Accuracy of Visual Estimates of Joint Angle and Angular Velocity Using Criterion Movements

Abstract

A descriptive study to document undergraduate physical education majors' (22.8 +/- 2.4 yr. old) estimates of sagittal plane elbow angle and angular velocity of elbow flexion visually was performed. 42 subjects rated videotape replays of 30 movements organized into three speeds of movement and two criterion elbow angles. Video images of the movements were analyzed with Peak Motus to measure actual values of elbow angles and peak angular velocity. Of the subjects 85.7% had speed ratings significantly correlated with true peak elbow angular velocity in all three angular velocity conditions. Few (16.7%) subjects' ratings of elbow angle correlated significantly with actual angles. Analysis of the subjects with good ratings showed the accuracy of visual ratings was significantly related to speed, with decreasing accuracy for slower speeds of movement. The use of criterion movements did not improve the small percentage of novice observers who could accurately estimate body angles during movement.

Structural Principal Components Analysis of the Kinematics of the Soccer Kick Using Different Types of Rating Scales

“…The influence of scale effects on rating accuracy was prominent (Figures 4 and 5, and Table 1), whereby the second hypothesis was accepted, and the predominance of a particular effect was specific to each variable. Faster Speeds of approach were estimated more accurately (EE = 0.44 ± 0.29 m/s) than slower speeds (EE = 0.61 ± 0.38 m/s), which concurs with current research [3] and contrasts with previous findings of dynamic observations [5][6][7]. The reason for this finding remains unknown, and may be related to the specialised function of the human eye as a motion detection system [8].…”

“…However, SDs were noticeably high, perhaps due to standard error (error due to a rater having ratings consistently different to other raters [1,2]). Thus, it can be suggested that a preliminary test is administered to ascertain the instructional requirements of individual raters; not all novice participants in Morrison et al's [3] study could be considered 'raters'. Findings for scale type and scale effects are shown below:…”

“…Mean and standard deviation (SD) absolute estimation error (EE), using the original metric units for each variable, was calculated. Normal Q-Q plots and Boxplots [22] were employed to remove extreme outliers from the data set (using a pre-selected cut-off value set by the experienced rater; e.g., 60°for the Angle of maximum knee flexion and Angle of maximum hip flexion variables, [3]).…”

“…These include relative, absolute, graphic, numerical, descriptive, and other [1,2]. In particular, analogue and Likert scale designs have enjoyed great popularity [3]. Up to the present time, however, there is lack of research regarding the type of scale that is more accurate.…”

“…Also, Knudson [7] proposed that dynamic observations are less valid and reliable than estimations of static postures. On the other hand, Freeman and Banks [8] have emphasised the capabilities of the human eye to perceive the speed of biological movement and Morrison et al [3] have suggested that, in fact, speed of movement may be "an important variable for qualitative analysis since it appears to be rated accurately by most novice observers" (p. 604). Thus, there is controversy in the literature regarding our ability to rate large versus more acute angles and slow movements compared to fast movements.…”