Gambling habits and Probability Judgements in a Bayesian Task Environment.

Area of Science:

• Cognitive psychology
• Behavioral economics
• Decision science

Background:

• Understanding how individuals form probability assessments is crucial, particularly in contexts like gambling.
• Existing research offers limited insight into the cognitive processes underlying gamblers' probability estimations.
• This study investigates factors influencing subjective probability assessments in gamblers and non-gamblers.

Purpose of the Study:

• To examine whether experienced online gamblers exhibit superior probability estimation accuracy compared to non-gamblers.
• To determine if accuracy in probability assessment differs based on gambling type (skill vs. non-skill games) and sex.
• To explore the influence of gambling frequency and cognitive reflection on Bayesian accuracy and information weighting.

Main Methods:

• Administered an incentivized Bayesian choice task to 465 self-reported gamblers and non-gamblers.
• Utilized the task to elicit subjective probability assessments and quantify the weighting of different information sources.
• Employed decision modeling to analyze information integration and accuracy in probability estimations.

Main Results:

• Hypotheses regarding superior accuracy in experienced gamblers, skill-game players, or by sex were not supported.
• Higher gambling frequency correlated with lower Bayesian accuracy, while higher cognitive reflection predicted greater accuracy.
• The negative impact of gambling frequency on accuracy was more pronounced in female participants.

Conclusions:

• Online gambling frequency is linked to diminished probability assessment skills, potentially hindering success in skill-based games.
• Reduced weighting of new evidence, relative to base rates, was observed in groups with lower accuracy.
• Cognitive reflection emerges as a key factor for accurate probability assessment, underscoring the importance of evidence integration.