DEVELOPMENT OF TESTS T0 MEASURE 1 mm: AND GROSS Empmo‘cEPmN m cmmREN Thesis for the Degree of Ph. D. M‘CHJGAN‘ STATE UNWERSITY‘ PAUL‘D. Roamson ' 1969 If Hittite» This is to certify that the thesis entitled Developmenf of Test to Measure Fine and Gross Proprioception In Children presented by Paul David Robinson has been accepted towards fulfillment of the requirements for A degree in Education and Recreafion Dam January 2!, l9 0-169 Health, Physical LIBRARY Michigan State University .t, ‘5 J \ “Pr-r «a: .e DEC 0 5.1096 W ABSTRACT DEVELOPMENT OF TESTS TO MEASURE FINE AND GROSS PROPRIOCEPTION IN CHILDREN By Paul D. Robinson The purpose of this study was to develop tests which would measure fine and gross proprioception in children. A series of seven tests Was developed to measure muscle tension variation, joint angle perception, labyrinthine function and the integration and coordination of these factors of proprioception. The tests developed were Weight Discrimination, Single Foot Balance, Parallel Blocks, Dis- tance Perception Walk, Weight Shifting, Thickness Discri- mination and Dot Repetition. Twenty one fifth and sixth grade boys were given the test series on three occasions at intervals of one month. The subjects were tested in groups of two to four boys and each subject was randomly assigned to a test within the series. All subjects were blindfolded for each test and at no time during the tests were they allowed I knowledge of their performance. : x Paul D. Robinson The reliability of each test was determined by the Pearson product moment coefficient of correlation from test-retest results. The presence of a learning curve within each test was illustrated by graphing the mean scores of each trial. Between month variation was deter- mined by plotting the test curves on the same graph. The results showed the most reliable tests to be Weight Discrimination (.85), Thickness Discrimination (.76), Single Foot Balance-—Dominant Foot (.99) and the Parallel Blocks test (.87). A learning curve was present in all of the tests but two and there was little between month variation. The author recommends the above four tests to be used in measuring the proprioception of children. v—_.__._-—————*_..-- DEVELOPMENT OF TESTS TO MEASURE FINE AND GROSS PROPRIOCEPTION IN CHILDREN By Paul D5 Robinson -.2# A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Health, Physical Education and Recreation 1969 Copyright by PAUL D. ROBINSON 1969 —..v-— A-.. 56550 M £7 To Jessica, Jason and Tracy ACKNOWLEDGMENTS The author wishes to thank Dr. Wayne Van Huss and Dr. Arthur H. Steinhaus for their advice and guidance; also David Anderson, Roger Allen, Archibald Young, Jr., Daniel Morabito and Judith Dormeier for their assistance and encouragement throughout this study. iv TABLE OF CONTENTS ACKNOWLEDGMENTS. LIST OF TABLES LIST OF FIGURES. LIST OF PLATES LIST OF APPENDICES. Chapter I. INTRODUCTION Statement of Problem. Need for the Study . Limitations of the Study Definition of Terms II. SELECTED REVIEW OF LITERATURE Brief Review of the Proprioceptive System. Review of Literature. III. METHOD Introduction Experimental Design . Tests and Test Administration. Analysis of Data . IV. RESULTS AND PRESENTATION OF DATA Introduction . . . . . Treatment and Presentation of Data . Results . . . . . . V. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS. Summary Conclusions. Recommendations BIBLIOGRAPHY. APPENDICES Page . viii ix LIST OF TABLES Table Page I. Weight Discrimination . . . . . . . . A5 II. Weight Discrimination——First Five Trials. . A8 III. Dot Repetition . . . . . . . . . . A9 IV. Dot Repetition--First Ten Trials . . . . 50 V. Distance Perception Walk . . . . . . . 52 VI. Distance Perception Walk-~First Five Trials. 5A VII. Parallel Blocks . . . . . . . . . . . 55 VIII. Parallel Blocks——First Ten Trials . . . . 56 IX. Thickness Discrimination . . . . . . . 57 X. Weight Shifting . . . . . . . . . . 60 XI. Single Foot Balance. . . . . . . . . 6“ vi LIST OF FIGURES Thickness Discrimination Apparatus. Parallel Blocks Apparatus. . . . Weight Discrimination . . . . Weight Discrimination (Bar Graph) . Dot Repetition . . . Distance Perception Walk . . . Parallel Blocks . . . . . . . Thickness Discrimination . . . . Thickness Discrimination (Bar Graph) Weight Shifting . . . . . . . vii Page 27 32 A6 47 51 53 53 58 61 63. .Htrim .on “.15; ._‘ .t ,‘ :JJ‘I. Plate J: \OODK‘IONU'I LIST OF PLATES Weight Discrimination Test Thickness Discrimination Test Dot Repetition Test Dot Repetition Test Parallel Blocks Test Parallel Blocks Test 0 0 Single Foot Balance Test. Weight Shifting Test Weight Shifting Test viii Page 25 28 30 30 3A 3A 36 38 38 A A“ - Appendix A. Q'IJLTJUOW LIST OF APPENDICES Weight Discrimination Test Single Foot Balance Test Parallel Blocks Test Distance Perception Walk Test Thickness Discrimination Test Weight Shifting Test Dot Repetition Test. ix Page 7A 75 77 8O 81 82 83 !,,‘-,.’