DIAGNOSING INDIVIDUAL DIFFERENCES IN BIAS DETECTION DURING REASONING – DIAGNOR

Although the human species has unique cognitive abilities to reason, our thinking is often biased by erroneous intuitions. From an educational and societal perspective such bias is a massive problem and cognitive scientists have been trying hard to identify its nature. Recent studies have suggested that despite their biased response, adults can detect that their answer is questionable and conflicts with logical-mathematical considerations. These conflict detection findings have been taken as support for the idea that heuristic bias typically results from a failure to override inappropriate intuitions, rather than from a failure to detect that these intuitions are unwarranted in the first place. However, a fundamental limitation of previous work is that it has paid little attention to possible individual differences. Research has focused on the modal or average biased reasoner. That is, analyses are typically run at the group level. These analyses allow us to draw conclusions about the typical nature of a biased response, as given by the “average” biased reasoner. However, there can be subgroups of individuals who are biased for different reasons. In other words, although group level analyses inform us about the reason for why most people are biased, they do not allow us to diagnose why a specific individual X or Y is being biased.

Identifying such individual variance is theoretically important but has even further reaching applied implications. Given the importance of sound reasoning for all aspects of life from the classroom to the office, it is not surprising that cognitive and educational scientists have been trying to develop educational “de-bias” interventions to help people avoid biased thinking. However, results of these interventions have been less than optimal. One reason lies in individual bias locus variance. If different individuals are biased for different reasons, they will obviously benefit from a different type of training. Hence, a straightforward solution to boost the efficiency of intervention programs is to target each type of program at those specific individuals that need them most. This requires the type of individual level analysis or diagnosis that current fundamental research fails to provide.

The DIAGNOR project will directly address the lack of individual level analysis in previous work. We propose an in depth and systematic exploration of the individual differences question with an ambitious combination of large scale behavioral, neuroscientific, and developmental studies. Our planned behavioral individual differences studies will use a simultaneous co-registration of different detection measures to sidestep the limited reliability of each individual measure. In addition, in our planned neuroscientific studies we will use (f)MRI and EEG to look for neural markers of individual differences in bias detection efficiency. Finally, all studies will be run with adult reasoners and younger populations to give us insight into developmental changes in the locus of individual bias differences.

Taken together, this project will allow us to diagnose the precise bias locus for each individual reasoner and help us to tackle one of the key outstanding issues in the reasoning field. This fundamental research will advance our insight into the nature of reasoning bias and allow us to optimize the efficiency of educational training programs to help reasoners avoid biased thinking.