Jumping to Conclusions About Delusions?
When someone forms a belief hastily, on scant evidence, and sticks to their conclusion, we accuse them of jumping to conclusions.
Cognitive and clinical psychologists who study delusions have often reported that people with delusions tend to jump to conclusions – even about things outside the realm of their delusions.
We can measure how people decide, and whether they jump to conclusions, using the urns or beads task.
Like a magician, we present the participant with two glass jars (or urns) of colored beads – one jar containing mostly red beads with some blue, the other jar containing mostly blue beads with some red.
We move the jars from the participant’s view, select a jar, remove the lid, and show them beads from the chosen jar, one at a time, without replacement.
Their task – based on the stream of beads – is to decide whether the
magician experimenter chose the mostly red jar or the mostly blue jar to draw from.
The canonical finding – the one you may have heard about in lectures on the psychology of delusions – is that people with delusions jump to conclusions about which jar the beads are coming from. They decide after one to three beads. Say the first bead is red, a patient with delusions might conclude immediately and confidently that the jar the experimenter chose was the jar with mostly red beads.
People without delusions take many more beads to make up their minds.
This jumping to conclusions bias has been the focus of much theorizing, and the target of cognitive therapies for delusions (with small to moderate effects).
If we can encourage people not to jump to conclusions, their delusions should get better.
Computational modeling analyses – where we write a mathematical account of participants’ belief formation and updating and combine that model with their actual choices to estimate the model parameters – suggest people jump to conclusions because their beliefs and choices are noisier, they switch beliefs excessively.
This is key.
Whilst people with delusions decide hastily, they don’t stick to their choice (which is what one might assume, based on the more colloquial meaning of jumping to conclusions, and the status of delusions as inordinately fixed beliefs).
Rather, if asked to keep watching beads, even after they have decided, they will flip-flop back and forth between the jars as more and more beads are presented.
The models suggested that was because they struggle to form a stable belief.
How and why that relates to fixed delusions remains to be seen.
Furthermore, recent data – including meta-analyses that look at all the data in aggregate – call into question whether patients with delusions have a jumping to conclusions bias.
We experimenters may have jumped to conclusions.
Meta-analyses have reported conflicting findings – depending on their study inclusion criteria – one meta-analysis reported a modest association between jumping to conclusions and delusions, another, no association with delusions in patients but a link between jumping to conclusions and delusion-like beliefs in people who do not have schizophrenia.
It is hard to conclude that this hastiness bias is a mechanism for the genesis of delusions in patients with schizophrenia.
That conclusion is further troubled by more recent results.
It seems that the jumping to conclusions effect may load heavily on participants’ general intellectual abilities – rather than their delusions.
Put simply, they may not understand the task and its associated probabilities.
Given a red bead, they don’t understand how likely how it is that bead came from the mostly red versus the mostly blue jar.
When we ensure that they do, and that they are appropriately incentivized, people with delusion-like beliefs and delusions-proper are actually more cautious in drawing their conclusions – they wait to decide until they have seen MORE beads than people without odd beliefs.
When we think about these behaviors and biases, we might benefit from an intuition from neuroimaging.
When we see a brain region apparently activated by a psychological process during some behavior, we should be careful to consider whether that activation is a cause, a consequence, or a correlate of that behavior. That is, does the activity underwrite the behavior? Is it induced as a result of the behavior? or is it driven by something with which the behavior is confounded?
Based on the evolving literature, it is yet unclear whether jumping to conclusions is a cause, a consequence, or a correlate of delusions.
What is certain – a simple jumping to conclusions bias does not explain the presence of delusions.