Research Article

The Distance Between Mars and Venus: Measuring Global Sex Differences in Personality

  • Marco Del Giudice mail,

    Affiliation: Department of Psychology, University of Turin, Torino, Italy

  • Tom Booth,

    Affiliation: Manchester Business School, University of Manchester, Manchester, United Kingdom

  • Paul Irwing

    Affiliation: Manchester Business School, University of Manchester, Manchester, United Kingdom

  • Published: January 04, 2012
  • DOI: 10.1371/journal.pone.0029265

Reader Comments (14)

Post a new comment on this article

The Distance Between North Dakota and South Dakota

Posted by jshyde on 05 Jan 2012 at 02:53 GMT

In their article, The Distance Between Mars and Venus: Measuring Global Sex Differences in Personality, Del Giudice, Booth, and Irwing challenge my Gender Similarities Hypothesis in the case of personality. Below I show that their methods lead to uninterpretible findings that fly in the face of contemporary personality theory. The Gender Similarities Hypothesis is still accurate and supported by massive amounts of data.
The main innovation in the Del Giudice paper is to introduce the use of Mahalanobis D to the measurement of the magnitude of gender differences. A staple of multivariate statistics for decades, D in this application measures the distance between 2 centroids in multivariate space [1]. It is a multivariate generalization of the d statistic used in many meta-analyses. What is not apparent from the Del Giudice paper, however, is that D is computed by taking the linear combination of the original variables that maximizes the difference between groups. What they have shown is that, if one takes a large enough set of personality measures and then takes a linear combination to maximize gender differences, one can get a pretty big gender difference. That is all they have shown – no more, no less.
An assumption of multivariate normality is crucial to Mahalanobis D if it is to be accurate [2]. The authors provide no statistical verification that their variables are distributed multivariate normally. In other research, apparent findings of large gender differences have crumbled when appropriate statistical methods were used for the non-normal, skewed distributions [3].
The gender difference that Del Giudice and colleagues have found is along a dimension in multivariate space that is a linear combination of the original variables transformed into latent variables. A point that is not mentioned in the Del Giudice article is that this dimension is the first discriminant function. Aside from the fact that the linear combination introduces bias by maximizing differences, the resulting dimension here is uninterpretible. What does it mean to say that there are large gender differences on this undefined dimension in 15-dimensional space created from latent variables? The authors call it global personality, but what does that mean? They promise to measure personality with greater “resolution,” yet in the end they have a single, undefined dimension of personality. They have blurred the question rather than offering higher resolution.
Reducing gender differences in personality to differences on a single dimension also flies in the face of contemporary personality theory, which emphasizes distinct dimensions of personality. There is abundant evidence for the Big Five factors of personality: Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscientiousness [4]. Others would argue for 16 personality factors, captured by the 16PF used in the Del Giudice study [5]. No personality theorists, to my knowledge, are arguing for a single factor or dimension, yet Del Giudice and colleagues base their findings on just that.
Another important point to note is that Del Giudice and colleagues’ methods rely on subjective self-reports of personality. When Del Giudice and colleagues talk about “error-free” scores, they are using psychometric terms that may mask the fact that the data are still based on subjective self-reports. As an example, Feingold’s meta-analysis [6] found gender differences in anxiety ranging in magnitude between d = -.15 and -.32. That is, the differences were small, with females scoring higher. All of the data were based on self-report personality inventories. In a meta-analysis of research on gender differences in temperament – some of it based on parent or other adult report, some of it based on behavioral measures – the effect size for the gender difference in fear was d = -0.12, i.e., a smaller difference than Feingold found for self-report measures [7]. Moreover, a behavioral study measuring children’s distress to the insertion of an intravenous needle showed no significant gender difference [8]. That is, the boys were as anxious and fearful as the girls. Too much of the research on gender differences has relied on subjective self-reports, when objective, behavioral measures may show much different results.
Moreover, subjective self-reports are vulnerable to the effects of stereotyping, and many personality traits are gender stereotyped [9]. Research indicates that people engage in gender-based self-stereotyping when reporting personality traits [10]. The two personality factors that show the largest univariate gender differences in the Del Giudice study are Sensitivity (d = -2.29) and Warmth (d = -.89). It is no accident that warmth and sensitivity are highly female stereotyped traits. Both of them, for example, are items on the Femininity scale of the Bem Sex Role Inventory [11] and on Spence’s Personal Attributes Questionnaire [12]. Most women would be reluctant to rate themselves as low on sensitivity and warmth. Many men, in contrast, would be loathe to describe themselves with the girlie attribute of sensitivity. In short, the larger gender differences found by Del Giudice and colleagues may represent no more than gender stereotypes, and may reveal little about actual behavior.
Finally, let us return to the original Gender Similarities Hypothesis [13]. It states that men and women are more similar than different on most (but not all) psychological variables. Evidence came from a review of 46 meta-analyses of psychological gender differences. Of the 124 effect sizes for gender differences extracted from the meta-analyses, 30% had d values falling between 0 and 0.10 (the trivial range) and an additional 48% fell between 0.11 and 0.35 (the small range). These effect sizes represented gender differences across a wide array of psychological variables, including mathematical problem solving, reading comprehension, attributions for success and failure, aggression, and leadership. Del Giudice and colleagues have focused on one particular topic, personality, and, using a technique to maximize differences between groups, they have found a large gender difference on a single, undefined dimension of personality. They argue that males and females are as different in personality as the distance between the planets Mars and Venus. Instead, the overwhelming evidence, across multiple psychological domains, is that men and women are more similar than different; the distance between them is more like the distance between North Dakota and South Dakota.

