Abstract
In a recent commentary, DeSteno (2010) critiqued the work of Levy and Kelley (2010) which investigated the relationship of attachment style to the sex difference in jealousy. This commentary addresses the concerns raised by DeSteno; I briefly review some of the literature that was not addressed by DeSteno's commentary and discuss directions that future research may take.
In a recent commentary, DeSteno (2010) critiqued the work of Levy and Kelley (2010) which investigated the relationship of attachment style to the sex difference in jealousy. In his letter, DeSteno brought up several critiques of the methodology used by Levy and Kelley in investigating the sex difference in jealousy work. Levy and Kelley used the Buss, Larsen, Westen and Semmelroth (1992) hypothetical forced choice dilemma where participants are asked to select whether an emotional or sexual infidelity is worse. DeSteno argued that the sex difference in jealousy is an artifact of measurement and that hypothetical responses do not mirror actual jealousy patterns. However, these critiques originate from an incomplete reading of the literature. Numerous studies (e.g., Bohner and Wänke, 2004; Weiderman and Allgeier, 1993) have demonstrated that the sex difference in jealousy emerges when using continuous measures of jealousy (see also Sagarin, Martin, Coutinho, Edlund, and Patel's, under review, meta-analysis, which found a significant average effect, g* = 0.237, p < .00001, across 44 independent samples that used continuous measures). Further, other studies have demonstrated that the sex difference in jealousy emerges when looking at retrospective reports of actual infidelity experience (e.g., Edlund et al., 2006). As such, the basic argument that the sex difference in jealousy is an artifact of measurement is simply unfounded.
DeSteno (2010) correctly pointed out that the sex difference in jealousy emerges quite robustly when using forced-choice measures (see also Harris, 2003). It is this robustness that makes this methodology the ideal choice for use in conjunction with variables that would be hypothesized to moderate this effect (as done in Levy and Kelley, 2010). When working with continuous measures, numerous factors unrelated to the sex difference in jealousy can influence the pattern of results (such as ceiling effects; Edlund and Sagarin, 2009). The appeal of the forced choice measure is in the forced choice – participants have to choose which is worse rather than being allowed to say that both options are extremely distressing. A moderator that would attenuate or eliminate the effect might get lost in error variance in a continuous-response format; in a forced choice study, the effect would be allowed to emerge more clearly. As such, rather than dismissing the use of forced choice methods in the sex difference in jealousy research, the field should embrace them when looking at moderators.
Importantly, I concur with DeSteno's (2010) call for a greater focus on assessing jealousy in a real time context. There has been a paucity of studies investigating real jealousy from staged interactions (due in no small part to the ethical considerations of such an endeavor). Importantly, these kinds of studies would be wise to use multiple methods of assessing jealousy (such as: forced choice, free-response, Likert-style responses, and physiological measures) while looking at the moderators of the sex difference in jealousy.