The emotional state of the police strongly influences how they interpret evidence – the investigation can go in a completely wrong direction researcher says
According to associate professor Karl Ask the emotional state of the police influences the interpretation of evidence. A sad policeman examines evidence more thoroughly than an angry policeman.
Why has the murder of Olof Palme not been solved?
Why was Thomas Qvick convicted of eight murders until it was realized that he had not committed even the first one?
Why were the police convinced that two little boys had killed four-year old Kevin in 1998, even though they didn’t?
Associate professor Karl Ask has an idea about why things went wrong in three of the criminal investigations that shook up Sweden. He considers the fundamental problem to be that the police got stuck on a hypothesis and forgot to consider other possible explanations for what happened.
For example the Palme-investigators had theory favorites.
“First they were on one track, then another. The chances of solving a crime grow smaller the more time passes. This is why crime investigators need to work on a wide front from day one” Ask emphasizes.
In June he visited an international forensic psychology conference that Åbo Akademi University hosted in Turku.
Ask, who works at the psychology department of the University of Gothenburg, has researched the police’s decision making processes, i.e. how the police looks for evidence, how they interpret it and make decisions based on it.
“We are especially interested in if there are systematic causes for mistakes that lead to the wrong conclusions” Ask formulates.
Causes for mistakes have indeed been found. Strong emotions, time pressure and expectations for instance, easily lead to the investigations only focusing on one single hypothesis or suspect. This could lead to a completely wrong conclusion.
Ask speaks of a study he conducted with his colleagues in which experienced crime investigators participated. The police had to conclude what had happened based on a case description.
Before the policemen were allowed to read the material the researchers manipulated their emotional state. Half of the group were manipulated to be angry, half to be sad.
“It was quite easy actually. We asked them to think of an event in which they had experienced very intense anger or sadness. They got to think about the event for a while, and afterwards they wrote a description of it and an assessment of what feelings the event evoked in them, and how intensely.”
After this the policemen read the case description of the crime. According to the description, which was based on true events, a 15-year-old boy had been assaulted. He was unconscious at the hospital, so he could not be asked about the perpetrator. The father was a suspect.
When the policemen had read the description they were given a new piece of evidence: a longtime neighbor had heard the fight and described the voices they had heard.
Half of the police were told that the neighbor recognized both the boy’s and father’s voices. The other group was told that, according to the witness, the second voice could not have belonged to the father but that it belonged to a younger man.
Ask and his colleagues were interested in how the policemen’s own emotional state influenced the interpretation of evidence.
The difference was significant. Angry policemen disregarded the witness statement and held on to their own preconceptions. The sad ones on the other hand paid a lot of attention to the witness testimony. If the witness statement contradicted the preconception they started to doubt the father’s guilt.
“The conclusion is that a sad state of mind leads to more a thorough examination of evidence. An angry state of mind on the other hand seems to lead to quick, simplified thinking were you hold on to your preconceptions”, Ask encapsulates.
“Many probably recognize this about themselves. When you are angry things are black and white. You can’t absorb information and instead you are convinced that you know what happened. When you are sad you think more about what has happened here and how it could have happened.”
Ask notes that even according to many other studies presumtions have a big impact on how the police investigate a case. It might even significantly affect the interpretation of details.
Ask gives an example of this.
A group of policemen read a case description again. According to it, a woman who had worked as a psychiatrist had been murdered in an apartment in which she also had her office.
Half of the policemen were told that the woman had received threatening phone calls. The woman had reported the phone calls and thought that the caller was one of her former patients. However, the woman was not able to name the caller.
The other group was told that the woman had been having a romantic affair with a patient and that the patients spouse had shown jealousy.
“We therefore brought up two motives, two suspects. Then the policemen got to read exactly identical case descriptions of what had happened.”
According to the case description, when the police arrived at the crimes scene the door to the apartment was opened by a woman who had cuts on her hand herself. The door was locked from the inside.
The policemen who knew about the threats over the phone interpreted this as the woman having locked the door to stop the killer from returning to the apartment. The policemen who had heard the jealousy story interpreted it as the woman obviously being the perpetrator.
“Two completely different conclusions were drawn from the same fact.”
The systematic biases of criminal investigations can according to Ask be avoided through proper work procedures. If there are none, the risk for misinterpretations grows larger.
However, he stresses that the experimental conditions created by the researchers are not completely equivalent to the way the police work in the field. Nevertheless the research gives indications as to how an investigation can end up on the wrong tracks because of cognitive bias.
There is however no complete template to avoiding the problems, Ask apologizes.
“There is no tested model for completely avoiding mistakes. It is an exciting time for research as different models’ operability are being tested now.”
