Priscilla A. LaBarbera

Associate Professor of Marketing

Director, Undergraduate Marketing Program

Department of Marketing

Jeffrey S. Simonoff

Professor of Statistics

Department of Statistics and Operations Research

Leonard N. Stern School of Business

New York University

Contact author: Priscilla A. LaBarbera, Associate Professor and Director, Undergraduate Program, Marketing Department, Leonard N. Stern School of Business, New York University, 44 West Fourth Street, Suite 8-90, New York, NY 10012-1126

Phone: 212-998-0517; Fax: 212-995-4006; e-mail: plabarbe@stern.nyu.edu

This document summarizes the formal statistical testing of hypotheses based on a survey of undergraduate students designed to examine the key factors that are involved in selecting a marketing major, as reported in the paper "Toward Enhancing the Quality and Quantity of Marketing Majors: An Empirical Study," by Priscilla A. LaBarbera and Jeffrey S. Simonoff.

H1: As compared to students who choose alternate majors, students who select marketing make their decisions later during their academic programs. Further, these students tend to place more importance on the qualitative nature of coursework, and less importance on the quantitative nature of coursework.

H2: Students who choose alternate majors perceive the marketing coursework to be less demanding and challenging as compared to those who select marketing as a major.

H3: There are perceptual differences between marketing majors and nonmarketing majors with respect to the image of marketing as a career.

H4: The importance attached to career-related factors including employment opportunities upon graduation, salary size upon graduation, quality of work life, prestige associated with the career, glamour associated with the career, and enjoyment of the career is not predictive of whether students will or will not choose marketing as their majors.

H5: While marketing majors are somewhat more optimistic about career opportunities for marketing majors and the competitiveness of starting salaries in the marketing field as compared to nonmarketing majors and undecided students, all three groups of students are generally pessimistic about these marketing career variables.

H6: Students who are undecided about their majors are open to outside influences. As compared to students who have selected marketing or a nonmarketing major, the undecided students attribute greater importance to parental and peer pressure.

In-depth interviews were conducted with New York University’s undergraduate business school marketing majors and nonmajors, business school academic advisers, and the New York University career placement staff regarding the choice of a business school major in general and marketing as a chosen major in particular. On the basis of these interviews and a study of the available literature, several hypotheses regarding the key factors that influence the choice for or against a marketing major were developed. These hypotheses were the basis for the development of a self-completion survey.

New York University is a private urban university in the northeast with a total of about 50,000 students. There are 16,253 undergraduate students with 2,100 attending the business school. During the Spring 1996 and Fall 1996 semesters, students were surveyed in one of two required courses (Introduction to Marketing or Introduction to Finance) as well as Consumer Behavior, which is a requirement for marketing majors. The data collection was designed so that there was no overlap in respondents. Two MBA level research assistants administered the self-completion survey, which took approximately ten to fifteen minutes to complete. There was a 100% response from the students in these classes. A total of 364 undergraduates (representing over 17% of the total population of undergraduates attending the business school) completed the survey. Of the total sample, approximately 76% were juniors and 24% were seniors, representing 52% males and 48% females. A copy of the survey (in MS-Word and html formats) and the resultant data (in ASCII format) can be obtained via the World Wide Web at the URL http://www.stern.nyu.edu/~plabarbe/survey.

The hypotheses are tested statistically using nominal logistic regression models. The response variable representing a student’s major has three categories: marketing, nonmarketing, and undecided. Let  ,  , and   be the probability that student i is a marketing, nonmarketing, or undecided major, respectively, and   be the value of a predicting variable x for that student. The nominal logistic regression model represents these probabilities using logistic functions:

In this example, the logarithm of the odds of being a nonmarketing major relative to being a marketing major, and the logarithm of the odds of being an undecided major relative to being a marketing major, are modeled as being (different) linear functions of the predictor x. That is, the marketing major category is the reference category. Given that any of the three categories can be chosen as the reference category without changing the model, the category that is most relevant for the specific hypothesis being examined will be used. In addition to examining an overall measure of the statistical significance of the predictor x as a classifier of the three categories, the statistical significance of the individual slope coefficient or can be used to assess whether the predictor differentiates between the individual category and the reference category. The model also can be generalized easily to include multiple predictors.