-- ‘ CHAPTER I INTRODUCTION All knowledge of our external and internal environ- ment comes to us through our sense organs. The sense organs initiate directly or indirectly even our simplest motor acts, and our most complex behaviour patterns are controlled by means of them. Ruch (7) suggests that pure sensation is an abstraction, occuring only the first time a baby experiences that sensation. "Thereafter past exper- iences, the blending of sensations, the comparison of one sensation with another, etc., transform sensation into perception" (7, p. 303). Perception is the process of organizing and giving meaning to experience. Boyd (12) points out that even without the use of the eyes, the appreciation of the relative positions of different parts of the body is exceedingly precise in Inan. The spatial relationships between our various body Ibarts and the orientation of the whole body in space depends upon afferent inputs from both somatic sensory arni vestibular receptors as well as from the eyes (6). ITHAS information gained from sensory and organs which ,tParlsmit inpulses from the muscles, joints and tendons is Lusually integrated with information from other sense IA organs (22, 23). It has been reported by Steinhaus (9) that in a motor nerve to a muscle approximately forty per cent of the nerve fibres are actually sensory fibres carrying impulses to the spinal cord and brain from mus- cles, tendons and related joint structures. These afferent impulses make us aware, at the conscious level, of move— ment, whether active or passive, movement variation in either speed, duration or direction, stretch and tension, weight or resistance variation, and shape variation or stereognosis. The sensory system found in and around the muscles, tendons and joints is the "proprioceptive" system. More than any other sensory system the proprio- ceptive system is responsible for the coordinated func- tioning of man throughout his lifetime. Various studies have concluded that there is a relationship between proprioception or kinesthesis and motor learning, however, the precise relationship is still unknown (10, 33, 37, 39, A9). Proprioception is most per- timent to the acquisition of motor skills since the aware— ness of movement is essential to successful performance whether it be a fine or gross motor skill (9). It has been suggested that proprioception is more important in the early stages of motor skill development than in the later stages (38). Proprioceptive awareness is essential for motor performance, and increased performance which results from training and practice may be due in part to the heightened proprioceptive sensitivity (A0). Through our development of proprioception we are able to estimate the power, distance and span of movements by muscle ad- justments and slight movements in coordination. Proprio- ceptive sensitivity in the different joints will vary with the type of sports in which an individual may regularly participate (27). Thus the more highly developed is one's overall athletic ability, the broader and deeper one's proprioceptive sensitivity will be. The "feel" of a movement produced by proprioceptive 'feedback and the way in which it is organized visually and mentally are interdependent (3, A, 22, 23). Proprio- ception and memory are the basis of voluntary movement and of learning. With these the performer is able to initiate a whole or to modify a part of a whole action. The dancer or the gymnast may have a visual picture of the desired position or movements and must have a proprioceptive memory of that position or movement (3). "Human move- ment differs from animal movement because man is able to think about his own movement. He can conceptualize his kinesthetic perception of his own movements" (18). It has been held that proprioceptive sensations, particularly from the tongue, lips and other voice organs together with implicit muscular responses giving rise to them, are the fundamental basis of the higher thought processes (2, AA). A In the actual learning of various movement patterns the relationship between proprioception and physical edu— cation is found. In learning new movement patterns it is necessary to have an awareness of body position or move- ment. An individual is able to judge the accuracy of his movements from the memory of previous sensations and the conscious awareness or subconscious recording of present ones. Proprioception is therefore the sense of the muscles, joints and associated structures together with the laby- rinth, which is vital in the learning and the performance of motor activities. The functions to which proprioception contributes and is associated with include coordination of body movements, skill development, posture, loco- motion, balance, manipulation and the perception of weight, speed, direction and shape, all of which are important elements within the teaching of physical education. Statement of Problem } No single test has been found to measure proprio- ceptive ability though several combinations of tests , have given a moderately high relationship (Al, A8, A9). l The problem is that proprioception is composed of several specific factors each of which can vary independently from the others. These factors are the perception of (a) changes in muscle tension, (b) changes in joint L - A angle, (c) joint movement and (d) changes in vesti- bular stability. The purpose of this study was to develop tests which would measure fine and gross proprioception in children. The tests were designed, as far as possible, to measure the specific factors of which proprioception