Intersectionality and quantitative methods: assessing regression from a feminist perspective

Nicholas A. Scott,Janet Siltanen

DOI: https://doi.org/10.1080/13645579.2016.1201328

2016-07-21

International Journal of Social Research Methodology

Abstract:This article examines the use of quantitative methods to advance feminist-inspired understandings of intersectionality. We acknowledge a range of conflicting opinions about the suitability of current quantitative techniques. To contribute to this debate, we assess the conceptualizations of intersectionality embedded in the most common approach to quantitative analysis, multiple regression. We identify three features of intersectional analysis highlighted in the feminist literature: (1) attention to context; (2) a heuristic approach to identifying relevant dimensions of inequality; (3) and addressing the complex, multidimensional structuring of inequality. Using these criteria, we evaluate: (1) multiple regression including context as a higher-order interaction; (2) multiple regressions run within different contexts and compared; and (3) multilevel regression including context as a higher-order level of analysis. We demonstrate with research illustrations that the models do a progressively better job at satisfying the criteria. We conclude that the third model offers a conceptualization of intersectionality that is the most consistent with the feminist literature.

social sciences, interdisciplinary