The results of the different logistic regression fits performed here are given in Table 1. The categories under "Hypothesis" describe the equation being fit, with the second category being the reference category. The fitted equation is given under "Equation," with associated p-value for the slope coefficient(s). If the coefficient is statistically significant at a .05 level, the estimated odds ratio is also given under "Odds ratio." The entry under "Overall p" is the p-value for the likelihood ratio test of the significance of the variable(s) in classifying to the three categories, with low values indicating significance. The entry under "Goodness-of-fit p" is the p-value for the deviance goodness-of-fit statistic, with low values indicating lack of fit. This value is not given for models with multiple predictors, as it is not appropriate in that situation due to sparsity of the resultant table of counts; see Simonoff (1998) for discussion of this point.

H1 addresses the question of whether marketing majors do, in fact, choose their majors later during their academic careers as compared to non-marketing majors. This is assessed using student responses as to when they made their decisions to choose their majors. The hypothesis also asserts that marketing majors are more qualitatively oriented and less quantitatively oriented than are nonmarketing majors in terms of the importance placed on these aspects of coursework.

The results support H1 on all three aspects, which is most apparent in a multiple regression model using all three factors as predictors. There is a significant relationship between when respondents choose their majors and the relative odds of being marketing versus nonmarketing majors. The estimated odds ratio of .80 implies that (holding importance of the qualitative and quantitative aspects of coursework fixed) an increase of one category in the time of decision (which roughly corresponds to one year) is associated with decreasing the odds by approximately 20% of being a nonmarketing major relative to being a marketing major. About 30% of the respondents decided on their majors during their junior or senior years, and of these 42% chose marketing. This can be compared to only 32% of the students who made earlier decisions about being marketing majors.

The coefficients for importance of quantitative and qualitative nature of coursework have opposite signs, consistent with greater importance of quantitative nature being associated with not being a marketing major, and greater importance of qualitative nature being associated with being a marketing major (holding all else in the model fixed). Roughly 10% of marketing majors view the quantitative nature of coursework as not at all important, compared with 5% of nonmarketing majors; 12% of marketing majors view it as extremely important, compared with 16% of nonmarketing majors. The results are stronger for the variable representing importance of the qualitative nature of coursework, with 77% of marketing majors finding this important, compared with 63% of nonmarketing majors and 60% of undecideds. Despite this, marketing majors apparently do not view their field as not requiring quantitative skills (only 18% of marketing majors disagree with the statement "Marketing coursework requires quantitative skills" based on a five-point Likert scale ranging from Strongly Disagree (1) to Strongly Agree (5)), although other majors are less sure (28% of nonmajors and 31% of undecideds, respectively, disagree with the earlier statement).

Several of Hugstad’s (1997) recommendations were based on the premise that the external perception of marketing as a major is different from the internal perception. H2 refers to issues of this type, as it relates to the perception of the difficulty of coursework. This is assessed using student responses to the statement "Marketing coursework is demanding and challenging." The results support H2, as both nonmarketing majors and undecideds have significantly lower ratings of the difficulty of marketing coursework in comparison to the marketing majors. While 18% of marketing majors strongly agree with the statement, only 8% of the nonmarketing majors and 10% of the undecideds strongly agree. Furthermore, while only 37% of the non-marketing majors strongly agree or agree with the statement, 65% of the marketing majors indicated agreement.

Another survey statement indicating the perception of the challenge afforded by the marketing major, "A marketing major is less difficult than other business school majors," was also examined. The results here were similar, but less clear-cut. Although there was a statistically significant difference between marketing majors and nonmarketing majors, it was weaker (p = .018 versus p < .001). Further, a multiple regression with both of these variables as predictors yields a model with the coursework variable highly significant and the major variable insignificant. Thus, it appears that the differences of perception are focused more clearly on the coursework associated with the marketing major, rather than on the major itself.

Are marketing major recruitment problems related to the image of a marketing career? H3 asserts that there is a difference in perception of the image of marketing between majors and nonmajors. Two survey items address this question: "A career in marketing is as well-respected as a career in accounting/finance," and "The field of marketing offers a better ‘quality of life’ than other business school majors." The data support H3, in that each variable is statistically significantly associated with major category. An even better model that includes both variables leads to similar conclusions. Nonmarketing majors are more likely to disagree that the marketing major is as well-respected as accounting and finance, and that marketing offers a better "quality of life," while marketing majors are more likely to agree with these statements. For example, 46% of marketing majors agree with the first statement and 44% agree with the latter one; this can be contrasted with 21% and 19%, respectively, of nonmarketing majors.

