Cognitive estimation and its assessment

Purpose. We evaluated the internal consistency and construct and criterion validity of a 10-item revision of the Cognitive Estimation Task (CET-R) developed by Shallice and Evans to assess problem-solving hypothesis generation. Method. The CET-R was administered to 216 healthy adults from the Aging, Brain Imaging, and Cognition study and 57 adult outpatients with schizophrenia. Results. Exploratory and confirmatory factor analysis (EFA and CFA) of the healthy sample revealed that seven of the 10 CET-R items constitute a more internally consistent scale (CET-R-7). Though EFA indicated that two CET-R-7 dimensions might be present (length and speed/time estimation, respectively), CFA confirmed that a single factor best represents the seven items. The CET-R-7 was modeled best by crystallized intelligence, adequately by fluid intelligence, and inadequately by visuospatial problem solving. Performance on the CET-R-7 correlated significantly with the neuropsychological domains of speed and fluency, but not memory or executive function. Finally, CET-R performance differed by diagnosis, sex, and education, but not age. Conclusions. This study identified an internally consistent set of items that measures the construct of cognitive estimation. This construct relates to several important dimensions of psychological functioning, including crystallized and fluid intelligence, generativity, and self-monitoring. It also is sensitive to cognitive dysfunction in adults with schizophrenia.