Learning Effectively From Experience: Distinguishing High from Low Performers

High performers learn from both success and failure making small adjustments. Conversely, low performers learned more from success.

Learning effectively from experience is a daunting task for any organism. For every good or bad outcome, there are an immense number of potential causes and associations to be considered. For many decisions, it can be nearly impossible to pick out the few relevant factors from the many irrelevant factors, even with extensive experience. A major stumbling block for learning in these multi-dimensional environments is the tendency to form spurious beliefs: i.e., to attribute a causal role to factors that have no actual bearing on the outcome.

The formation of spurious beliefs is universal, from Skinner’s observations of superstitious pigeons [1] to an athlete’s belief in a lucky hat. In some situations, these beliefs are essentially harmless; by-products of learning mechanisms, but in other settings their impact can be severe. For example, spurious associations can have literal life-or-death consequences when they affect the complex decisions made by physicians. These expert decision-makers must extract and distill relevant features from a myriad of tests, symptoms, and personal histories, and employ these features to make critical medical decisions. Consequently, it is important to understand how spurious associations form and how they can bias subsequent decisions.

Spurious learning and false belief formation happens when the dorsolateral prefrontal cortex fails to distinguish correctly between important and unimportant associations.

The authors conclude:

High performers learned from both successes and failures, and made smaller rule adjustments after feedback. Conversely, low performers learned disproportionately from successes, and made larger rule adjustments …

Taken together, the behavioral and neuroimaging results suggest that success-chasing and confirmation bias may underlie the relative pervasiveness of premature, asymmetric learning and the resultant poor performance of the majority of physician subjects in the present study. The general human bias towards confirmation over disconfirmation in hypothesis-testing has been extensively documented in a variety of non-medical contexts, such as the Wason Card Task. Conversely, the necessity for disconfirmation learning in empirical investigations is a key principle identified by the philosopher of science, Karl Popper [28]. Conceivably, providing medical professionals with formal training in disconfirmation learning could improve their ability to learn effectively from clinical experience in real-world settings. Exploring this possibility would be an important area for future research.

In conclusion, the results of this study show distinct patterns of learning, both behaviorally and neurally, between effective and ineffective learners among physicians making decisions in a medically framed learning task. The tendency to chase successes and ignore failures provides a simple computational model of how spurious beliefs might be formed, and how different individuals seeing similar data might learn very different sets of associations. The neural differences observed could conceivably be developed into useful biomarkers for essential differences in individual learning styles. These may in turn prove useful in identifying those individuals who can resist the impulse to chase successes, and hence learn most effectively from experience. Finally, we note that although this study focused upon the specific case of medical decision-making, the findings may be also be relevant to many other fields in which experts must make high-stakes decisions by drawing upon personal experience.


Accurate associative learning is often hindered by confirmation bias and success-chasing, which together can conspire to produce or solidify false beliefs in the decision-maker. We performed functional magnetic resonance imaging in 35 experienced physicians, while they learned to choose between two treatments in a series of virtual patient encounters. We estimated a learning model for each subject based on their observed behavior and this model divided clearly into high performers and low performers. The high performers showed small, but equal learning rates for both successes (positive outcomes) and failures (no response to the drug). In contrast, low performers showed very large and asymmetric learning rates, learning significantly more from successes than failures; a tendency that led to sub-optimal treatment choices. Consistently with these behavioral findings, high performers showed larger, more sustained BOLD responses to failed vs. successful outcomes in the dorsolateral prefrontal cortex and inferior parietal lobule while low performers displayed the opposite response profile. Furthermore, participants’ learning asymmetry correlated with anticipatory activation in the nucleus accumbens at trial onset, well before outcome presentation. Subjects with anticipatory activation in the nucleus accumbens showed more success-chasing during learning. These results suggest that high performers’ brains achieve better outcomes by attending to informative failures during training, rather than chasing the reward value of successes. The differential brain activations between high and low performers could potentially be developed into biomarkers to identify efficient learners on novel decision tasks, in medical or other contexts.