A major implied premise of Hugstad’s (1997) analysis of the status of marketing as an undergraduate major is that the choice is unrelated to the objective decision-making criteria of the student. For example, the existence of employment opportunities upon graduation may be a factor that is of lesser or greater importance to students who choose the marketing major as compared to those who do not. This will complicate recruitment strategies because the market for marketing majors might be limited to those students who attach less importance to such opportunities. H4 addresses this issue by examining the relationships of major choice category to the importance of the following factors: employment opportunities upon graduation, salary size upon graduation, quality of work life, prestige associated with the career, glamour associated with the career, and enjoyment of the career.

At a marginal level, the data generally, but not completely, support the hypothesis. Each of the variables is not statistically significantly related to category membership, as shown in Table 1, except for the importance of the enjoyment of the chosen career. While 75% of marketing majors consider enjoyment of a chosen career extremely important in choosing a major, only 50% of nonmarketing majors and 56% of undecideds feel that way (the estimated odds ratios around .5 imply that an increase of one level of importance attached to career enjoyment doubles the odds of being a marketing major relative to being a nonmarketing major or undecided).

There is also a more complex relationship between major category and these objective criteria that does not surface in the marginal logistic regressions. A multiple regression using these variables reveals that the importance of prestige and glamour together are predictive, even though they are not marginally predictive. This result can be explained by the summary shown in Table 2. The lack of significance of each variable marginally is reflected in the fairly constant percentages of marketing majors in the two margins under Total. The positive coefficient of Prestige in the regression fit implies that, holding the importance of glamour constant, an increase in the importance of prestige is associated with a decreasing likelihood of being a marketing major. This is apparent in the observed percentages, which tend to decrease along each row.

By contrast, the negative coefficient of Glamour implies that, holding the importance of prestige constant, an increase in the importance of glamour is associated with an increasing chance of being a marketing major, as can be seen in the observed percentages, which tend to increase going down columns. It should be noted, however, that views on the importance of prestige and glamour are strongly associated, which can be seen from the concentration of observations along the diagonal in Table 2. Thus, the practical importance of effects based on holding one variable fixed while varying the other may be limited.

H5 asserts that marketing majors are more optimistic about career and salary opportunities in the marketing field than are nonmajors. This hypothesis is only weakly supported by the data. There is a marginal effect distinguishing marketing from nonmarketing majors based on feelings on career opportunities for marketing majors, but it is somewhat illusory. While 34% of marketing majors agree with the statement "There are many career opportunities in the marketing field for new graduates," only 24% of the nonmarketing majors agree with this statement. Opinions about salary size competitiveness follow a similar pattern, with 22% of marketing majors and 12% of nonmarketing majors agreeing with the statement "The field of marketing offers competitive starting salaries relative to other business fields." The major point about these findings, however, is that all groups have a pessimistic view of the marketing job market and the competitiveness of starting salaries in the field.

The previous hypotheses have focused on the differences between marketing and nonmarketing majors, which is reasonable as an investigation of what factors are related to choice of major. The students who have not yet chosen a major are also of interest, in that they represent potential future marketing majors. In the analyses thus far, there has been little distinction between undecideds and marketing majors. The two variables relating to sensitivity to outside influences do show a difference. H6, which the data support, states that undecideds attach more importance to both peer pressure and parental influence in choosing their majors as compared to either marketing or nonmarketing majors. While 72% of marketing majors and 65% of nonmarketing majors do not consider peer pressure important, only 44% of undecideds feel this way. Of the two outside factors, peer pressure is apparently more important as a predictor of whether students will be undecided in the choice of their majors, as the parental influence variable is no longer statistically significant when peer pressure is included in the model.

Hugstad, Paul (1997). "Marketing the Marketing Major." Journal of Marketing Education, Spring: 19-4-13.

Simonoff, Jeffrey S. (1998). "Logistic Regression, Categorical Predictors, and Goodness-of-Fit: It Depends on Who You Ask." American Statistician, February: 52, forthcoming.

Table 1. Summary of logistic regression model fits.

Table 2. Percentage of students who have chosen a major that chose marketing, separated by their views of the importance of prestige and glamour of a career in choosing a major.