Who's afraid of their economics classes? Why are students apprehensive about introductory economics courses? And empirical investigation.

Benedict, Mary Ellen ; Hoag, John

I. Introduction

Economics can often evoke strong emotions of fear and anxiety on the part of students. Some of this fear may have a positive effect causing students to work harder. But if the anxiety is too great, it may have a negative effect on student performance, the number of majors, and the general perception of economics. Our experience suggests that students are apprehensive about economics classes. If we want to reduce some of that apprehension, it would be useful to know the sources of the apprehension and the links between the apprehension and demographic characteristics. Understanding the demographics associated with this emotion would enable economics instructors to better prepare for teaching introductory economics classes. This study uses data on four large lecture classes at an Ohio public university for the 1997-1998 academic year to analyze: (1) which students are more likely to be apprehensive about taking an economics class; and (2) what reasons students provide for being apprehensive. Descriptive statist ics and probit regression analysis provide empirical evidence regarding the relationship of students' demographic characteristics (including age, gender, major, and previous economics exposure) and academic ability (as measured by college grade point average (GPA) and math ACT scores) to apprehension about taking an economics course.

We find this topic relevant to understanding the pedagogy of the economics profession. Teaching any subject requires a knowledge of the students who comprise the audience. Although some students take economics as an elective, economics is a required course for a number of majors, including students who may not possess adequate technical skills for learning the subject or who may not desire the economic skills used in the assessment of social problems. Understanding what type of student is more likely to be apprehensive about economics helps us to understand who it is we are trying to teach and how to best approach teaching methods (e.g., perhaps separate courses for business and nonbusiness majors are necessary).

II. The Link Between Apprehension and Economic Education

The economics education literature has examined student attitudes toward learning economics, but generally not in the context of student apprehension. These studies usually focus on whether the student likes to study economics or whether the subject is considered useful for further coursework or future careers. Karstensson and Vedder (1974) developed a questionaire and summary measure of attitude that included student responses about interest in the subject and usefulness to college and post-college work. The authors found a statistically significant and positive relationship between the course grade in economics and precourse attitude. A similar summary measure of attitude was used by Wetzel et al. (1982) and Charkins et al. (1985), linking learning styles and teaching styles to attitudes about studying economics. The research suggests that attitudes about learning economics are less positive when there is mismatch between the learning style of the student and the teaching style of the instructor.

Soper and Walstad (1983) reported that in 1979, the Joint Council on Economic Education commissioned the development of a survey to assess student attitudes toward economics and economic sophistication. The survey (Survey on Economic Attitudes, SEA) consisted of 28 questions, with 14 questions in each sub-area of attitude and sophistication. After considerable development, the survey was successfully tested for reliability and validity, using a national sample of high school students. The data were used in several subsequent studies investigating student attitudes toward learning economics. Fizel and Johnson (1986) examined whether the sequencing of micro and macro courses affects the attitudes of students toward the subject. They found that the order of micro and macro had an ambiguous effect on attitude. Grimes et al. (1989) found that student attitudes were no different from courses without economics U$A than when the course incorporated Economics U$A. (1) A more recent study by Phipps and Clark (1993) us ed a factor analysis with the SEA data. The investigation revealed that three primary factors were related to the high school students' attitudes toward economics: enjoyment of economics as a subject, usefulness of economics, and difficulty of economics. Of the 14 questions in the survey's attitude section, 12 loaded strongly on these factors and two questions were spread over the three factors. The authors concluded that the above-mentioned factors are key to a student's attitude about the subject of economics.

In reviewing other disciplines, we find that the psychology field has developed an explanation of student performance and its relationship to self-concept. In particular, some studies use a self-concept approach to identify factors that lead to math anxiety, which affects performance in math classes (Hackett and Betz 1981; Hackett, 1985). Marsh (1990) examined the endogeneity of performance and self-concept. However, even though Marsh improved the empirical modeling in this area, the static nature of his empirical work does not clearly examine the theory, which requires a dynamic model that allows for continual changes in performance and self-concept. This type of examination would require panel data on individuals over the time, a data issue that has stymied many in this field.

We build on the above studies in the economics education and psychology literature by investigating what determines one student attitude about economics, apprehension. We develop an argument based on the theory of risk perception that originated in psychology and decision science. The ideas that have been developed about risk perception suits our economics perspective. The basic notion behind this theory is that uncertainty increases an individual's fear about a future event. In order to reduce uncertainty, or at least a perception of uncertainty, individuals follow simple heuristics, or information-gathering techniques, that help to reduce the complexity in judgement processes. Kahneman and Tversky (1974) developed a theory of probabilistic judgement that includes three major heuristics:

1. Representativeness. Individuals look for similarities between two events and estimate probabilities based on how similar or different the events are. (e.g., If Dr. Smith is a male with a pocket protector and tape on his glasses, we are more likely to judge him to be an economist than an artist, based on our past academic or work experiences.)

2. Availability. Individuals assess probabilities based on events that can easily be brought to mind. (e.g., We are likely to estimate a higher probability of dying in a plane accident after a major plane crash than during a period of safe air travel.)

3. Adjustment and Anchoring. Individuals make estimates from an initial value (an anchor) and make adjustments as they search for a final answer. With this heuristic, choosing an initial anchor or starting point is key to a final judgement about risk. This heuristic is often used to explain why individuals overestimate or underestimate the risk of some diseases or accidents (Slovic, Fischhoff, and Lichtenstein 1986).

