both no-STEM [ꭓ²(1, n = 229) =15.42, p < 0.001] and STEM majors [ꭓ²(1, n = 398) = 12.26, p < 0.001]. In these analyses, students who withdrew from a course and those who completed it but received a failing grade were grouped into one category and compared with students who received a passing grade.

When final test grades, measured as pass or fail, were submitted to the chi-square test, a different pattern emerged. In no-STEM majors, males were more likely to get a passing grade than females [ꭓ²(1, n = 206) = 5.60, p = 0.018]. In STEM majors, both male and female students were more likely to fail than to pass [ꭓ²(1, n = 348) = 1.63, ns].

It is important to note that an examination of synchrony effects (i.e., individuals with a particular chronotype obtain higher academic performance than individuals with the opposite chronotype when they are evaluated at the time of the day that fits their chronotype) [63] was unfeasible since only a small number of participants could be uniquely classified as either morning or evening types. Thus, a point biserial correlation was conducted on course grades and class timing (morning = 0 versus afternoon/early evening = 1) in each of the four sub-samples created by gender and major. The timing of course offering did not appear to be related to female students’ course performance [no-STEM: r = +0.041, n = 170, ns; STEM: r = +0.054, n = 290, ns]. However, evening preferences were inversely related to the performance of male students [no-STEM: r = −0.442, n = 59, p < 0.001; STEM: r = −0.195, n = 108, p = 0.044]. Thus, the analyses of the contribution of individual differences to performance, which are displayed below, included the timing of the course in which students were enrolled.

6.2. Assessment of Contributions to Performance

We asked whether morningness, self-efficacy, and time of course offering (a situational variable) accounted for students’ course grades by conducting a linear regression with these factors as the predictors and course grades as the outcome variable. We conducted separate analyses for female and male students in no-STEM and STEM majors to detect whether there were different patterns of contributions. Furthermore, to recognize that students who withdrew from a course failed differently from those who completed it but received a poor grade, such students were assigned a score of 0%.

Table 3. Regression analyses for course grades of males and females in no-STEM and STEM majors.

Predictor Variables	B	SE	Beta	t	Sign.
No-STEM Male Students
Constant	77.705	20.109
Morningness	−0.366	0.470	−0.095	−0.780	ns
Self-Efficacy	2.433	3.370	0.087	0.722	ns
Timing of Course	−23.709	6.657	−0.430	−3.561	0.001
No-STEM Female Students
Constant	64.525	16.294
Morningness	−0.162	0.383	−0.033	−0.422	ns
Self-Efficacy	2.480	2.100	0.092	1.181	ns
Timing of Course	2.653	4.417	0.047	0.601	ns
STEM Male Students
Constant	20.933	17.079
Morningness	1.095	0.376	0.271	2.908	0.004
Self-Efficacy	2.696	2.442	0.103	1.104	ns
Timing of Course	−11.086	5.119	−0.200	−2.166	0.033
STEM Female Students
Constant	47.179	10.987
Morningness	0.125	0.276	0.027	0.453	ns
Self-Efficacy	3.875	1.755	0.130	2.208	0.028
Timing of Course	2.685	3.551	0.044	0.756	ns

Note: no-STEM Males: R = 0.462; no-STEM Females: R = 0.110; STEM Males: R = 0.339; STEM Females: R = 0.144.

shows that the course performance of male and female STEM students was sensitive to different factors. Namely, the more STEM female students were confident in their abilities, the higher were their course grades. Instead, the more STEM male students were morning type and avoided enrollment in afternoon/evening classes, the higher were their course grades. In contrast, in no-STEM majors, the more male students avoided enrollment in afternoon/evening classes, the higher were their course grades, whereas the performance of female students was insensitive to the selected factors.

The same analysis conducted on final test scores again illustrated that test performance was sensitive to students’ major selection (no-STEM versus STEM) and gender (see

Table 4. Regression analyses for final test grades of males and females in no-STEM and STEM majors.

Predictor Variables	B	SE	Beta	t	Sign.
No-STEM Male Students
Constant	67.888	24.249
Morningness	−0.624	0.567	−0.134	−1.101	ns
Self-Efficacy	5.964	4.063	0.178	1.468	ns
Timing of Course	−24.169	8.028	−0.367	−3.011	0.004
No-STEM Female Students
Constant	44.102	16.609
Morningness	−0.013	0.390	−0.003	−0.034	ns
Self-Efficacy	2.531	2.141	0.093	1.182	ns
Timing of Course	−2.032	4.503	−0.035	−0.451	ns
STEM Male Students
Constant	−13.509	18.632
Morningness	1.536	0.411	0.341	3.739	0.000
Self-Efficacy	5.326	2.664	0.183	1.999	0.048
Timing of Course	−7.219	5.584	−0.117	−1.293	ns
STEM Female Students
Constant	33.289	11.164
Morningness	0.259	0.280	0.054	0.923	ns
Self-Efficacy	3.464	1.783	0.113	1.943	ns
Timing of Course	−8.764	3.608	−0.142	−2.429	0.016

Note: no-STEM Males: R = 0.447; no-STEM Females: R = 0.103; STEM Males: R = 0.384; STEM Females: R = 0.184.

