The multidimensional nature of the Cognitive Estimation Task: Evidence from healthy aging and patient groups

Objective: The Cognitive Estimation Test (CET; Shallice & Evans, 1978; MacPherson et al., 2014) is widely used in clinical and research settings as an off-the-shelf measure of executive function. The CET is thought to assess the ability to produce reasonable estimates to items that individuals would not know that the exact answer (e.g., “How fast do race horses run?”) and is thought to rely upon processes such as reasoning, the development and application of appropriate strategies and response plausibility checking, as well as general knowledge, numeracy etc. Given that there are many different ways of performing or indeed failing the CET, it is important to understand how individuals produce cognitive estimates and whether specific estimates might be better at identifying impaired estimation abilities. Participants and methods: One-hundred and eighty healthy adults aged between 18 and 85 years performed an online version of the CET-A and CET-B, providing cognitive estimate responses, as well as the strategies used to generate these cognitive estimates. A second group of neurological patients with lesions due to brain tumours, stroke or neurodegenerative disease were assessed on the CET-A and/or CET-B. Item analyses of the CET items was conducted across the different patient groups in an attempt to improve the quality and accuracy of the CET.Results: Item-total correlational analyses revealed that, depending on the patient group, distinct items were better at discriminating between patients who performed well and patients who perform poorly on CET-A and CET-B. Further analyses of the strategies adopted by healthy participants’ to perform the CET items were examined in relation to their CET performance and the patients’ performance. Conclusions: These results suggest that CET items measure distinct constructs and different items are better at discriminating impaired performance in different patient groups. These findings provide further support for the multidimensional nature of the CET.