People use the three heuristics to estimate the probability of future events. The estimation of probability is complex, even for experts (Chapman and Chapman 1971). Therefore, individuals find ways to make judgements about risk based on one or all of these rules. (2) The heuristics simplify the information gathering process that reduces estimated uncertainty about the future.

We assume students follow the same rules of judgement regarding the risk of taking an economics course. In this context, risk assessment (and ultimately, apprehension level) centers around the student's estimation of the final performance in a course. However, we believe that the word "performance" means more than final grades. For example, students become apprehensive about taking a course if they believe the experience will hurt them scholastically or if the experience lowers their level of self-esteem. Therefore, apprehension goes beyond estimating a low probability of getting a "good" grade in the course. It also incorporates a sense of deficiency or of unfairness. In other words, if students hear colleagues say that the grade earned did not reflect learning, apprehension is like to increase. (3) Finally, apprehension about a course may stem from the lack of options a student has. Course requirements within a given degree program reduce the option of avoiding an unpleasant academic experience. Furthermor e, students can only select a course section that is constrained by scheduling conflicts.

Students select an economics course and section given their course requirements and time constraints. How does the student assess future risk in this course? First, the student uses the unofficial university "grapevine" to develop an initial anchor for estimating the probability of a good performance. Thus, course and professor reputation are used to develop this anchor. What students hear from friends and colleagues about the department's courses, how individual professors grade, and in general, whether students fare well in economic classes, are used to estimate a probability of success or failure in the class.

One factor affecting student apprehension that has been studied extensively, although in contexts not related to apprehension, is student performance. A number of studies (Park and Kerr 1990; Anderson et al. 1994; Dynan and Rouse 1997) indicate that previous performance in earlier grades or general skill development, as measured by general aptitude tests, and in particular, mathematical skills, have a positive relationship with final grades in introductory economics classes. Furthermore, students who performed well in a previous economics class, tend to perform well in a subsequent class or take additional classes (Karstensson and Vedder 1974). Dynan and Rouse (1997) argue that the primary factor in determining whether students majored in economics or not was whether the students had an aptitude for the subject or believed they had an aptitude for the subject. Generally, students whose grade in principles was above their graduating GPA were more likely to major in economics. These results suggest that knowin g something about economics from a previous class or having a base set of technical skills will likely raise the initial anchor when a student is estimating future class performance.

A thorough examination of gender differences in economics course performance also suggests that initial anchors may be different for men and women. Tests of gender differences in general economics aptitude have been mixed. Generally, female students on average score lower on economics aptitude tests (Anderson, et al. 1994; Bolch and Fels 1974; Heath 1989); however, a recent study finds no gender difference (Williams, et al. 1992). Furthermore, a number of studies find that female students tend to do less well in their economics classes relative to other classes than male students do (Dynan and Rouse 1997). Thus, if women gather their information about courses from other women, they may tend to lower their estimated probability of success. (4)

Individuals use the representativeness heuristic when they relate their economics course to courses that share similar characteristics. It is likely that students who generally have trouble with business courses (the nonbusiness major required to take a few introductory level business courses) or those who have recent and unsuccessful experience in courses requiring mathematics or abstract thinking, will be more apprehensive about taking economics.

An additional area of concern for students may be the mathematical aspect of economics, and math anxiety has been extensively studied. Much of the work in this area has used primary and secondary school students as the focus (Henbree 1988; Meece, et al. 1990) although there are investigations of college students (Zanakis and Valenzi 1997). The general conclusion is that performance (as measured by grades) and perception of ability are strongly related. Students who lack confidence in their mathematics ability tend to have lower grades than those who are confident. Furthermore, a meta-analysis by Henbree (1988) finds that test anxiety is also inversely related to self-esteem. Math anxiety and self-perception of mathematics ability may also be differentiated by gender. Hackett (1985) found that female undergraduate students have a lower tendency to select math-related majors than males. She found that women have lower confidence in their mathematics ability, not because of actual ability, but because of two co njunctive reasons: (1) socialization influences create gender differences in attitudes toward mathematics and learning mathematics; and (2) women are less prepared than men for mathematics classes.

These results suggest that students who lack mathematics skills perform relatively poorly in courses requiring mathematical tools. The self-perception of not being able to do well may increase the likelihood of poor performance, thus raising the level of apprehension about taking a course. Female students may be more apprehensive than males if culture has created a female bias against disciplines using mathematics or if females lack mathematical preparation.

The final heuristic, availability, is most likely used by students measuring their future class performance when they use their current overall college performance to assess future performance. Marsh (1990) finds that academic achievement affects subsequent academic self-concept, which in turn affects future achievement. Students who are currently doing well in school will have more confidence about their future performance, which likely increases their chances of doing well and lowers their apprehension level.

In summary, the use of heuristics to determine future performance in an economics class lead to the following hypotheses: (1) students are more apprehensive if they estimate low probabilities of success based on anchors that use the reputation of the course, the professor, and the discipline; (2) students who do not perform well in related courses or in courses building technical tools are more apprehensive because they estimate the probability of classroom success as lower; (3) students who perceive that they are not doing well in college are more apprehensive about all of their courses, including economics; and (4) women are more apprehensive than men because they tend to do less well in economics and they tend to have less confidence in their math skills.

