If college students are more narcissistic today than when I was in college, I haven’t seen it. In fact, my students seem even more darling today than they did in 1994 when I taught my first course at the University of Notre Dame. For instance, I used to be accused of giving “tricky” tests. But now students do not complain at all about my tough exams. A number of them stay after class to ask intellectual questions about that day’s lecture. And a great many of them seek service positions, such as Teach for America, after they graduate and before they launch their careers.
My enchantment with how good the students are today lines up with the fact that youth crime has decreased in recent years (U.S. Bureau of the Census, 2006). Moreover, high-school students’ volunteerism has increased (1).
How then could Jean Twenge and her colleagues conclude that college students today are more narcissistic than my generation was (2)? Their evidence for this claim was derived from college students’ scores on the Narcissism Personality Inventory (NPI) in 85 samples from 1982-2006. They found that the students’ average score on the 40-item inventory rose steadily from 1982 to 2006. It went from around 15 points to 17 points.
However, when the NPI scores were analyzed separately by the sex of the respondents in the 44 studies that had coded for sex, the results showed that the entire rise was accounted for by the increase in female college students’ scores. As Twenge and her colleagues pointed out, whereas women students were lower in narcissism than men in 1982, they had almost caught up to the men by 2006. However, the authors downplayed these findings, saying that the analyses “may not be representative.” The reason they gave was that a number of studies had not published their results by sex; thus they had gotten the means directly from the authors. Their dismissal of this subset of data is ironic, however, because just as those 44 samples are not necessarily representative of the population of college students, the 85 samples are not either! In neither case were the samples drawn at random from any broader population of college students, which precludes the generalization of either set of findings to college students.
If I had published those same data, I would have not concluded that the study offered evidence of a rise in narcissism among college students in general. The lack of random sampling to ensure that the samples represented the population of college students was a very big problem. After all, an entirely different set of results might come from a new set of studies done the exact same way due to the variability of the samples. This problem is known as sampling error. Moreover, given that the women drove the rise in narcissism in the 44 studies that analyzed the sexes separately, it might be appropriate to theorize about changes in college women per se. What looks like an increase in narcissism overall may merely be a bi-product of the growing equality and leadership roles that women have enjoyed over the years. Thus, college women today, as compared with my generation at college age, might be more inclined to pick the narcissistic choice in questions from the NPI such as “I am not sure if I would make a good leader” versus “I see myself as a good leader”; or “I am assertive” versus “I wish I were more assertive.”
So on behalf of my Generation Me Too, I apologize to college students everywhere for accusing them of being more narcissistic than we were as college students. As I’ve just shown, the empirical foundation for that claim must be called into question.
1. Bachman, J., Johnston, L., & O’Malley, P. (2006). Monitoring the future: Questionnaire responses from the nation’s high school seniors. Ann Arbor, University of Michigan.
2. Twenge, J.M., Konrath, S., Foster, J. D., Campbell, W. K., & Bushman, B. J. (2008). Egos inflating over time: A cross-temporal meta-analysis of the Narcissistic Personality Inventory. Journal of Personality, 76, 875-901.