On the use of Multiple Correspondence Analysis to visually explore affiliation networks

• We propose Multiple Correspondence Analysis to visually explore affiliation networks.
• We show how MCA assigns a different role to actors and events.
• We outline how the adopted doubling perspective improves result interpretation.
• We use asymmetric maps to jointly represent actors and events.
• We propose a way to add actor and event attributes to the analysis.

In this paper we discuss the use of Multiple Correspondence Analysis to analyze and graphically represent two-mode networks, and we propose to apply it in a Greenacre's doubling perspective. We discuss how Multiple Correspondence Analysis: (i) properly takes into account the nature of relational data and the intrinsic asymmetry of actors/events in two-mode networks; (ii) allows a proper graphical appraisal of the underlying relational structure of actors or events; (iii) makes it possible to add actor and event attributes to the analysis in order to improve results interpretation; and (iv) gives different results with respect to the usual Simple Correspondence Analysis.