III. Analysis of Apprehension and Economics

A. Description of the Apprehension Data

This study's data are derived from a survey given to Bowling Green State University economics principles students. (5) Students in four large lecture classes from the 1997-1998 academic year completed questionnaires. These classes were selected for a number of reasons: the large lecture classes were taught at the same conceptual level and they have the same prerequisite; both classes were in same room; and, only two professors were assigned to teach these classes for the academic year (one for microeconomics and the other for macroeconomics). Although the reasoning for our sample selection was driven by the original purpose of the survey (to discover the relationship between seating choice and performance) the large lecture sample also lends itself to the study on apprehension because it is likely that the level of the class, the room environment, and the professor can affect the level of comfort for learning. By controlling for room and professor, we can better focus on how apprehension is related to student demographics.

In the middle of the fall of 1997, we administered the survey to the two large lecture classes, one micro and the other macroeconomics, that had an initial total registration level of 338 students. We repeated the survey administration in the spring of 1998 to classes with an initial total registration of 357 students, but conducted the survey earlier in the term because we experienced a number of absentees during the fall. (6) In compliance with the Human Subjects Review Board protocol at the university, the survey made clear that the respondent had the option of participating in the study, and all but seven who filled out the form granted permission to use their information. We deleted those seven individuals from the analysis, as well as those who did not fully complete the survey. The total number of students who completed all relevant survey information was 547 of the original 695 students.

Two other deletions were made. First, given the documented evidence that respondents in academic studies tend to inflate their personal performance measures (Maxwell and Lopus 1994), we verified information with the registrar's office on current grade point average, college entrance test scores, semester hours completed, and college affiliation. A number of students either were missing from the registrar's data file or were missing key information on the Math ACT score or the current GPA. In addition, some students dropped the class before the end of the term. They were also eliminated from the sample. The final sample after all deletions included 399 individuals, of which 186 took the micro class and 212 took the macro class.

The deletions originally concerned us because it could be that those who miss class are generally poor performers (Douglas and Sulock 1995). Additionally, perhaps those who dropped the class may have been individuals who did not have a positive view of their potential success in the course. However, there is no clear evidence of such patterns because overall GPAs for the full sample versus the reduced sample are almost identical (2.787 versus 2.788). Further, bivarate probit model estimates did not provide evidence of a sample selection bias. (7)

We focus on Bowling Green State University because we were able to gather information at our own institution with relative ease. Bowling Green State University is an Ohio public university that provides educational services to more than 15,000 students. The school has six colleges, and one (the College of Business Administration) requires that all students take the microeconomics and macroeconomics courses. A subset of disciplines in other colleges also requires that at least one economics course be taken at the principles level. (8) Economics can also be taken by students to fulfill a university general education social sciences requirement.

The number of disciplines that include individuals who take economics courses is relevant to our study. Students studying in the business related disciplines (BA students) may have some predilection toward courses like economics or they may have overcome their apprehension once they have taken other business courses, such as statistics or accounting, or prerequisites, such as calculus. Of course, it may also be that poor performance in a previous business course makes a student more apprehensive.

Students from nonbusiness-related disciplines face a different set of experiences. Some disciplines require some business courses, but students are less likely to have a good preparation for economics courses. For example, because calculus is required for a College of Business degree, BA students are more likely to have been exposed to some business concepts with applied mathematics examples (e.g., marginal analysis) when they begin economics. On the other hand, nonbusiness disciplines might require at most College Algebra, where business examples are fewer in number and in complexity. (9) Further, some nonbusiness majors elect to take economics and these students likely come into the classroom with much less apprehension than those who are required to take the course.

Thus, business majors face a very different set of factors than other college majors do. Business majors are required to take economics, they face a different set of prerequisite work and they may have more interest in business-related classes. How students from the College of Business and other colleges respond to our question of apprehension will be analyzed in the next sections.

B. Descriptive Analysis of Apprehension

Apprehension is a subjective word that can take on different meanings. In the context of our survey, we approached the idea of apprehension as does Webster's Dictionary: "viewing the future with anxiety or alarm." Within that context, the last two questions in the survey asked the following:

Were you apprehensive about taking this course? (Circle one) Yes No

If you were apprehensive, it was because: (Choose all those reflecting your views):

1. The course has a reputation for being hard.

2. The professor has a reputation of being hard.

3. Economics is boring

4. Economics is not relevant to my career.

5. Economics is too abstract.

6. Other reason (lines and space provided for writing). (10)

The definition of apprehension embedded in this question reflects the factors found by Phipps and Clark (1993): enjoyment of economics, Item 3, usefulness of economics, Item 4, and difficulty, Items 1, 2, 5. This question also reflects the heuristic of representativeness.

Our prior beliefs about apprehension and economics courses led to the above listed questions. As described in Section II, our hypothesis is that students become apprehensive about a course due to the uncertainty surrounding their potential success. This uncertainty can occur because a student has no prior experience with the subject of economics or has prior experience with economics or a related course (business or mathematics) and performed poorly. While it is clear that Items #1, 2, 4, and 5 relate to our hypothesis, Item #3 is not readily obvious. We include Item #3 because young adults use the word "boring" as an all-inclusive negative assessment of a situation. It can mean "uninteresting" or "hard" or "bad." Although the fact that this term is all-inclusive limits how we can interpret responses to Item #3, the term may elicit at least a small piece of information from the students on why they are apprehensive.