1.Stevens JP (2009). Applied multivariate statistics for the social sciences. 5th ed. New York: Routledge.
2.Overall JE & Klett CJ (1972). Applied multivariate analysis.New York: McGraw-Hill.
3.Pedersen WC, Mill LC, Putcha-Bhagavatula AD, Yang Y (2002). Evolved sex differences in the number of partners desired? The long and the short of it. Psych Sci, 13, 157-161.
4.McCrae RH, Costa PT (1997). Personality trait structure as a human universal. Amer. Psychologist 52: 509-516.
5.Conn SR, Rieke ML (1994) editors. The 16PF fifth edition technical manual. Champagne, IL: Institute for Personality and Ability Testing.
6.Feingold A (1994). Gender differences in personality: A meta-analysis. Psych Bull 116: 429-456.
7.Else-Quest NM, Hyde JS, Goldsmith, HH, Van Hulle, CA (2006). Gender differences in temperament: A meta-analysis. Psych Bull 132: 33-72.
8.McCarthy AM Kleiber C, Hanrahan K, Zimmerman MB, Westhus N, Allen S (2010). Factors explaining children’s responses to intravenous needle insertions. Nurs Res 59: 407-416.
9.Gerber GL (2009). Status and the gender stereotyped personality traits: Toward an integration. Sex Roles 61: 297-316.
10.Oswald DL, Chapleau, KM (2010). Selective self-stereotyping and women’s self-esteem maintenance. Pers Indiv Diff 49: 918-922.
11.Bem SL (1974). The measurement of psychological androgyny. J Consult Clin Psych 42: 155-162.
12.Spence JT, Buckner CE (2000). Instrumental and expressive traits, trait stereotypes, and sexist attitudes. Psych Women Quart 24: 44-62.
13.Hyde JS (2005). The gender similarities hypothesis. Amer Psychologist 60: 581-592.

The author thanks Kelley Kidwell and Michael Newton for their very helpful comments on an earlier draft of this paper.

Competing interests declared: I am the author of the Gender Similarities Hypothesis.

RE: The Distance Between North Dakota and South Dakota

Bootlegger replied to jshyde on 05 Jan 2012 at 03:26 GMT

<em>What does it mean to say that there are large gender differences on this undefined dimension in 15-dimensional space created from latent variables? </em>

Indeed! The authors seem to gloss over quite a few of their assumptions.

No competing interests declared.

RE: What is the Overlap Between North Dakota and South Dakota

thalia replied to jshyde on 05 Jan 2012 at 23:34 GMT

While your metaphor is provocative, I think it is misleading to your theory. You could state, "The Rocky Mountains are intrinsic to the geology of the area; the Rocky Mountains do not alter at the state line between North and South Dakota."

While the distance between N Dakota and S Dakota may not be large, a person can only be a citizen of one of those states.

No competing interests declared.