), but in a way that differed from the pattern that emerged from course grades. Indeed, among STEM male students, both morningness and self-efficacy were beneficial to their final test performance. That is, the greater the predilection for morning activities and the higher the confidence in their abilities, the better were their final test grades. Avoidance of evening classes was only beneficial to the test grades of no-STEM males and STEM females. Thus, although both course grades and final test grades could be considered indices of summative assessment, the restricted timeframe of a two-hour exam created unique assessment conditions that differed from those that defined math courses across an entire semester.

7. Discussion of Results

The results of the present study can be summarized in two points: First, although there were no differences involving gender and major in students’ confidence in their abilities to overcome challenges (i.e., self-efficacy), there were gender differences in morningness (as measured by overall scores on the scale of Smith et al. [70]) among no-STEM majors (i.e., females displayed greater morningness than males), but not among STEM majors. Furthermore, females’ course performance was largely insensitive to the time at which math courses were scheduled, whereas males tended to exhibit higher course performance in morning classes. Yet, the fact that most students fell into the intermediate chronotype category suggests that changes in the scheduling of classes from one semester or school year to another might have succeeded in tempering chronotypes, albeit much more for females than for males. BaHammam et al. [73] also found that Saudi college students had no preference for morningness or eveningness, a finding that is replicated in other parts of the world [74,75]. The absence of significant differences in self-efficacy instead may be due to the rather high levels reported by the students in our study (i.e., a ceiling effect).

Second, individual differences were more or less valuable indices, depending on not only the gender and major of the students but also the global versus time-restricted nature of the assessment. For instance, in the constrained timeframe of final tests, both morningness and self-efficacy made a positive contribution to the math performance of males in STEM majors, whereas females in the same majors were insensitive to both factors. Thus, during debriefing, we informally examined whether math anxiety might have weakened the contribution of self-efficacy and morningness to test performance in the remaining students. Informal conversations with students and instructors during debriefing suggested that even though the view of math as a challenging and demanding academic subject was widespread, math anxiety was less likely to be mentioned by STEM males. The latter has been thought to drive both male and female students to select no-STEM majors [76,77], as well as to be a key determinant of performance, thereby potentially weakening the impact of the individual difference measures selected for the present investigation. Less concern about failing math final tests as well as courses by male STEM majors may reflect gender stereotypes of males’ superior computational abilities, along with the absence of stereotype threats in the traditionally male-dominated fields of STEM [17,18,19], even in the face of substantial failure rates.

When the assessment was distributed across the entire semester and included both tests and homework assignments, morningness made a positive contribution to the overall math performance of males in STEM majors, whereas self-efficacy made a positive contribution to the overall math performance of females in the same majors. These findings suggest that if a broader window is taken of STEM students’ performance, the challenges faced by STEM female and male students come into clear focus. Self-efficacy may benefit STEM females due to the gender stereotypes favoring males in STEM fields combined with the fact that such fields have only recently been open to women. STEM males, reinforced by favorable stereotypes, may suffer from entitlement [77] and thus benefit from morningness, which serves as an indicator of the effort devoted to academic demands as opposed to the leisure activities that are typically undertaken by SA youth after sundown.

The present findings contribute to the extant literature on the role of individual differences in STEM education [78] by offering a window into an understudied student population. They differ from the findings of research that has highlighted gender differences in the self-efficacy of Middle Eastern (i.e., Iranian) students pursuing STEM majors [41], suggesting that regional diversity needs to be investigated. The absence of gender differences in chronotype among STEM students is inconsistent with the gender differences reported by Mirghani et al. [61] in Saudi STEM students. The selective contribution of morningness to the performance of STEM male students also contrasts with the findings of other studies of no correlation between chronotype and academic performance in a STEM field such as medicine [58,61], suggesting that diversity may exist among student sub-populations in STEM fields.

