Test-Retest Reliability of the Test of Infant Motor Performance.

Test-Retest Reliability of the Test of Infant Motor Performance

Campbell SK (Department of Physical Therapy, University of Illinois at Chicago, Chicago, III), Pediatr Phys Ther. 1999;11:60-66.

This study explored the reliability of the Test of Infant Motor Performance (TIMP), a functional motor scale designed for use with infants between the ages of 32 postconceptional weeks and 4 months corrected or chronological age. The TIMP evaluates both spontaneous movements and the movements elicited in response to placement in various positions, to visual stimuli, or to auditory stimuli. The primary purpose of this study was to determine the test-retest reliability of the Test of Infant Motor Performance (TIMP). Test-retest reliability was also compared when the 2 paired tests were administered by the same tester and when 2 testers were used.

The subjects in the study were 106 infants representing white, black, and Latino ethnic groups. Care was taken to ensure that subjects represented all age groups across the age range of the test and across the 3 levels of risk for developmental disability (high, medium, or low). Seven physical therapists and 3 occupational therapists participated in the study as testers. All 10 therapists had completed training in administration of the test and were rated as consistent examiners, as evaluated by the FACETS computer program for Rasch psychometric analysis.

Following receipt of parental consent, each infant was tested twice within 3 days. Ten infants were also tested on 2 different occasions, resulting in a total of 114 pairs of usable data (2 infants were excluded from the analysis because of incomplete test results). Fifty-nine percent of the tests were performed by the same tester on each of the 2 paired tests, and 41% of the paired test sessions were performed by different testers.

Multiple regression analysis was used to compare the raw scores on the tests. The Pearson product moment correlation coefficient between the paired tests was .89 with a difference in mean scores of 3.3 points. The maximum total raw score of the TIMP is 170 points. When data for same versus different raters were factored in, no significant difference between the paired test scores was noted.

The author concluded that the test-retest reliability on the TIM P over a 3-day period exceeded the recommended minimum reliability of .80 and that differences in test scores between 2 different testers were not statistically significant. She thought that the high test-retest reliability was related to the emphasis on functional infant motor behaviors rather than focusing on reflexes, postural tone, or manipulation of the infant's behavioral state. The TIMP was recommended by the author as a useful clinical tool for evaluating functional motor behavior of infants in the nursery who are not dependent on mechanical ventilation and who are considered to be healthy enough to withstand testing.

Donna Cech, PT, PCS Midwestern University Downers Grove, Ill