Skip to main content
Sydney Science Forum: Not Guilty

5 things we learnt in ‘Not Guilty: the psychology of crime investigations’

30 September 2019
The science of how our brains work in a crime scene
Dr Celine van Golde, from our School of Psychology in the Faculty of Science, gave a hugely popular Sydney Science Forum talk ‘Not Guilty: the psychology of crime investigations’ on 18 September. Here are five things we learnt from her.
Dr Celine van Golde

Why do wrongful convictions happen?

Wrongful convictions can and do happen – it's the unfortunate truth of the legal system. But it’s not just a legal matter; the science of how our brains work plays a big part too.

There are many reasons why wrongful convictions happen. In Australia, we don’t have a database that collects this sort of information, but other countries do, and that data shows some of the reasons wrongful convictions happen.

The Innocence Project in the US collects such data, and their database of cases shows how human error causes wrongful convictions. The data shows the top reasons for wrongful convictions are:

  1. eyewitness misidentification of suspect – 72% of cases
  2. unvalidated or improper forensics – 47% of cases
  3. false confessions or admissions from suspect – 27% of cases
  4. informants or snitches – 15% of cases

The percentages above add up to more that 100% as each one is the percentage of the total cases examined, showing that wrongful convictions can have more than one cause. 

In Australia, we’ve had high profile cases of wrongful conviction

The most famous case of wrongful conviction in Australia is Lindy Chamberlain, who was charged with murdering her baby daughter, Azaria, in 1980.

Lindy Chamberlain served three years in prison of her life sentence, and was later exonerated due to errors in the forensic science used in her case. A luminol test had been carried out in her car, which showed a large patch of fluid in the car, that investigators interpreted as the baby’s blood. Later, this fluid in the car was determined to be non-biological, and most likely a sound-deadening compound from a manufacturing overspray.

Other high profile Australian cases include John Button who was convicted of manslaughter of his girlfriend Rosemary Anderson by running her over with his car and he was later exonerated; and Sue Neill-Fraser who was convicted of murdering her partner of 18 years Bob Chappell by throwing him off a yacht and her case is currently being appealed.

Brain image and memory

Our memories can fail us

It’s clear that human error plays a big part in wrongful convictions, and our memories are especially to blame. Memories are important in criminal cases through the memory of eyewitnesses, suspects and victims.

Memory evidence is very influential in trials, as juries are more likely to convict based on eyewitness evidence than other forms of evidence, but unfortunately our memory can be very unreliable.

Memory is not like Netflix

Unlike Netflix, where you can go back to a precise series, episode and scene, and play it back exactly, your memories are reconstructed when we recall them. Your memory is also affected by many factors, such as your emotions or who you’re recounting your memory to.

There are three stages to our memories: encoding, storage and retrieval. Each stage is affected by numerous factors that influence how you will remember events.  

In general, within the first hour after an event happens, we lose quite a bit of detail in our recall memory of the event, and within 15 hours of an event, we lose around 70% of the details of the event.

We fill in the gaps in our memories with details we either assume to be correct or with details that others provide. That means we’re very open to leading questions or suggestions from others.

The ‘grandmother’ of forensic psychology memory research, Professor Elizabeth Loftus, did huge amounts of pioneering research on the malleability of memory, especially in the case of eyewitness memory. She showed that how people are questioned about what they saw, affects what details people respond with, including reporting false memories of details in a film of a car crash she showed.        

We’re also subject to confirmation bias, where we search for or interpret information in a way that confirms our beliefs, and we ignore information that contradicts our beliefs.

All this doesn’t mean that memory evidence is invalid in criminal investigations, but that we need to be very careful in how memories are elicited and how people are questioned about what they saw or experienced.

Not Guilty project

Psychology and law come together in Not Guilty project

The University of Sydney’s Not Guilty project brings together students and staff in psychology and law to work together to assess potential cases of wrongful conviction.

Psychology plays a very important role in righting wrongful convictions and also preventing them in the first place through policy reform.

Watch Dr Celine van Golde’s whole Sydney Science Forum talk 'Not Guilty: the psychology of crime investigations'.