Test-Retest Reliability of the Assessment of Motor and Process Skills in Elderly Adults

Abstract

We examined the test-retest reliability of the Assessment of Motor and Process Skills (AMPS) in a heterogeneous sample of 55 elderly adults. Subjects were reassessed within 1 to 10 days (M=4.0, SD=2.7 days). The subjects' Time 1 and Time 2 ability measures were highly correlated (Motor: r(55)=.88, P<.001; Process: r(55)=.86, P<.001). Measures were stable in 92% of the subjects. Significant variability occurred in only 8% (N=9) of the subjects' ability measures. Measurement error was a concern less than 2% of the time; 6% of the subjects' measures differed as a reflection of real changes in their task performances. Although some variability was attributed to fluctuations in subjects' medical conditions or motivation to adhere to the task requirements, not all of it could be attributed to a specific factor. These data provide support for the reliability of the AMPS, and serve as a reminder that not all test—retest differences represent measurement error.