Objective: To investigate the reliability, concurrent validity, and error of a new video digitizing system for evaluating posture when applied to inanimate objects.
Design: Delayed repeated measures of digital images of inanimate objects.
Setting: University laboratory.
Methods: Digital video images of inanimate objects (5 parallelograms) of different sizes and shapes were obtained with the BioTonix postural evaluation system. Three examiners digitized video images of inanimate objects twice; the second data collection was 1 week after the first set. The objects were digitized with both high- and low-resolution settings of the video screen. The Tonix's measurements were statistically compared with the actual object dimensions. Statistical evaluations of reliability and validity were conducted.
Results: For distances, both intraclass and interclass correlation coefficients were very high, 0.99 for the estimate. The low- versus high-resolution settings were comparable for distances. For angles, on the low-resolution setting, both intraclass and interclass correlation coefficients were very high: 0.969 and 0.953. On the high-resolution setting, for angles, both intraclass and interclass coefficients were well above 0.99. The difference of the actual size and the means of the digitized measurements of the means were small: at most 1.5 degrees for angles and 3.3 mm for distances. The standard deviations were small, and the confidence intervals were narrow.
Conclusions: Our results demonstrate that the BioTonix's video system has high degrees of reliability and validity. Thus this system would seem suitable for clinical use in the analysis of posture.