Experimental and self-reported measures of impulsivity: A reconsideration

Impulsivity is a complex construct that most people seem to understand intuitively, but ordinarily struggle to define precisely. This term readily lends itself to everyday definitions, including acting without thinking, rash and quick decisions, and impatience. However, definitional issues regarding this term are widespread across scientific studies. Impulsiveness is often considered as a personality trait or a transient state within the psychology literature, whereby the former either refers to an underlying trait or the product thereof while the latter concerns itself with action or choice impulsivity. Additionally, economists typically only consider impulsivity that would be referred to by psychologists as choice impulsivity. We sought to better understand the relationship between self-reported impulsiveness as commonly measured by psychologists and incentivised impulsivity tasks as commonly used by experimental economists. Analysing data collected in 2012, we consider how a risk preference task and a time preference task are related to the Barratt Impulsiveness Scale-11 (BIS-11). We conduct joint maximum likelihood estimation to estimate discounting and risk preference parameters, and compare them to the BIS-11 total score and subscales. Our results show that the BIS-11 is related to time preferences, while only the motor impulsiveness and nonplanning subscales of the BIS-11 are related to risk preferences. These results suggest that the BIS-11 is not a direct substitute for experimental economic tasks when assessing impulsivity