ABSTRACT
This chapter expands upon findings in the Fox and Farrington article in Psychological Bulletin, which presents a systematic review and meta-analysis of 426 publications on offender profiling from 1976 to 2016. The chapter begins with a departure from the original article to more fully discuss the origins of evidence-based offender profiling (EBOP), tracing the history and inspiration back through to David Canter’s formation of the sub-field of investigative psychology in 2000. The assumptions, analytical techniques, and major findings of research from this field are reviewed. Next, the statistical profiles developed by Farrington and Lambert in the 1990s are discussed to illustrate the transition from the use of quantitative statistics to not only develop the profiles but also to analyze the statistical links between certain types of offenses and certain types of offenders. Finally, the EBOP method, which is discussed in more detail in Chapter 4, is reviewed, so that readers understand the need for EBOP, the basic principles of the new field, and how EBOP is applied in academic research and police investigations.
Some key findings from the meta-analysis of offender profiling literature are discussed, such as the most common types of crimes that have been studied were homicides (27%) and sexual assaults (13%), and most research was conducted in the United States (38%) or the United Kingdom (26%). The most prolific authors have been Richard Kocsis, Laurence Alison, and David Canter. Over time, the scientific rigor of studies has increased. In regard to authors, 47% of reports in 1976 to 1985 were written by Federal Bureau of Investigation profilers, but none in 2006 to 2016. In contrast, only 9% of reports in 1976 to 1985 were written by psychologists, but this increased to 47% in 2006–2016. The most common offender profiles that have been proposed for different types of crimes are presented, and unlike in the original article, a full discussion of the resulting and recurring profile types across each crime group and potential reasons for the lack of replicability in offender profiling research are discussed.
It is concluded that future offender profiles should be based on an empirical approach that relies on advanced statistical analysis of large datasets and that profiles should be implemented in police investigations and evaluated for their effects on arrest rates in unsolved cases using an experimental design, as EBOP recommends.
