The conjunction fallacy in profiles of victims of homicide

Using basic probability theory, estimates of the characteristics of the average homicide victim are calculated using the notion of disjoint probability. The assumption of disjoint events (e.g., the victim's race bears no effect on the offender's weapon choice) is then tested empirically using the Uniform Crime Report‐Supplementary Homicide Report. Exploratory results suggest that many demographic and situational characteristics taken together are only slightly more related than chance. Put simply, the average profile of homicide victims portrayed by the media becomes less likely as demographic variables are added. A survey was conducted to test whether individuals conjoined these characteristics, thinking they were more likely to occur together. Consistent with the conjunction fallacy, many participants overestimated the likelihood that certain demographic or situational characteristics will occur together, and some overestimated it to a mathematically impossible degree. These two experiments showcase the difficulty in displaying statistical profiling and how it affects the public's perception of offenders.