Using Tools to Help Us Think: Actual but Also Believed Reliability Modulates Cognitive Offloading.

Abstract

A distributed cognitive system is a system in which cognitive processes are distributed between brain-based internal and environment-based external resources. In the current experiment, we examined the influence of metacognitive processes on external resource use (i.e., cognitive offloading) in such systems. High-tech working environments oftentimes represent distributed cognitive systems. Because cognitive offloading can both support and harm performance, depending on the specific circumstances, it is essential to understand when and why people offload their cognition. We used an extension of the mental rotation paradigm. It allowed participants to rotate stimuli either internally as in the original paradigm or with a rotation knob that afforded rotating stimuli externally on a computer screen. Two parameters were manipulated: the knob's actual reliability (AR) and an instruction altering participants' beliefs about the knob's reliability (believed reliability; BR). We measured cognitive offloading proportion and perceived knob utility. Participants were able to quickly and dynamically adjust their cognitive offloading proportion and subjective utility assessments in response to AR, suggesting a high level of offloading proficiency. However, when BR instructions were presented that falsely described the knob's reliability to be lower than it actually was, participants reduced cognitive offloading substantially. The extent to which people offload their cognition is not based solely on utility maximization; it is additionally affected by possibly erroneous preexisting beliefs. To support users in efficiently operating in a distributed cognitive system, an external resource's utility should be made transparent, and preexisting beliefs should be adjusted prior to interaction.