Table 1 presents the descriptive statistics for the variables used in this study. Panel A of Table 2 presents the frequency distributions of course apprehension by demographic characteristics. Chi-Square tests between relevant subgroups (male/female, class, college, micro/macro, and previous exposure to economics or not) suggest that the percentage point differences between subgroups are statistically significant at the five percent level. As noted in Table 1, more than 38 percent of the entire sample responded that they were apprehensive about taking economics. Females were almost twice as likely as males to respond that they were apprehensive, while BA students were less likely to state they were apprehensive compared to their colleagues from other colleges (32 percent versus almost 53 percent for Arts and Science students and 49 percent for students in the colleges of Education and Technology).

As noted in Table 2, women have higher stated levels of apprehension in almost all categories, with almost twice as high a percentage point difference compared to males. The only exception is for the students in the Education, Technology, or Health and Human Services Colleges (Other), where males have a higher rate of apprehension than females (37.5 versus 16.7 percent).

That females are generally more apprehensive than males is not surprising given the earlier discussion. Females may estimate their chance of success in an economics class lower than males and therefore may be more apprehensive. As noted earlier, empirical evidence suggests performance differences between males and females in economics courses, with females earning lower grades than males on average. It is very unlikely that female students know the estimated gender differences in grades in economics courses; however, it is possible they anchor an assessment of success based on information provided by female friends. Females may also lack confidence in the technical abilities related to economics. These issues are discussed below.

In addition to gender and college differences, students taking microeconomics were more likely to be apprehensive than those taking macroeconomics (44 percent versus 32 percent), while those students with no previous economics background were more likely to be apprehensive than those students with some economics course work (approximately 45 versus 35 percent). Several reasons can account for these differences. First, because microeconomics is often taken first at Bowling Green, students may be uncertain of what to expect from the microeconomics course and therefore be more apprehensive because it is their first economics class. Second, the micro course probably has a reputation as a more theoretical and less policy-oriented course with the "more theoretical" causing anxiety. Finally, the result may in part be due to the different reputation of the micro and macro professors.

Panel B of Table 2 reviews student responses regarding the reasons for being apprehensive: One hundred and eighty-three responses were coded 1 through 5 from the survey question. Using the coded responses, we find that approximately 64 percent of the apprehensive students were so because of the reputation of the class or the professor. Furthermore, those who cited the professor's reputation as a reason to be apprehensive also cited the reputation of the class as a reason for their apprehension. Not surprisingly, students appear to use the information provided by others to develop anchors about their potential success in economics. Further, there is some difference by gender in the use of the student network to get information on the economics classes (67 versus 59.2 percent for females and males, respectively), and women were more likely to code that the reputation of the professor made them apprehensive.

The other coded responses indicate that finding economics "boring" is the second highest reason (14.8 percent) for being apprehensive. However, because 68 percent of the time, (11) students used this code in conjunction with another code, it is our opinion that the students may be indicating a negative attitude toward economics, rather than using "boring" as a reason for being apprehensive.

The last two coded responses were used by only a few of those who were apprehensive: relevancy of the course (6 percent) and level of abstraction in economics (10.9 percent). This outcome suggests that conceptual issues cause some students to be apprehensive, with males more concerned about the relevancy of the course (11.3 percent versus 2.7 percent for females) and abstraction (14.1 versus 8.9 for females). Clearly, these issues weigh far less in a student's level of apprehension when compared to the reputational effects of the course, regardless of gender.

C. Probit Analysis of Student Apprehension about Economics Courses

A probit regression estimates the relationship between apprehension about principles of economics courses and student characteristics and includes the following exogenous variables:

1. Math ACT Score. We include this measure of mathematical ability because it is assumed that math ability underlies an individual's perception of future success in an economics class. Those with higher mathematics scores will be less apprehensive because they are less worried about the skills they bring into the course.

2. GPA. The current grade point average of a student is included as a measure of college success at the time of the survey. We assume that students who are doing well in college, as indicated by a relatively higher GPA, will be less anxious about their economics courses. (12)

3. BA. As mentioned earlier, this variable indicates whether a student is housed within the College of Business or not. We assume that BA students will be less apprehensive, ceteris paribus, given their underlying interest in business and exposure to related courses.

4. JRSR. This variable indicates the class status of the individual and equals one if the student is a junior or senior, and zero if the student is a sophomore or freshman. (13) Upperclassmen have more college experiences and they have a better information network to reply upon. Therefore, controlling for their previous experiences (see (6)), we expect juniors and seniors to be less apprehensive.

5. GENDER. As noted earlier, females are more likely to be apprehensive about their economics courses and in general, their success in school. We assume that females will have a higher probability of being apprehensive than males.

6. PREVECON. This binary regressor equals 1 if a student had a previous economics course in college or high school, and 0 if never exposed to economics before. We assume that previous exposure to economics, holding all else constant, will reduce the level of anxiety about taking another course. Previous exposure introduces the student to the graphs and language of economics so that they may be more comfortable with the material.

