The Male Advantage: Why women can't resist the Outlier Male

Product Information

So far, we have emphasized the relevance of potential sex differences in visual-spatial working memory to their relatively narrow implications for spatial abilities research. However, visual-spatial working memory tasks also are used widely for assessment and theory building in clinical (Alonso-Recio, Martín-Plasencia, Loeches-Alonso, & Serrano-Rodrigues, 2014; Barrett, Kelly, Bell, & King, 2008), developmental (Almela, van der Meij, Hidalgo, Villada, & Salvador, 2012; Teixeira, Zachi, Roque, Taub, & Ventura, 2011), and purely experimental settings (Hegarty, Montello, Richardson, Ishikawa, & Lovelace, 2006; Martin & Chaudry, 2014). Thus, visual-spatial working memory applies to many contexts and knowing whether it produces sex differences has important implications for how data are interpreted in these various contexts. Essentially, if we determine that sex differences exist in visual-spatial working memory, it might require an adjustment of norms in clinical settings, consideration of potentially different developmental trajectories for males and females, and additional components to existing theories. Yet, sex as a factor often is ignored in this area of research, as we discovered in our literature search. This means that the implications of the present paper go beyond establishing a potential link between sex differences in visual-spatial working memory and spatial ability. This paper also has implications for how we use visual-spatial working memory in clinical, developmental, and experimental settings. Mathew, J., Bernier, P.-M. & Danion, F. R. Asymmetrical relationship between prediction and control during visuomotor adaptation. eNeuro. https://doi.org/10.1523/ENEURO.0280-18.2018 (2018). Cansino, S., Hernández-Ramos, E., Estrada-Manilla, C., Torres-Trejo, F., Martínez-Galindo, J. G., Ayala-Hernández, M., & Rodríguez-Ortiz, M. D. (2013). The decline of verbal and visuospatial working memory across the adult life span. Age, 35(6), 2283–2302. doi: 10.1007/s11357-013-9531-1

Because location seems to be a central component of so many tasks, it is important to distinguish the studies sampled here from the ones that were sampled by Voyer, Postma, Brake, and Imperato-McGinley ( 2007) in their meta-analysis of sex differences in object location memory. Their analysis focused mostly on tasks similar to the one proposed by Silverman and Eals ( 1992), in which a large array of objects is memorized and, after an intervening object identity memory task, participants are tasked to identify moved and unmoved objects. Essentially, the time interval between encoding and retrieval is typically too long and the number of objects is too numerous to fit within the limits of working memory (Cowan, 2008). Therefore, the location tasks included in the present analysis are distinct from the type of task that was discussed by Voyer et al. ( 2007). Age of participants Bargary, G. et al. Individual differences in human eye movements: An oculomotor signature?. Vis. Res. 141, 157–169 (2017). Maselli, A. et al. A whole-body characterization of individual strategies, gender differences and common styles in overarm throwing. J. Neurophysiol. https://doi.org/10.1152/jn.00011.2019 (2019). A number of variables were coded as factors that might moderate sex differences in visual-spatial working memory. Specifically, characteristics relevant to the samples themselves (sample level variables) and factors inherent to the tasks used in each study (measure level variables) were considered. Sample level variables

References

Hernández-Balderas, M. Á., Rángel-Félix, G., Zavala-González, J. C., Romero-Romero, H., Silva-Pereyra, J. F., del Rio-Portilla, I. Y., & Bernal-Hernández, J. (2012). Sex differences in the visuospatial sketchpad in scholar children. Journal of Behavior, Health & Social Issues, 4(2), 103–115. doi: 10.5460/jbhsi.v4.2.34111 With 25 samples (36 effect sizes), the data obtained with tasks measuring memory for patterns were analyzed with multilevel modeling. However, this analysis failed to reveal any moderator accounting for significant variance in effect sizes (all p> 0.24), despite a significant variance component, χ 2(24) = 59.02, p< 0.001.