is composed. Need for the Study There is a lack of basic specific information about the variability of proprioception in children or adults. Though proprioception is thought to improve with training or practice there is little evidence to support this View. The development of a few reliable, economical, feasible and efficient measuring devices is necessary so that the teacher at any school grade level can determine proprioceptive weaknesses in children and prescribe possible suitable practices for their improvement. Limitations of the Study 1. The environmental conditions under which the subjects were tested were not consistent for all sub- jects. For example,temperature and humidity varied con— siderably within each series of tests for it was not Possitale to test all the subjects at the same time. 2. It was impossible to control the diet and feeding habits of the subjects from one series of test to the next. Some boys were tested not having eaten for four or five hours. 3. The amount of physical activity in which the subjects participated prior to testing was beyond the author's control. A. The possibility of some subjects practicing some of the tests between the one series and the next must also be accepted as being beyond the author's control. Definition of Terms Proprioception.--(Latin:--proprius meaning one's own; perceptio meaning perception) (l). Appreciation of position, balance and changes in equilibrium on the part of the muscular system. Appreciation of one's own body position in space. Proprioceptive Impulses.--Afferent nerve impulses originating in receptors in muscles, tendons, joints and vestibular apparatus of the internal ear. Their reflex functions are concerned with locomotion and maintenance of posture. Proprioceptor.--A receptor located in a muscle, teridon, joint or vestibular apparatus whose reflex furlction is locomotor or postural. — l l, I CHAPTER II SELECTED REVIEW OF LITERATURE Brief Review of the Proprioceptive System The morphology and physiology of the propriocep- tive system is only dealt with to the extent which is necessary for the author's purpose. Sherrington's term "proprioception" is not fully understood by many readers and for these the term "kinesthesia" or "kinesthetic perception" as used in experimental psychology may be substituted. The term "proprioceptive" is almost uni- versally used in physiological texts. Through proprioceptive sensations movement and position of the joints are perceived. Goldscheider in 1898 established that appreciation of movement of the limbs is derived mainly from stimulation of the joints rather than the muscles (6). In harmony with these findings there is, according to Rose and Mountcastle (6% no evidence for and strong evidence against the notion that impulses provided by the stretch receptors in the muscles provide information for the perception of movement or the position of joints. The propriocep- tors found in muscles, tendons and joints and in the ' - _'.f§’1-!'\' 3' T—————___———— . vw—V‘. - ‘vw‘ _ V'— v—vwv‘F-FV 8 labyrinth provide information concerning the movements and position of the body in space. The muscle spindles are found within the body of the muscle. They are excited by stretch of the muscle but cease to discharge as the muscle is shortened by alpha motoneuron action. However the gamma motoneurons produce upon discharge an increase in spindle organ activity even when the muscle shortens. Thus the spindle activity may vary from zero to maximum independently of the length or tension of the muscle. It is for this reason that these receptors cannot indicate muscle length or joint angle (6). 1 The Golgi tendon organs are found in the tendons close to the origin of their muscle. They discharge afferent impulses at a rate related to tension. The length of the muscle determines the tension to which these organs are subjected. The length depends upon the joint angle and upon the force exerted against a . resistance by the active contraction of the muscle. It should be noted that the number of Golgi organs and their rates of discharge are not completely dependent upon the angle of the joint, thus they cannot measure joint angle or muscle length alone (6). Howeven they do signal the tension in the muscle which is important in the discrimination of weight or resistance. — From histological work there are indications of three types of receptor organs in articular tissue. The most common are the Ruffini endings identified by Boyd (ll) in the joint capsule but not in the synovial lining membrane. They are slowly-adapting sense—organs capable of providing accurate information about the relative position of the bones forming the joint (12). They are able to signal the steady position and the direction, rate and extent of joint movement. During a movement, the rate of discharge is a function of its speed and extent. Though this work was performed on the knee joint of the cat, articular innervation has been found to be uniform in all species studies. Another slowly- adapting receptor similar to the Golgi tendon organ has been found associated with the ligaments of joints, but it is less numerous than the Ruffini type endings (6). Boyd and Roberts (12) revealed a fast-adapting sense organ also present in the