In addition to the above variables, there are other several controls to the model: MACRO indicates whether the class is a macroeconomics or microeconomics course and controls for topic and professor differences, and SPRING indicates whether the course was taught in the fall or the spring. Recall that one professor taught both macro sections and another taught the micro sections. Also, we wanted to control for timing differences in the survey administration between semesters. The fall semester students had already been tested through quizzes by the time we conducted the survey. Such assessment, although minor, may have created differences between the two semesters. Whether early performance assessment would increase apprehension or reduce it is not clear; students who do well on quizzes and homeworks may be more relaxed about future performance, but if students performed poorly before taking the survey, they would likely be more apprehensive.

We also included an indicator for whether an individual is in the sample twice to control for individual specific error in the regression model. Although only 8 percent of the sample was in the survey twice, it could be that seeing the survey previously affected the second set of responses from those individuals.

Table 3 presents the probit results on the probability of being apprehensive. Column (1) presents the model for the full sample. The results indicate a reasonably good predictability of the overall model, as indicated by the statistical significance of the likelihood ratio test and the percent correctly predicted (67 percent). We find that, as hypothesized, students who come into an economics class with better mathematical skills are less likely to report being apprehensive, and the conditional probability of apprehension falls 1 percentage point with every unit increase in the ACT Math score. This outcome suggests that math skills reduce the level of uncertainty a student has about classroom success in economics, although the effect is not large. In addition, the probability that BA students reported being apprehensive is lower by 17 percent than for students from other colleges, holding all else constant, suggesting that a general business background may reduce uncertainty about classroom success. It may a lso be that BA students have a better information network about the professors or courses and, therefore, come into the economics principles with less apprehension. Students in the spring semester also reported a lower probability of being apprehensive (by 11 percentage points), indicating that our concern about the timing of the survey may be true. The fact that students in the fall term had some idea of their performance by the time we gave the survey may have made increased the probability of apprehension about taking the course.

The other statistically significant factor is gender. Female students reported a 26 percentage point higher level of apprehension than males, holding all else constant. Because we have controlled for math ability, current college success and college type, it appears as though females have more general concern about their success in the economics class than males do. It could also be that cultural differences make the female students more willing to state their apprehension. If females are more likely to state their feelings than males, the statistical result may be an artifact of society norms.

To further test for gender differences, we ran separate regressions for males and females. We find that females drive the full sample model's college type results. The Math ACT score variable was not statistically significant in either gender model and the partial effect was very similar for both (1.6 percentage point reduction in the reported probability of being apprehensive for women and 1 percentage point reduction for males). The probability of reporting apprehension is lower by 13 percentage points for males in the spring semester, suggesting that males may be quicker to adjust their anchors regarding class success or they use the most recent performance measures (quizzes and tests) to assess their future success in the class more readily than females do.

IV. Conclusion

The empirical analysis provides three main conclusions for this study. First, according to the descriptive statistics, the reputation of the course is the most important variable in determining if the student is apprehensive. However, this result may be endemic to the department used in the study and may not be generalizable to other schools. The other two results, however, are more likely to be universal. Clearly, females express more apprehension about economics than males. In addition, math ability seems to matter in terms of which students are apprehensive, and this effect appears to be larger for women than men. Given these results, the main questions are: Do we wish to do something about the levels of apprehension we find in our students? And if we do care to do something, what will we do? Much depends on the source and impact of the apprehension. If reputation reflects the fact that economics is somehow more abstract and mathematical than other fields, it is unlikely that we can do much about it. If t he apprehension in the end somehow induces students to work harder and be more attentive, then perhaps it is not such a bad thing. In either of these cases, working to alter the apprehension may not be productive.

If, on the other hand, the apprehension causes students to come to the classroom with a negative attitude toward economics, or causes students to do poorly because of the high level of stress, then it may be important for us to see what alternatives are available to reduce the anxiety. As noted by many in the mathematics education field, anxiety leads to poor performance in mathematical classes (Hacket, 1985). Further, there is empirical evidence, at least for those students in remedial math courses, that beliefs about mathematics (how students perceive the learning process of mathematics) were related to success or failure in the classroom (Stage and Kloostermann 1995). If students perceive that economics can be learned in a particular way, does it make the learning of economics more difficult for them?

The results suggest the following. One interpretation of finding a gender difference in apprehension may be that females are more honest about their feelings than males. Another interpretation is that women experience a "chilly" climate in the economic classroom (Ferber 1995). A third may be that women are more apprehensive generally and see costs more clearly than men. In the latter case, it may not be desirable to do much about the problem. If, however, apprehension negatively affects performance and stands in the way of success, or if the nature of teaching economics creates gender differences in performance outcomes, some intervention is necessary. One possible way to address this apprehension is to be sure that extra effort is offered to women and others who are apprehensive about mathematics. Help sessions aimed at mathematics used in the economics class may be useful for those with low math skills. A number of business colleges, including our own, have recently instituted Supplementary Instruction Pro grams, aimed at those business majors who are struggling in prerequisite mathematics classes. (14) A second approach is to use class examples that are gender neutral or to incorporate language that appeals to students of both genders so that women and men can better learn to make connections between economic ideas. These ideas can create learning environments that not only reduce the level of apprehension, but prepare the student for the type of learning economic courses require.

