Measuring specific, rather than generalized, cognitive deficits and maximizing between-group effect size in studies of cognition and cognitive change.

摘要

While cognitive impairment in schizophrenia is easy to demonstrate, it has been much more difficult to measure a specific cognitive process unconfounded by the influence of other cognitive processes and noncognitive factors (eg, sedation, low motivation) that affect test scores. With the recent interest in the identification of neurophysiology-linked cognitive probes for clinical trials, the issue of isolating specific cognitive processes has taken on increased importance. Recent advances in research design and psychometric theory regarding cognition research in schizophrenia demonstrate the importance of (1) maximizing between-group differences via reduction of measurement error during both test development and subsequent research and (2) the development and use of process-specific tasks in which theory-driven performance indices are derived across multiple conditions. Use of these 2 strategies can significantly advance both our understanding of schizophrenia and measurement sensitivity for clinical trials. Novel data-analytic strategies for analyzing change across multiple conditions and/or multiple time points also allow for increased reliability and greater measurement sensitivity than traditional strategies. Following discussion of these issues, trade-offs inherent to attempts to address psychometric issues in schizophrenia research are reviewed. Finally, additional considerations for maximizing sensitivity and real-world significance in clinical trials are discussed.