Individual differences require as much consideration as does the broader cultural context in which they exist. In a society in transition to a sustainable economy fully integrated into the global marketplace, changes cannot be limited to natural resources (e.g., solar power) but must include human capital [9,11]. Women of college-age are asked to contribute to the Saudi economy while they are still facing the remnants of a system that has relegated their mothers and grandmothers mostly to the home [12,13]. Their presence in STEM fields has been promoted from the top, but their success still heavily depends on choices made well before entering college [79]. Such choices are likely to be driven by enduring gender stereotypes that may prevent young women from fully appreciating STEM fields and the educational and professional opportunities in such fields that are currently available to them [14,15]. Math competency is generally considered a key aspect of academic and professional success in STEM fields [80,81]. Thus, although the larger cultural context of past practices may impact young women’s choices, at the college level, the promotion of gender equity is likely to be key to female students’ successful learning of math [7,8]. At the selected institution, women’s determination to succeed, as evidenced by their higher pass rates in the math courses of our research, can be considered one of the most powerful antidotes against persisting gender stereotypes along with the promotion of gender equity in and outside the classroom and in the workplace. Males may also benefit from an atmosphere of gender equity as it can counteract stereotypes built on a patriarchal view of society and, potentially, improve their math performance through enhanced motivation and effort.

At the selected institution, findings collected in the classroom are reviewed by faculty and administrators, first to determine their impact on teaching and learning, and then to guide interventions to promote students’ academic success [82]. Thus, they are locally relevant in addition to the relevance that might arise from their contribution to the extant literature. The protocol used is that of action research whereby findings of studies conducted in the classroom are used to understand teaching and learning as well as to benefit the participants’ lives [1,2]. For instance, the present findings have been used to plan interventions for STEM students differentiated by gender. The planned intervention has focused on developing awareness of the impact of self-efficacy on academic success in female STEM majors, and on developing or merely reinforcing morning habits in male STEM majors. Of course, differentiated training has been envisioned to be linked to other more traditional initiatives intended to ensure proper course placement and the nurturing of math competencies through sound tutoring and advising. On the other hand, the absence of a significant contribution of the selected factors to math performance in no-STEM students has suggested that other individual differences might be relevant to such students. The identification of such factors is underway.

8. Conclusions

Our study aimed to uncover individual differences that may contribute to male and female students’ math performance in STEM and no-STEM majors. The selection of individual differences was informed by the extant literature and motivated by faculty who were concerned about learning in their math classes. Its findings have had two noticeable implications. They have stimulated discussions on teaching and learning in math courses, which have then guided the development of targeted interventions intended to ameliorate math learning at the selected institution. Furthermore, they have been understood by faculty and administrators as starting points from which further probing of individual differences can be carried out. Thus, the findings of the present investigation have clear implications for math learning and teaching. Namely, they highlight the need to consider individual differences locally (i.e., at a particular institution) as a way to develop suitable interventions to enhance students’ academic success. Of course, we do not claim that our findings inevitably apply to other institutions and student populations. We claim though that the procedure used to investigate individual differences, which has involved the very faculty who are responsible for teaching a given subject matter, has the potential to enhance their engagement in educational research, an outcome that is likely to extend to other institutions. A deeper understanding of students is also likely to enhance educators’ sense of agency, another outcome that has been often reported to us.

The present study has limitations that will be addressed in future research. First, it is noteworthy to mention the lack of information about students’ prior math attainment [83] and any quantitative assessment of math anxiety. Second, college students from SA may be in a privileged position compared to those of other Middle Eastern countries as they receive substantial financial support and are given an unparalleled range of academic resources to support their educational endeavors (e.g., state-of-the-art facilities, curricula approved by Western foreign educational organizations to meet international standards, etc.). Thus, the impact of individual difference variables may be weaker than in other populations. Third, although the uneven sample of male and female students is typical of studies conducted in SA, see [61], and could be partially attributed to student enrollment rates, it could also reflect the low engagement of male students in activities that fall outside the assessment protocol of a course. Low engagement may be fostered by a sense of entitlement in men of a society emerging from gender inequity [13]. Thus, gender differences in engagement might need to be explored further to determine their nature and exact origin. Fourth, although the timing of course delivery was either morning or afternoon/early evening, most students were classified as intermediate types, thereby reducing the morning and evening types to very small groups, and postponing the assessment of synchrony effects to future research. Fifth, the contribution of self-efficacy to math performance may vary as students’ academic knowledge and experience increase [83] and their educational journey progresses towards its end. As such, a longitudinal perspective may need to be adopted. Sixth, the learning of math may be hindered or challenged by study materials and activities presented in a second language [84]. Although at the institution selected for the present investigation, students’ English language proficiency is systematically documented through standardized tests administered before admission and reinforced in preparatory courses, the role played by a second language in the acquisition of math needs to be further explored. Evidence regarding this issue, which is available in the extant literature, is not clear-cut. On one hand, it is recognized that the language in math texts is different from ordinary prose in that it relies on words that are novel or whose meanings may differ from those of everyday language, all embedded in a variety of unfamiliar symbols and graphs [85]. Thus, studies have linked higher math performance to learners’ reliance on their first language [86,87]. On the other hand, studies have also shown that students may prefer to rely on English (i.e., their second language) in math courses, thereby leading to the question of the extent to which preferences may shape performance above and beyond a language’s prior use [88,89].