This study is a first step in analyzing what causes students apprehension about taking principles level economics classes. However, because our survey was designed for a different purpose, we advocate further research on the subject. Given the results here, a more extensive measurement tool on student attitudes is warranted to examine what factors affect student apprehension, and as a next step, how apprehension is related to student performance. TABLE 1 Descriptive Statistics Variable Description Total Appren = 1 If a student is apprehensive; 0.381 0 otherwise (.486) Math Math Act Score 21.98 (3.91) GPA Grade Point Average at the time 2.782 of the survey (.620) BA = 1 If a student is from the College 0.644 of Business; 0 otherwise (.479) JR./SR. = 1 If a student is a junior or senior; 0.278 0 otherwise (.449) Gender = 1 If a student is female; 0.441 0 otherwise (.497) Prev Econ = 1 If a student had a previous; 0.687 0 otherwise (.467) Macro = 1 If a student was in macroeconomics; 0.469 = 0 if a student was in microeconomics (.500) Spring = 1 If a student compltd survey in the spring; 0.534 = 0 If a student compltd survey in the fall (.499) Second Survey = 1 If a student completed the survey twice; 0.080 0 otherwise (.272) N 399 Expected Sign Variable Female Male in Regression Appren 0.511 0.278 (.501) (.449) Math 21.55 22.32 - (3.81) (3.960) GPA 2.911 2.680 - (.605) (.613) BA 0.642 0.646 - (.481) (.479) JR./SR. 0.261 0.291 - (.441) (.456) Gender + Prev Econ 0.653 0.700 - (.489) (.459) Macro 0.489 0.453 - (.501) (.400) Spring 0.574 0.502 - (.496) (.501) Second Survey 0.108 0.058 not clear (.311) (.235) N 176 223 a) Source: Bowling Green State University data complied by the authors during the 1997-1998 academic year. Final Statistics are based on 399 students. b) Standard deviations are in parentheses. TABLE 2 Apprehension and Economic Courses Panel: A General characteristics of Students who are Apprehensive Females Number Percent Appre- Group Percent Number Group Apprehensive hensive Full Sample 38.1 152 399 Gender: Female 51.1 90 176 Male 27.8 62 223 Class: Fresh/Sop 37.8 109 288 52.3 68 Jr./Sr. 38.7 43 111 47.8 22 Colleges: BA 31.9 82 257 57.5 65 A&S 52.8 26 53 63.3 19 Other * 49.1 44 89 16.7 6 Course: Micro 43.9 93 212 Macro 31.6 59 187 Economics Background: Previous Econ 34.7 94 271 47.0 54 No Previous Econ 45.3 58 128 59.0 36 Semester: Fall 1997 44.0 82 186 56.0 42 Spring 1998 32.8 70 213 45.5 46 Males Number Total Percent Appre- Total Group Females Apprehensive hensive Males Full Sample Gender: Female Male Class: Fresh/Sop 130 25.9 41 158 Jr./Sr. 46 32.3 21 65 Colleges: BA 113 23.6 34 144 A&S 30 30.4 6 23 Other * 36 37.5 21 56 Course: Micro Macro Economics Background: Previous Econ 115 25.6 40 156 No Previous Econ 61 38.8 26 67 Semester: Fall 1997 75 36.0 40 111 Spring 1998 101 19.6 22 112 Panel B: Reasons Why Students Are Apprehensive Percent of Female Reason Responses Number Percent Number Reputation of Class/Professor 63.9 117 67.0 75 Reputation of Professor 4.4 8 5.4 6 Economics is boring 14.8 27 16.1 18 Economics is not relevant 6.0 11 2.7 3 Economics is too abstract 10.9 20 8.9 10 183 * 112 * Male Reason Percent Number Reputation of Class/Professor 59.2 42 Reputation of Professor 2.8 2 Economics is boring 12.7 9 Economics is not relevant 11.3 8 Economics is too abstract 14.1 10 71 * * The total is larger than the number of apprehensive students because respondents could select more than one answer. Source: Survey developed by the authors. TABLE 3 Probit Regressions Estimating Student Apprehension Full Marginal Females Marginal Variable Sample Effects Only Effects Constant 1.008 ** 2.180 *** (.450) (.670) Math -0.034 ** -0.013 ** -0.040 -0.016 (0.020) (.008) (.033) (.013) GPA -0.121 -0.046 -0.227 -0.091 (.126) (.047) (.201) (.080) BA -0.455 *** -0.172 *** -0.642 *** -.256 *** (.145) (.055) (.220) (.088) JRSR -0.034 -0.013 -0.232 -0.093 (.164) (.062) (.239) (.095) Gender 0.689 *** 0.260 *** (.141) (.153) PrevEcon 0.018 0.007 -0.006 -0.002 (.171) (.065) (.283) (.113) Macro -.223 -0.084 -0.123 -0.049 (.171) (.064) (.277) (.110) Spring -0.284 ** -0.107 ** -.144 -0.057 (.143) (.054) (.223) (.089) Second Survey -0.154 -0.058 -0.026 -0.010 (.273) (.103) (.340) (.135) Log-Likelihood -238.520 *** -112.378 ** Percent Correct 67.0 % 61.9 % N 399 176 Males Marginal Variable Only Effects Constant 0.715 (.620) Math -0.029 -0.010 (.026) (.009) GPA -0.070 -0.023 (.167) (.054) BA -0.266 * -0.087 * (.199) (.065) JRSR -0.180 0.059 (.233) (0.076) Gender PrevEcon 0.011 (0.004) (.220) (.072) Macro -0.356 -0.116 (.231) (.075) Spring -0.409 ** -0.133 ** (.192) (.062) Second Survey -0.498 -0.162 (.541) (.175) Log-Likelihood -123.170 ** Percent Correct 72.6% N 223 Data Source: Survey by the authors of Bowling Green State University students in four large lecture introductory economics courses. Standard errors are in parentheses. The asterisks represent the criterion for the level of statistical significance of a Type I error of 10 percent (*), 5 percent (**) and 1 percent (***); one tailed tests used when appropriate. Partial derivatives were calculated by LIMDEP Version 7.0

Notes

(1.) Economics U$A consists of 28 half hour television tapes (14 micro and 1 macro). A textbook was also keyed to the tapes together with other ancillaries to support learning. These tapes could be used as part of a regular class or as a stand alone substitute for a standard economics class.