Balasubramanian, S., Melendez-Calderon, A., Roby-Brami, A. & Burdet, E. On the analysis of movement smoothness. J. NeuroEng. Rehabil. 12, 112 (2015). Despite their small size, the direction of sex differences is fairly consistent in the present analysis, with only one category reflecting a significant female advantage (memory for location in Table 2). This consistency emphasizes the importance of not disregarding completely any small effect as it might have implications in specific contexts. For example, the mere existence of sex differences in visual-spatial working memory should warn clinicians to be cautious when comparing male and female patients. In fact, if the test publisher feels that there is a need to provide separate norms for the Wechsler Adult Intelligence Scale—Third Edition (WAIS-III; Wechsler, 1997) despite controversial data concerning the existence of sex differences in intelligence (Halpern, 2013), a demonstrated case as we found here should clearly warrant separate norms for males and females. Similarly, it might be advisable to account for the sex difference in building models of visual-spatial working memory. Essentially, much empirical research is required before we can state without a doubt that males and females process visual-spatial working memory tasks in the same way. In the meantime, it might be advisable to examine model fit separately for male and female samples. Results of the overall analysis clearly show a male advantage in visual-spatial working memory with an estimated effect size of 0.155 that is significantly different from zero. However, this also reflects a small effect that essentially fits with the notion that most sex differences are small and that similarities tend to be the norm (Hyde, 2005; 2014). To put this finding in context, an effect size of 0.155 would require a total sample of 654 participants to achieve 80 % power at the 0.05 level of significance. This is much larger than the typical individual study sample size as presented in Table 1. This suggests that open archiving of data might provide an interesting avenue for researchers in that it would provide a cumulative record of available data that would likely improve our understanding of the underlying factors.The second step in data analysis considered the fact that many of the moderator variables that were discussed earlier are confounded with specific task. Therefore, this second step involved the application of mixed-effects meta-analysis (when all effects sizes were independent or did not include enough samples to compute robust standard errors (Raudenbush & Bryk, 2002) or multilevel modeling (in the case of non-independent effect sizes with enough samples to compute robust standard errors) to examine the influence of moderator variables within each specific task. van Donkelaar, P. & Lee, R. G. Interactions between the eye and hand motor systems: Disruptions due to cerebellar dysfunction. J. Neurophysiol. 72, 1674–1685 (1994). Evardone, M., & Alexander, G. M. (2009). Anxiety, sex-linked behaviors, and digit ratios (2D:4D). Archives of Sexual Behavior, 38(3), 442–455. doi: 10.1007/s10508-007-9260-6 Alonso-Recio, L., Martín-Plasencia, P., Loeches-Alonso, Á., & Serrano-Rodríguez, J. M. (2014). Working memory and facial expression recognition in patients with Parkinson’s disease. Journal of the International Neuropsychological Society, 20(5), 496–505. doi: 10.1017/S1355617714000265

Examination of the combined effect size was performed by computing a null model where the test of significance for the intercept is examined (Raudenbush & Bryk, 2002). Results of this analysis revealed a mean estimated d of 0.155 (95 % confidence interval (CI) = 0.087-0.223), indicating that males significantly outperformed females on visual-spatial working memory tasks, t(97) = 4.54, p< 0.001. Egger, M., Davey Smith, G., Schneider, M., & Minder, C. (1997). Bias in meta-analysis detected by a simple, graphical test. British Medical Journal, 315, 629–634. doi: 10.1136/bmj.315.7109.629 Ingalhalikar, M. et al. Sex differences in the structural connectome of the human brain. Proc. Natl. Acad. Sci. U. S. A. 111, 823–828 (2014). Gallivan, J. P., Chapman, C. S., Wolpert, D. M. & Flanagan, J. R. Decision-making in sensorimotor control. Nat. Rev. Neurosci. 19, 519–534 (2018).Aside from pointing out once more the distinct nature of location memory, this reanalysis suggests that when visual-spatial working memory tasks show a male advantage, this advantage appears first in the category that includes the age associated with puberty (13 years)—a finding similar to what Voyer et al. ( 1995) reported with spatial abilities. In accounting for this finding, hormonal changes, the most obvious factor associated with puberty, provides an interesting option. After all, there is ample evidence of the influence of sex hormones on visual-spatial working memory in humans, although the data remain unclear concerning the respective role of estrogen and progesterone (Duff & Hampson, 2000; Postma, Winkel, Tuiten, & van Honk, 1999) or testosterone (Cherrier et al., 2001) and whether these hormones improve or impair visual-spatial working memory. This is clearly an area that requires much more research. Martin, G. N., & Chaudry, A. (2014). Working memory performance and exposure to pleasant and unpleasant ambient odor: Is spatial span special? International Journal of Neuroscience, 124(11), 806–811. doi: 10.3109/00207454.2014.890619 Watson, N. V. & Kimura, D. Right-hand superiority for throwing but not for intercepting. Neuropsychologia 27, 1399–1414 (1989). Critics decry all this as unfair. Sex is one of the most powerful and consistent predictors of performance. Males have significant advantages over females in nearly every sport (see chart). Some teenage males outperform women’s world-record holders. In running, they are consistently around 10% faster than females, whether in sprints or the marathon. In Ms Hubbard’s sport of Olympic weightlifting, the gap is 25% or more, even after correcting for an athlete’s size. The Olympic record for females in the 69kg bodyweight category, for instance, is a total of 286kg. For males of the same weight, it is 357kg. Without sex categories, females would be absent from most elite sport. Because year of publication was coded routinely, it was included as a potential moderator. This factor often is interpreted as an indirect way to assess how social changes might promote fluctuations in sex differences (Feingold, 1988). Measure level variables