Do High Standards, Sensitivity to Failure, and Lack of Perseverance Interact to Predict All-or-Nothing Behaviour?

Abstract

Why do some people tend to begin new goals or activities with a high level of commitment and determination, only to drop substantially in their involvement within a short time? The purpose of the present research was to identify stable individual differences that influence the tendency to manifest this pattern of behaviour (i.e., starting strong then dropping off), which is subsequently termed “all-or-nothing behaviour”. In Study 1, participants completed questionnaire items that were expected to tap all-or-nothing behaviour. Common factor analysis was performed to examine the factor structure of the data, and revealed three independent constructs that were labelled “high standards”, “sensitivity to failure”, and “lack of perseverance”. The purpose of Study 2 was to assess the temporal stability of these constructs. Participants completed measures of the constructs at two times separated by 20 weeks. High correlations between scores at Time 1 and Time 2 revealed that the constructs are stable over time. The purpose of Study 3 was to examine the distinctiveness of the target constructs from existing individual difference measures, and to examine their association with theoretically similar and dissimilar constructs. Participants completed measures of the target constructs along with several other individual differences, and correlations among the measures were computed. Moderate to high correlations were found between the target constructs and theoretically related constructs, but they were not were not so high such that they appeared to represent the same underlying dimension. Further, relatively low correlations were found between the target constructs and theoretically unrelated constructs. The purpose of Studies 4 and 5 was to investigate the influence of the target constructs on all-or-nothing behaviour in the domains of group fitness attendance (Study 4) and campus club involvement (Study 5). In both Studies 4 and 5, participants completed measures of the constructs that were believed to relate to an all-or-nothing behavioural pattern, and their adherence to the target behaviours was assessed. I hypothesized a three-way interaction among the target constructs such that the presence of all three would be necessary to predict all-or-nothing behaviour. Results from both Studies 4 and 5 were consistent with this hypothesis such that individuals who were high in each of the three constructs were more likely to exhibit the all-or-nothing pattern of behaviour than were other individuals. Theoretical implications for understanding processes involved in behaviour change, and practical implications for interventions targeting at-risk individuals, are discussed.