Some studies have reported that children with poor motor functioning tend to improve over time. However, much existing research does not account for regression towards the mean (RTM). Here, we examine measurement stability among 589 children aged 4–5 years.
We administered the Movement Assessment Battery for Children 2nd Edition annually to 269 children initially scoring above and 252 at or below the 16th percentile. We measured agreement between year 1 and year 2 standard scores using Pearson correlation and derived expected regression towards the mean (RTM). We then regressed follow-up on baseline scores, controlling for relative age, sex, and exact interval between assessments. Finally, we performed a small illustrative simulation.
Outcomes and results
The mean score in the poor-coordination group rose from 5.6 (SD = 1.5) to 7.2 (SD = 2.8). Year 1 and year 2 scores were correlated at r = 0.66, corresponding to predicted RTM in the MI group of 1.56, close to the observed change of 1.57. Degree of change was not associated with time between assessments.
Observed improvements in motor functioning were consistent with measurement error. The stability of motor functioning may be greater than it appears from past research, and reported functional improvements in some studies may be illusory.
What this paper adds?
-Impaired motor coordination is common and can meaningfully affect functioning.
-Our analysis suggests that coordination is generally stable.
-Apparent improvement in research may often be produced by measurement error.