Overcoming Selective Observation and Faulty Reasoning in Policing
Start a criminal justice degree at American Military University.
By Scot DuFour, alumnus, American Public University
There are things I have heard and seen during my policing career that seem to allude to certain truths, yet the evidence doesn’t stack up. For example, many cops hold the belief that critical incidents, like officer-involved shootings, come in sets of three. Or that our shift is going to be busy when there is a full moon. Similarly, most cops have, at some point, been told in roll call that they are doing a great job because crime statistics are down—but why do command staff assume that a decrease in reported crime is because of the efforts of patrol officers?
[Related: Avoiding Crusadism in Police Work]
These examples point to a flawed human psychology, which can also cause us to conduct incomplete criminal investigations. In order to conduct more fair and thorough investigations, officers need to understand and recognize normal cognitive biases like confirmation bias and the tendency to misunderstand things like statistical data.
Understanding Confirmation Bias and Selective Observation
Humans like to find consistency and patterns in their daily lives; it makes us feel better to think there is some kind of correlation or cause when things go right or wrong. Take the game of blackjack or poker, for example. Every real player knows that to win over the long-term, you must play the odds. Relying on a gut feeling might let you win here and there, but if you play with your gut feeling every hand, you will quickly be out of money.
As police officers, we try to mitigate the risks we take by utilizing sound tactical training and treating unknown situations as worst-possible scenarios until we can prove otherwise. An officer would never saunter into a hotel room where a man was reportedly armed with a rifle because they had a gut feeling he was not planning on shooting anyone that day. That being said, officers are still susceptible to making common errors in everyday reasoning.
One common error in reasoning is called confirmation bias or selective observation. This is when people, often subconsciously, only take note of things that align with their existing beliefs (Bachman & Schutt, 2015). For example, a cop who responds to three critical incidents within a work week will be quick to point out how it proves their theory that critical incidents happen in groups of three. That same officer likely does not take note of the next week when only one critical incident occurs.
[Related: Sentencing Bias: Why More Research Is Needed]
This same type of selective observation is why some officers might believe more crime occurs during a full moon. An officer working on a particularly busy night will say to himself, “it must be a full moon tonight!” and then, looking into the sky, see a full moon and conclude that his theory is correct. But busy nights without full moons are not noticed or counted against his theory.
Selective Observation Compromises Police Work
Humans are naturally biased and there has been increased attention on implicit bias in policing in recent years (Blake, 2016). For officers, there are many pitfalls of acting unethically or even just inefficiently during an investigation. For example, selective observation can lead to racism and discrimination at one end of the scale to complacency and poor investigations at the other.
Officers must actively work to stop selective observation from bolstering their implicit biases. The obvious example is in cases involving racism, but even an officer who believes “domestic violence victims never leave their abuser” or “the courts don’t prosecute offenders anyway” is in danger of ineffective policing.
Misunderstanding Statistical Data
Another common error in reasoning in addition to selective observation is misunderstanding statistical data. People in general do not understand what is called the clustering illusion. A clear example of the clustering illusion from Gilovich (1991) showed that in a series of 10 coin flips, there are often clusters of tails or heads. Over a long series of coin flips, there would be something close to a 50/50 split, but if we focus on the patterns we perceive based on a small sample, it can lead us to incorrectly believe those results are representative of the actual statistical probability. So while well-trained crime analysts are familiar with statistics, the average street cop is probably in danger of seeing the crimes he or she responds to as a pattern rather than a random distribution of events.
Another error related to the clustering illusion is failure to recognize statistical regression. For example, Gilovich (1991) provides the example of how people might believe a decrease in crime immediately following a crime wave is the result of a new law enforcement policy. This is because people tend to assign too much meaning to random events. People who have taken a statistics or research class know that statistical regression guarantees that extraordinary data will be followed by a deterioration in that data. Think of sports players who go on hot streaks or crime rates that go up and down. Crime rates always fluctuate and they always will so a short drop in crime rarely relates to the level of effort given by patrol officers. It also means that your police force is not suddenly underperforming if crime rates rise again, which they will.
Errors in Reasoning Can Lead to Investigative Mistakes
Cops are in the business of finding evidence to prove that a person committed a crime. But studies have shown that officers who believe an offender is lying cannot be made to change their mind easily even if there is evidence that says otherwise. Instead, officers who think an offender is guilty tend to choose further investigative steps that they believe will gather more evidence against the offender. Furthermore, the more severe the crime is the more easily officers are convinced of the accused’s guilt (Rassin, Eerland, & Kuijpers, 2010).
Exculpatory evidence, which is evidence that clears a person of blame, is an incredibly important tool to avoid arresting the wrong person. Cops should recognize their common errors in reasoning and actually focus on disproving their hypothesis. In the fields of science and academic research, to garner accurate results researchers are constantly trying to disprove their theories rather than find more examples that confirm their belief. The only things you risk by searching for exculpatory evidence is clearing the innocent of wrongdoing and building a stronger case against the guilty.
About the Author: Scot DuFour has been a police officer since 2004 and is currently a Field Training Officer with a police department in Colorado. He was previously an investigator in a domestic violence prosecutions unit for a district attorney’s office, a police officer with the Phoenix Police Department, and a Task Force Officer with the Drug Enforcement Administration. He is a graduate of American Public University with a bachelor’s degree in philosophy and a master’s degree in criminal justice. To contact the author, please email IPSauthor@apus.edu. For more articles featuring insight from industry experts, subscribe to In Public Safety’s bi-monthly newsletter.
Bachman, R., & Schutt, R.K. (2015). Fundamentals of research in criminology and criminal justice (3rd ed.). Thousand Oaks, CA: Sage Publications.
Blake, D. (2016). Unpacking implicit bias in policing. Policeone.com. Retrieved from https://www.policeone.com/patrol-issues/articles/239038006-Unpacking-implicit-bias-in-policing/
Gilovich, T. (1991). How we know what isn’t so: The fallibility of human reason in everyday life. New York, NY: The Free Press.
Rassin, E., Eerland, A., & Kuijpers, I. (2010). Let’s find the evidence: An analysis study of confirmation bias in criminal investigations. Journal of Investigative Psychology and Offender Profiling, 7, 231-246.