joint structures. Upon histological examination Boyd (ll) determined them to be modified Pacinian corpuscles. The Pacinian corpuscles are widely distributed in the fascia of muscles and especially beneath the tendinous insertion of muscles at the joint. Pressure is exerted on them when the muscles are stretched or contracted. "The thesis that the sense of position and movements of the joints is dependent upon joint receptors themselves fits well the clinical observations (6, p. A15)." V< 10 "Upon entering the spinal cord, sensory fibres are regrouped so that the fibres for cutaneous and deep sensibility are no longer separate (8, p. 322)." The sensory impulses of proprioception are conducted in the posterior columns. The sensations served by the posterior columns give knowledge of limb position in space, fine discrimination of weight, size, shape and texture of objects handled (8). Proprioceptive activity is relayed in the medial lemniscal system (6). According to Ruch (8),it is apparent that discrimination of fine spatial functions requires a vast number of neurons arranged in dense, specially organized fields. The thalamus does not possess such fields but the cortex does. For this reason regions with a high degree of special dis- criminative ability, such as the tongue, have a large proportion of cortex devoted to them. From his exper— ience and reading the author would postulate that the impulses reaching the cortex give rise to sensations which are stored in the cortex as memories. The mem- ories are related to incoming sensations which allows the individual to form a perception. When sufficient memory has developed in relationship to motor behavior, it is possible for the individual to mentally perform an act. In doing this tensions are set up in the muscles required to perform the act and thus the performer "feels" the act. 11 The sense organs of the labyrinth contribute stim— uli to postural tonus through the impulses produced by movement of the head. Only slight stimuli are necessary to evoke reflexes which keep the body erect. The major stimuli of postural tonus arise from the labyrinth, tonic neck-righting reflexes, optic reflexes, kinesthetic sense, extensor reflexes, spinal stretch reflex and plantar reflexes (3). A person whose labyrinth function is lost must rely upon muscle and tendon proprioceptors together with visual stimuli to maintain his balance. The sway of the body produces the stimuli to evoke the righting reflex. This is especially evident in normal individuals standing erect with their eyes closed and demonstrates the importance of proprioception in bal- ance and maintenance of posture. Review of Literature Some of the first work done on the development of tests of proprioception was in 1909 by J. H. Leuba (30) who devised a special apparatus to study the perception of motion in joints. After experimenting with his apparatus on the influence of the duration and rate of arm movements upon the judgement of the length of the movement he concluded that "An apprehension of duration and rate is sufficient," to judge the length of a move— ment and that the existence of local signs in the joint _ . 12 sensations was questionable (31). Since that time many proprioceptive testing instruments have been devised but a determination of the reliability of each instrument, in most instances, has been lacking. The instruments which have been developed and modified from one author to the next are concerned primarily with measuring the perception of joint angle, balance, size, weight, mus- cular force and shape. A few tests have involved the perception of movement. By far the most common test has been the measure- ment of joint angle perception in both the upper and lower limbs. Young (A8) used arm raising to the side 900 and a leg raise of 20°. Stevens (50) also measured the ability of subjects to assume specific arm and leg positions by means of the kinesthetic sense. In a fac- torial analysis of measures of kinesthesis Witte (51) identified various positional factors related to arms and legs. These were:(l) arm positioning for short arm movements on the vertical plane; (2) arm position- ing in long arm movements of the vertical plane; (3) arm positioning on the horizontal plane; and (A) leg positioning. These four factors together with three concerned with the force of muscular contraction of arm and leg were identified as being basic to the thirty three tests of kinesthesis that he studied. Witte therefore suggests that kinesthesis cannot be d . (I) 13 thought of as a general trait. Wiebe (A5) did a careful statistical evaluation of existing tests of kinesthesis. He found that fifteen of the tests had reliability coef- ficients greater than .65 and recommended them as useful testing instruments. Unfortunatelg the reliabilities were computed from alternate halves of the results which only measures the consistency of test performance on that single occasion. Arm-static function and leg— static function were found to have an acceptable degree of validity in measuring kinesthesis. However, low intercorrelations between tests indicated that there was no general kinesthetic sensitivity but that there might probably be numerous specific factors. Clapper (A9), Phillips (38), Roloff (39) and Scott (Al) all used arm positioning in measuring kinesthetic perception. Phillips (38) pointed out that the superiority of the dominant or non-dominant arm in any given movement may be influenced by whether or not the individual habi- tually uses the movement or one similar to it. The accuracy with which the movement can be reproduced through proprioception is also affected by whether or not the movement is used in daily activities. Lloyd and Caldwell (32) studied the accuracy of active and passive positioning of the leg on the basis of kinesthetic cues. They found that the mode of move— ment and the goal positions significantly influenced ‘ " 7”. ”LT..