There are promising examples. According to Ask, one of them is a method that is used in the investigation of crimes committed against children in Finland.
Docent of forensic psychology Julia Korkman says that the goal in Finland in these situations is to test different hypothesis.
“We consider what else could explain the suspicion if the crime did not occur” she elaborates.
Korkman is a researcher at Åbo Akademi University and has been involved in developing procedures at the Forensic Psychology Center for Children and Adolescents at HUS.
For instance, a small child’s sexually oriented playing does not necessarily indicate abuse, the child could have been exposed to sexual content by mistake, online for example.
In custody battles you always have to take into account the possibility of one of the parents trying to influence the child’s testimony.
“Hypothesis testing can in my opinion work in favor of a wrongly accused person as well as a real victim. If alternative hypothesis are not tested they can of course not be ruled out” Korkman emphasizes.
According to her the system does however not completely eliminate the risk of confirmation bias. Confirmation bias means that the individual only sees the evidence that support his or her preconceptions.
“The system does of course not guarantee that the assessments are always correct, the court should always critically assess them as well. Nevertheless this is a step in the right direction” Korkman says.
Karl Ask wishes the research results over time will have an influence on the police’s way of working. So far the Swedish police has not reacted to them very quickly. The Norwegian police on the other hand has been more active.
The Norwegian police has been particularly influenced by Ask’s and his colleague Ivar Fahsing’s research in which Norwegian and British criminal investigators ways of working were compared.
Norwegian and British policemen in the study were asked to come up with as many explanations to a person’s disappearance as possible.
“Ordinary people usually only come up with one explanation: the individual was kidnapped and most likely killed”, Ask says amused.
The more clever ones realize that there are other alternatives as well. The person could have disappeared voluntarily, been in an accident or gotten ill. In reality there are dozens of alternatives.
“We had an expert group, which came up with all fathomable scenarios, so we had an ideal performance.”
In the study they notices that there was no difference between young and experienced criminal investigators in Norway. They came up with quite a small number of hypothesis.
“So there is no advancement in Norway.”
In Britain on the other hand the novices were bad but the experienced criminal investigators were exceptionally good. They came up with a great deal of different hypotheses.
The most likely reason for the results is the difference in criminal investigators’ training.
In Norway, like in Sweden and Finland, the police training is quite long and you become a criminal investigator without a specialized training program.
In Britain on the other hand the basic training for policemen is quite short. At an early stage the young policeman decides what to specialize in. If he decides to specialize in criminal investigation there is a comprehensive training program for that. Further training is also continuously available.
In British training searching for alternative hypotheses is emphasized and there are strict requirements on documentation. For example, the police have to precisely log why someone has become a suspect of a crime. This makes the system more transparent, and the investigator also feels personally accountable for the decisions.”
Ask’s and Fahsing’s research results have already caused enthusiasm in Norway.
“The change has happened really fast. Our research was finalized a few years ago, and already the highest police management is looking into how training for criminal investigators can be improved in Norway. We have not seen this in Sweden.”
Why did Jari Aarnio’s subordinates not see their boss’s crimes?
Cognitive bias also has an impact on people not seeing the truth right under their nose. The police are no different in this regard than other people.
This is why associate professor Karl Ask does not find it surprising that the investigation into former chief of the Helsinki drug squad, Jari Aarnio, did not begin on account of his subordinates.
Ask reminds us that we are all guided by our expectations. These do not include a policeman straying into the criminal underworld.
“Because of this, especially in the police force, it was not possible to even imagine anything like this. I couldn’t imagine either that my closest colleagues would cheat in their research, even though I have read about well-known scientific frauds” Ask compares.
If you do not expect to see something, it is hard to see. Furthermore the individual might interpret even alarming signs according to their own preconceptions. This is why Aarnio’s surplus of cash did not raise suspicion.
“In these situations we choose the explanation that fits our preconceptions”.
Aarnio’s subordinates might have also thought they might end up in an odd light if the unit’s chief turns out to be a criminal.
“They might have had a subconscious motive to maintain a positive image of their workplace”, Ask notes.
Ask thinks that some of Aarnio’s subordinates must have noticed something strange in Aarnio’s activity, but that they for some reason or another did not want to bring it up. Groupthink could have influenced this: in homogenous groups thinking becomes more uniform.
“If you present a dissenting view in this kind of group, you easily can become silenced or marginalized, pushed out of the group. It’s not that strange”.
Docent of forensic psychology Julia Korkman agrees with Ask’s assessment. She mentions her investigation from e few years back in which a priest was suspected of abusing young boys for several decades.
Many had seen very suspicious things but not intervened, Korkman says.
“They saw things but did not believe their eyes. Many said that they would never have believed the priest in question could have done such things.”