Relative Size versus Controlling for Size

The corpus callosum is the main tract of fibers providing communication between the left and right cerebral hemispheres. The question of a sex difference between human males and females in the size of the corpus callosum relative to the size of the brain has been actively debated for almost 20 years. One group of investigators notes that when a ratio is calculated with corpus callosum size in the numerator and brain size in the denominator, the value is, on average, larger in females. The other group argues that ratios are misleading on this question and that better statistical methods, such as analysis of covariance, do not indicate a sex difference. I show that this debate has been unnecessary. The issue is resolved by recognizing the difference between statistical control, in which focus is on a single trait for which a second trait is defined as a confounding effect, and proportionality, which focuses on the relationship between two traitsthe relationship that statistical control serves to remove. The calculation of ratios is not effective for statistical control but is essential for the evaluation of proportionality. Interpretation of the ratio is affected in important ways by dimensionalitythe units of the numerator and denominator (lengths, areas, or volumes)and this too has been mostly overlooked. A metaanalysis of published data indicates that the corpus callosum is relatively larger in females. However, the mean difference between sexes is small, males and females mostly overlap in relative size, and we do not know what the biological implications might be.