On sex/gender related similarities and differences in fMRI language research.

Again and again, attempts have been made to find correlates of sex/gender differences in the human brain. Despite the insistence with which differences have been stated, empirical results have not been unequivocal: evidence for and against the influence of sex in the makeup of men's and women's brains has been presented. This article focuses on the relevance of sex/gender related differences in fMRI research, especially with regard to language processing. By discussing some crucial criteria from fMRI examinations, we demonstrate the existence of paradigmatic, methodological and statistical defaults that interfere with assessing the presence or absence of sex/gender differences. These criteria are, among others, the use of contrast analyses, the function of the variable sex/gender as a co-item and the "publication bias". It is argued that dealing with the sex/gender variable will, at least to some degree, inevitably lead to the detection of differences rather than to the detection of similarities.