Research using techniques from social network analysis have expanded dramatically in recent years. The availability of network data and the recognition that social network techniques can provide an additional perspective have contributed to this expansion. Social network data are not always in a standard network form and, in many instances, consists of two distinct groups with ties between groups and no within group ties. For example, people attending events or meetings, authors collaborating on research outputs, or directors on boards of companies. Such data are known as two-mode data. Recently, Everett and Borgatti suggested a general approach for analyzing two-mode data. They suggested forming two one-mode data sets, analyzing these separately, and then recombining the results using the original data. One under-explored area in their work is in how this method can be applied to centrality problems; an issue we seek to begin to address here.