Base Rate Fallacy

Understanding the Base Rate Fallacy

The tendency to ignore or underuse base rate information and instead to be influenced by the distinctive features of the case being judged is known as base rate fallacy.

The Base rate fallacy is a common cognitive error that skews decision-making whereby information about the occurrence of some common characteristic within a given population is ignored or not given much weight in decision making. Many cognitive errors are the results of people not paying attention to base rates.

Tversky and Kahneman (1973) demonstrated that people had a tendency to neglect base-rate or statistical information in favor of similarity judgments. In a typical study, the participants were asked to predict the field of study of a graduate or the profession of somebody on the basis of a brief description. The description contained some personality traits that were similar to the stereotype of a profession, for example, of lawyers or engineers.

As expected, the participants’ judgments turned out to be determined by the degree of similarity between the description and the stereotype of the profession. This happened even when the participants were made familiar with the base rates, that is, the frequencies of law and engineering students and professionals in the population. This is known as the base-rate fallacy.

This heuristic is often equated with the heuristic of representativeness: an even is judged probable to the extent that it represents the essential features of its parent population or of its generating process.

It means, among other things, that people in situations of uncertainty tend to look for familiar patterns and are apt to believe that the pattern will repeat itself. This tendency has important implications for understanding error judgments made by profilers.

For example, the profiler may focus on a specific offender, pushing into the background useful information about the population of offenders with similar characteristics.

Base rate neglect is especially likely to happen if the profiler encounters a case that s/he perceives is unique and outside the usual cases within a particular offense category. It also happens when the profiler believes s/he is better equipped for dealing with the case based on prior experience. The conclusion the profiler neglect or underweight the base-rate information, that is, s/he commit the base-rate fallacy.

Rainbow et al. (2011) provide an excellent example of how investigators and profilers may become distracted from the usual crime scene investigative methods because they ignore or are unaware of the base rate.

The case involved a 90-year-old woman who was found dead in her home. At the crime scene, her heart had been removed from her body and placed on a silver platter. Blood had been drained from her body and poured into a small container, which had the traces of lip marks on the rim. Candles had been arranged to suggest some kind of ceremony had occurred, and fireplace pokers were placed at her feet in the shape of a crucifix. Adding to the drama, the murder had happened on an island off the coast of Wales that was devoted with ancient Druid ruins. It would be tempting to view this as a horrific illustration of a cult-related murder and assume that a small group of individuals was involved. However, investigators in this case were wise enough to consider base rate data—who kills the elderly?

The Ultimate Managed Hosting Platform

As Rainbow et al. (2011) note,

In terms of prioritizing suspects, base rate information from research into elderly homicide together with a logical crime scene interpretation strongly indicated that the offender was likely to have some association to the victim and probably lived in close proximity. (p.44)

Forensic evidence, including a footprint left at the scene, led to the arrest a 17-year-old from the same village who had delivered newspapers to the victim’s door the previous 3 years and was aware that she had money and jewels stashed in her home.

Investigators concluded it was neither a ritualistic sacrifice nor an occult ceremony, but a straightforward robbery-murder situation. The 17-year-old killer, in an attempt to diver attention away from himself, set the stage to make it appear to be a mysterious ritualistic murder.

Expressions of Uncertainty

Few if any profilers would be so foolish as to indicate that the perpetrator definitely possessed certain characteristics. Almost invariably, they will make statements framed as probabilities, communicating that there is some uncertainty in their assessment.

According to Heuer (1999), however, probabilities of something happening may be expressed in two ways.

In many instances, subjective probability statements are ambiguous and misunderstood by police investigators. As Heuer reports, “To say that something could happen or is possible may refer to anything from a 1-percent to a 99-percent probability” (pp. 152-153). The profiler should communicate more clearly by placing a personal percentage on the prediction (i.e., 30%) so that investigators can judge how strongly the profiler believes the event will occur. Nevertheless, it should be emphasized that this is a probability, not a definitive prediction.

It is likely that clinically based profilers will resist the notion of attaching a percentage figure to their predictions—this seems to fly in the face of intuition or clinically judgment. Nevertheless, according to Heuer (1999), without such guidance, investigators may be inclined to interpret ambiguous probability statements as highly consistent with their own preconceptions of the case. It is very important that police investigators be open to alternative viewpoints, and it is equally important that profilers help create alternative ideas.

In this context, the profiler should be comfortable enough to consult with outside experts and colleagues whenever possible to formulate alternative perspectives. These colleagues may see things or ask questions that the profiler has not seen or asked. It is likely then, that a team of profilers working together will produce a more accurate profile than a lone individual.

The Ultimate Managed Hosting Platform