Geometric Data Analysis (GDA) - an alternative approach to the analyses of gender differences

Resumé

The aim of this paper is threefold: First, the criticism of quantitative methods raised by feminist and gender researchers is reiterated and illustrated using gender differences in job attribute preferences as an example. Second, the paper compares this ‘standard quantitative methods’ approach to Geometric Data Analysis (GDA), an approach that e.g. makes use of principal components analysis. I argue that GDA breaks with many of the problematic features of traditional statistics by being multi-dimensional (as opposed to one-dimensional), having a statistical model formulated at the individual level (as opposed to treating individuals as mere ‘residuals’) and visualising the results (as opposed to just presenting the results exclusively in numbers). Third, the empirical analyses from the first part of the paper are then used as an example and analysed again, thereby introducing the basic concepts and principles which comprise GDA. Data used in the paper stem from the study Youth on the margin where a sample of young men and women from the North Denmark Region were asked to fill out a battery of job attribute preferences among other things. This is an important topic because such preferences are widely thought to be closely related to the continuing segregation of the Danish labour market.