Implications of within-person variability in cognitive and neuropsychological functioning for the interpretation of change

Abstract

Samples of adults across a wide age range performed a battery of 16 cognitive tests in 3 sessions within an interval of approximately 2 weeks. Estimates of within-person variability across the 3 assessments were relatively large and were equivalent in magnitude to the cross-sectional age differences expected over an interval of 15-25 years. These findings raise questions about the precision of assessments based on a single measurement and imply that it may be difficult to distinguish true change from short-term fluctuation. Because there were large individual differences in the magnitude of this variability, it is proposed that change might be most meaningfully expressed in units of each individual's own across-session variability.

Figure 1

Schematic illustration of the ambiguity of change when there is variability in level of performance at each measurement occasion (i.e., T1 and T2). The right panel portrays the proposal to assess change in terms of the t value between the distributions of scores from the two occasions.

Figure 2

Average levels of within-person variability and of 10-year cross-sectional age differences for four cognitive abilities scaled in between-person standard deviation units.