(2.) The fact that individuals use heuristics to make judgements does not imply that the final judgements are correct. The social decision science literature is filled with examinations of how these heuristics lead to flawed estimations of future events (Kahneman, Slovic, and Tversky 1985; Borcherding, Larichev, and Messick 1990).

(3.) The feeling of unfairness can be either because the grade was too low (students believe they deserved a better grade) or too high (students believe they received credit for material they did not understand).

(4.) An additional concern is the lack of students of color in economics. However, there is a absence of research on the relationship between race and economic class performance or economic majors. We did not investigate race in our study because there are too few students of color in our classes to make statistical estimation of differences by race possible.

(5.) The survey is available upon request.

(6.) Neither professor took attendance and both used pop quizzes as a mechanism to encourage attendance. Because we visited the classrooms on days when quizzes were not given, and because each class developed a pattern for the timing of quizzes, a number of students probably missed the days of the surveys knowing that the risk of having a quiz was very low. To question students before such patterns were apparent, we elected to conduct the spring 1998 survey two weeks into the semester. However, it is not clear to us that taking the survey early helped to increase our sample substantially. We had an increase of 28 responses, but the spring classes had 19 more students when compared to the fall classes.

(7.) Because the sampling bias could be toward students who are less likely to be apprehensive about taking the course, we used a sample selection correction (Green 1992) for possible sample selection bias in our probit regression analysis. The model assumed self-selection on class standing, gender, and association with the business college. Various formulations of the model indicated that the outcomes were sensitive to what was included on the right-hand side of the regression. Oftentimes, the models only converged with loose convergence criteria. Further, the coefficient on the Heckman's sample correction variable in the corrected probit was not statistically significant with any model. Results are available from the authors on request.

(8.) Other colleges requiring at least one principles level course includes: Arts and Sciences (General Business, Economics, International Studies, Fashion Merchandising, Interior Design, Computer Science w/Business Specialization, Environmental Studies); Education (Specialization in Education or Finance, Marketing and Social Studies); Health and Human Services (Environmental Health and Social Work); and Technology (Construction Management and Architecture/Environmental Technology).

(9.) There are nonbusiness disciplines requiring much more mathematics, such as engineering. For this study, however, the students from other disciplines taking business courses generally do not major in highly technical courses (See Endnote 7).

(10.) In our original research, we controlled for student apprehension about the course in an ordered probit model of performance. Apprehension about the class had an impact on final grade in the course, with those who were apprehensive more likely to receive a lower grade in the course (Benedict and Hoag, 2002).

(11.) Available from the authors on request.

(12.) On the other hand, we all know the students who do well and still worry about every homework, test, and course. However, we assume that number to be small and that GPA and apprehension are negatively correlated.

(13.) It may be that a more refined split would discern apprehension differences between seniors and juniors (the senior desperate to get out and graduate, the junior with a little more time to finish college) or freshmen and sophomores (with sophomores having more experience in college). However, there are very few freshmen or seniors in the principles courses in the sample, so we elect to have only two groups.

(14.) One counter argument to this notion of helping is that there is a presumption that students have some mastery of this material before they come to the class. Why is there a need for us to do this remedial work? And why are some other students getting the material elsewhere and these students are not? Should not students take responsibility for the fact that they did not achieve the levels of success in previous courses that they should have achieved? While this point is important, the fact is that the students who show up in our classes do not have the skills or at least are apprehensive about their abilities. If it is true that the apprehension causes poor performance, it may be necessary for us to intervene.

References

Anderson, Gordon, Dwayne Benjamin, and Melvyn A. Fuss, 1994. "The Determinants of Success in University Introductory Economics Courses." Journal of Economic Education, 25 (Spring): 99-119.

Benedict, Mary Ellen and John Hoag, 2002. "Seating Choice in Large Lectures: Are Students Sitting in the Back of the Room Disadvantaged?" Unpublished paper.

Bolch, Ben W. and Rendigs Fels, 1974. "A Note on Sex and Economic Education." Journal of Economic Education, 6 (Fall): 64-67.

Borcherding, Katrin, Oleg I. Larichev, and David M. Messick, 1990. Contemporary Issues in Decision Making. Amsterdam: North-Holland.

Chapman, Loren J. and Jean P. Chapman, 1971. "Test Results Are What You Think They Are." Psychology Today, 5 (November): 18-22, 106-110.

Charkins, R. J., Dennis O'Toole, and James Wetzel, 1982. "Linking Teaching and Student Learning Styles with Student Achievement and Attitudes." Journal of Economic Education, 13 (Winter): 33-39.

