“…However, in the vignette comparisons, the woman's reputation in the community (whether she was faithful or rumored to be having an affair; whether she was well liked or not) was highly influential for a decision to prosecute, as were the dead man's family's preference for her to be prosecuted, and whether an advocacy group was involved, highlighting a potential discrepancy between intentions and actual impact. Supporting the likelihood of extraneous variables influencing decisions to prosecute is the match between those factors that did influence decision makers' likelihood to prosecute with anecdotal literature warning that particular stereotypes or case discrepancies with self-defense criteria may influence outcomes (e.g., Gillespie 1989) or with factors previously identified through quasi-experimental research manipulating personal variables of battered women defendants as problematic variables for these cases (e.g., Follingstad et al 1996). The fact that this study identified personal aspects of the battered woman (i.e., her size, her drinking, her reputation), or the preference of the deceased man's family, or the woman's demeanor after the killing as producing significantly different orientations toward prosecution strongly suggests the need for greater understanding of these discretionary decisions.…”