- -_ i_-, Pfi— . I r l 1A the accuracy of positioning. Since the range of greatest accuracy for active movement coincided with the normal walking arc of the lower limb the authors suggest that accuracy of positioning is best in the range of move— ments in which there has been greatest practice. It is also hypothesized that the distribution and rate of receptor firing may favour discrimination in this range. Caldwell and Herbert (15L studying the judgment of angular positions in the horizontal plane on the basis of kinesthetic cues and Caldwell (IA) studying the accuracy of constant angular displacement of the arm in the horizontal plane also found that the accuracy of arm positioning was dependent upon the locus of the goal position. In addition,it is dependent upon the direction of the primary adjustive movements. They found that positioning accuracy tended to increase as the extremes of arm flexion and extension were approached, and to decrease as the arm was moved to a more medial position. They thus suggested that movement against an opposing force produced by the muscle antagonists themselves is associated with greater positioning accu- racy than movement with an aiding force produced by the relaxation of antagonist muscles. Wyke (A7) did a comparative analysis of proprio- ception in left and right arms and the effect of head rotation to left and right. She used the "kinesthetic M‘.‘"‘ 's- : :. «vmw «i- ‘vs “m":rr' ~-~~w=s ~ i I 15 memory for the target," technique. With the head nor— mally oriented the right arm (dominant), is better than the left in accuracy of target pointing. Accuracy of pointing was found to be greater with the target directly in front of the body than when it lay to either the left or right side. When the head is rotated the direction of pointing error was inversely related to the direction of rotation.~ It is suggested by Wyke that the precision of control over the arm is related to the varying ability of individual subjects to correlate limb movements with the prevailing orientation of the body, especially of the neck and head. A part of the clinical examination of deep sensibility is to test the "sense of position." The patient with his eyes closed has a limb placed in an unusual position and is then asked to duplicate the pos- ture with the other limb. Measurement of error is purely subjective. Another maneuver is the "finger to finger" test, error of which can be measured by interposing a piece of board between the two fingers and marking their j positions (7). The perception of movement whether it be rate, duration or direction has been tested by several authors. Clapper (A9) and Roloff (39) used arm swinging and arm circling to measure the kinesthetic perception of move— ment. Scott (Al) attempted to measure the perception of a more controlled movement, horizontal line drawing 16 but was only able to get a correlation of .52. Obukhova and Roman (36) studied the role of vision and kinesthesis in reproduction of passive movements. The movements are reproduced (a) without visual control, (b) with visual control of the movement and (c) with the eyes open but without visual control. The authors found that subjects were more accurate in reproducing passive movements when blindfolded than when allowed to look at the arm. Grif- fith (2A) reported a study on the importance of kines- thesis in learning the skill of driving a golf ball. He concluded, as in the above study, that the group which had been blindfolded during the first four weeks of prac— tice, and thus encouraged to rely on proprioception to "feel" the swing, surpassed the group which had prac- ticed with visual cues. Brown 2141;. (13) in a fine study on the accuracy of positioning reaction as a func- tion of direction and extent, had subjects move their right hand from a point of rest to a terminal test posi— tion in two directions in each of three planes. They found that subjects tend to overshoot short distances and undershoot longer distances. The only exception to this was in the case of downward movements in which there was a constant overshooting which the authors sug— gest may be due to gravity. Movement away from the body had smaller percentage of error than movement toward the body. Perhaps it may be questioned whether the sub- jects in the above studies were perceiving movement from — 17 duration and rate, as suggested by Leuba (31) or whether they are solely conscious of the pre and post movement joint angles. It is most probable that both processes are functioning and both contribute to the making of accurate judgments. The