Douglas, Stratford and Joseph Sulak, 1995. "Estimating Educational Production Functions with Correction for Drops." Journal of Economic Education, 26 (Spring): 101-112.

Dynan, Kare E. and Cecilia Elena Rouse, 1997. "The Underrepresentation of Women in Economics: A Study of Undergraduate Economics Students." Journal of Economic Education, 28 (Fall): 350-368.

Ferber, Marianne, 1995. "The Study of Economics: A Feminist Critique." The American Economic Review: Papers and Proceedings, 85 (May): 357-361.

Fizel, John and Jerry Johnson, 1986. "The Effect of Macro/Micro Course Sequencing on Learning and Attitudes in Principles of Economics." Journal of Economic Education, 17 (Fall): 87-98.

Greene, William H., 1992. LIMDEP, Version 6.0. New York: Economics Software, Inc.

Grimes, Paul, Timothy Krehbiel, Joyce Nelson, and James Niss, 1989. "The Effectiveness of Economics U$A on Learning and Attitudes." Journal of Economic Education, 20 (Spring): 139-152.

Hackett, Gail, 1985. "Role of Mathematics Self-Efficacy in the Choice of Math-Related Majors of College Women and Men: A Path Analysis." Journal of Counseling Psychology, 32 (Jan.): 47-56.

Hackett, Gail and Nancy E. Betz, 1981. "A Self-Efficacy Approach to the Career Development of Women." Journal of Vocational Behavior, 18 (June): 326-339.

Heath, J.A., 1989. "An Econometric Model of the Role of Gender in Economic Education." American Economic Review: Papers and Proceedings, 79 (May): 235-238.

Henbree, R., 1988. "Correlates, Causes, Effects, and Treatment of Test Anxiety." Review of Educational Research, 58 (Spring): 47-77.

Kahneman, Daniel, and Amos Tversky, 1974. "Judgement Under Uncertainty: Heuristics and Biases." Science, 185: 1124-1131.

Kahneman, Daniel, Paul Slovic, and Amos Tversky (eds), 1986. Judgement Under Uncertainty: Heuristics and Biases. London: Cambridge University Press.

Karstensson, Lewis and Richard K. Vedder, 1974. "A Note on Attitude as a Factor in Learning Economics." Journal of Economic Education, 5 (Spring): 109-111.

Marsh, Herbert W., 1990. "Causal Ordering of Academic Self-Concept and Academic Achievement: A Multiwave, Longitudinal Panel Analysis." Journal of Educational Psychology, 82 (Dec.): 646-656.

Maxwell, Nan L. and Jane S. Lopus, 1994. "The Lake Wobegon Effect in Student Self-Reported Data." American Economic Review: Papers and Proceedings, 84 (May): 201-205.

Meece, Judith L., Allan Wigfield, and Jacquelynne S. Eccles, 1990. "Predictors of Math Anxiety and Its Influence on Young Adolescents' Course Enrollment Intentions and Performance in Mathematics." Journal of Educational Psychology, 82 (March): 60-70.

Park, Kang H. and Peter M. Kerr, 1990. "Determinants of Academic Performance: A Multinomial Logit Approach." Journal of Economic Education, 21 (Spring): 101-111.

Phipps, Barbara J. and James E. Clark, 1993. "Attitudes Toward Economics: Uni- or Multidimensional?" Journal of Economic Education, 24 (Summer): 195-212.

Slovic, Paul, Baruch Fischhoff, and Sarah Lichtenstein, 1986. "Facts Versus Fears: Understanding Perceived Risk." Judgement Under Uncertainty: Heuristics and Biases, eds. Kahneman, Daniel, Paul Slovic, and Amos Tversky. London: Cambridge University Press, 463-492.

Soper, John C. and William Walstad, 1983. "On Measuring Economic Attitudes." Journal of Economic Education, 14 (Fall): 5-17.

Stage, Frances and Peter Kloosterman, 1995. "Gender, Beliefs, and Achievement in Remedial College-Level Mathematics." Journal of Higher Education, 66 (May-June): 294-311.

Wetzel, James, W. James Potter, and Dennis O'Toole, 1989. "The Influence of Learning and Teaching Styles on Student Attitudes and achievement in the Introductory Economics Course: A Case Study." Journal of Economic Education, 13 (Winter): 33-39.

Williams, Mary L., Charles Waldauer, and Vijaya G. Duggal, 1992. "Gender Differences in Economic Knowledge: An Extension of the Analysis." Journal of Economic Education, 23 (Summer): 219-231.

Webster's Ninth New Collegiate Dictionary, 1988. Merriam-Webster, Inc., Springfield, Mass.

Zanakis, Stelios H. and Enzo Valenzi, 1997. "Student Anxiety and Attitudes in Business Statistics." Journal of Education for Business, 73 (Sept-Oct): 10-16.

Mary Ellen Benedict * and John Hoag **

* Bowling Green State University, Bowling Green, OH 43403, e-mail: mbendedi@cba.bgsu.edu

** Bowling Green State University, Bowling Green OH 43403 e-mail: jhoag@bca.bgsu.edu

The authors thank session members of the 1999 American Economics Association meetings in New York City and an anonymous referee for their helpful comments, and Jerry Chen, Bong Li, and Anna Kazmierkiewicz for their date entry efforts. We also thank Guenther Fanter for making us "think twice."