The Metacognitive Role of Giving-up Tendencies

Abstract

The decision to give up reflects a metacognitive strategy which can minimise errors, time and effort when solving problems. Despite its utility, there is a dearth of research into this behaviour. This thesis presents the results of three studies involving a total of 546 university student participants that examined giving-up behaviour during problem-solving from individual differences and cognitive perspectives. This research had three aims: (1) to establish whether giving-up decisions were systematic across tasks, (2) to determine the factors that underly and/or relate to giving up, and (3) to determine the performance outcomes within the relevant cognitive tasks and beyond that relate to giving up. To achieve this, giving-up decisions were captured as an opt-out option embedded in cognitive tasks and were used to derive metrics such as time spent until giving up and giving-up frequency. Additional metrics were also captured, including other cognitive and metacognitive variables, decision-making styles and performance metrics. This research established three key findings. First, participants varied systematically in giving-up behaviour across different cognitive tests, as supported by factor analyses and latent profile analyses. Second, metacognitive, cognitive and decision-making variables related to giving-up behaviour at both item and individual level. At item level, giving-up decisions were made more frequently when feelings of confidence were low and were made faster when a problem was perceived as unsolvable. At individual level, giving-up tendencies correlated positively with cognitive ability and rational decision-making. Finally, this research established that different giving-up profiles corresponded with adaptive or maladaptive performance outcomes. This thesis contributes to the understanding of giving-up decisions during problem-solving and provides a strong foundation for its further investigation within meta-reasoning and decision-making frameworks.