The science of Human Factors and Ergonomics (HF/E) often relies on self-report. This is a cause for concern because subjective methods are inherently susceptible to bias. Here, we present, examine and discuss a puzzling association between age and self-reported cognitive failures as assessed with Broadbent's Cognitive Failures Questionnaire (CFQ). Despite many well-established age-associated forms of cognitive decline, older persons actually report almost equivalent, or even less, cognitive failures on the CFQ than younger persons. Our present analysis indicates that this paradox may be resolved through the fact that people show age-related learning/adaptation/compensation and by the observation that the CFQ measures peoples' beliefs with respect to an individually idiosyncratic reference. Yet, at the heart of the paradox may be the idea that people cannot remember their own cognitive failures, pointing to even greater concerns with all forms of subjective self-report and its use in HF/E. Practitioner Summary: Scientists and practitioners often try to understand and improve human performance with the help of self-report questionnaires. Our paper discusses the validity of self-reported errors measured with the Cognitive Failures Questionnaire (CFQ). We look to resolve the curious paradox that older persons tend to report fewer failures than younger persons do.
Keywords: cognitive impairment; ergonomics tools and methods; human error; memory.