second most tested trait of proprioception has been the perception of muscle tension variation. Two kinds of test have been used, weight discrimination and force application, using either the upper or lower limbs. Stevens (50) developed a six item battery of tests which included two arm force tests and one leg force test. Using this battery she found that individuals trained in motor movements showed more highly developed proprio- ception than the untrained. Wiebe (A5) evaluated existing tests and substantiated that some, but not all, proprio- ceptive tests appear to be related to motor ability, one of which was the dynamic function of the arm as shown by the arm force test. Using a factor analysis of twenty five tests, Witte (51) identified seven as being relevant measures of kinesthesis, one of which was force of mus- cular contraction of arms and legs. Mumby (35) working with a similar constant pressure testing device as used by Henry (26) found that the results appeared to be significantly related to wrestling ability. Kerr and Weinland (29) also have shown that athletes were superior in muscular perception based partially upon the results from an arm force test. 18 The appreciation of muscle tension has been studied clinically by determining the ability to detect differ— ences in weights of objects by lifting them (7). The Weber-Fechner law (5) states that the smallest discrim— inable difference between two weights is a constant frac- tion of the weights themselves. This is known as the Weber fraction which, for weights, is approximately one thirtieth. This means that it is possible to discrimi- nate between 31 grams and 30 grams. Sekuler (A3) has pointed out that ability to discriminate weight increases over the range 100-AOO m.sec., and reaches a plateau between AOO—8OO m.sec. It is not known if longer dura- tion produces an adaptation to the weight. Sekuler sug— gests that secondary adjustments of small oscillatory movements may represent the output of a feedback system designed to optimize the ability to discriminate lifted weights. Hawkins (25) using the method of constant stimuli with two decision categories of "heavier" and "lighter" determined that in weight discrimination re- tardeds were inferior to normal subjects. Weight dis— crimination as a test of proprioception has been seldom used outside medicine and psychology yet it shows poten- tial in the accuracy to which it can be measured. Testing the perception of size or width is a relatively recent innovation to the area of proprio- ceptive measuring techniques. Comalli, et al. (17) _ 19 studied whether changes in muscular involvement affect tactual—kinesthetic perception of size. They determined that with muscular involvement the apparent distance between rods which were held by thumb and forefinger was significantly smaller than with no muscular involvement. Evans and Howarth (21) in a similar study determined that increased grip tension was found to reduce the accuracy of width judgment in terms of constant error but only slightly affected the variance. McPherson and Renfrew (3A) have shown that objects of equal width or size held simultaneously in each hand tend to be judged to be unequal and that objects in the dominant hand are perceived to be smaller by the majority of subjects. Churchill (16) is not in agreement with this latter study but argues on the basis of testing the hands sep- arately to show no difference, whereas McPherson and Renfrew (3A) tested the hands simultaneously. A few tests of balance have been used as measures i of proprioception. Bass (10) and McClOy (33) have ! determined that as well as semicircular canal function, kinesthesis was an important factor in both static and dynamic balance. Conversely,Wiebe (A5) found the testing of balance to be an important factor in studying proprioception. Espenchade (19) using the beam walking test devised by Seashore (A2), showed a positive rela- tionship between dynamic balance and physical abilities 2O important in physical education. Estep (20) and Katsuta (28) found the same with regard to static balance and gross motor performance. Thus it may be concluded that since proprioception is an integral part of balance then it is also an important factor in motor ability or performance. The vast majority of studies have found a posi- ltive relationship between motor ability and performance on the tests of proprioception (19, 20, 27, 28, 29, 35, 39, A0, A5, A8, 50). As to the effect of proprioception upon motor learning itself, B. E. Phillips (37) found a low positive correlation between some proprioceptive tests and the early stages of learning motor skills. This was later confirmed by M. Phillips and Summers (38). However, Clapper (A9), though she found a sig— nificant relationship between a battery of proprio— ceptive tests and rate of motor learning, the relation- ship, when tests were considered individually was insignificant. CHAPTER III METHOD Introduction The purpose of this study was to develop tests which measure both fine and gross proprioception in children. After reviewing the literature seven such tests were developed and the reliability of each was then determined. Experimental Design Twenty one boys from the fifth and sixth grades of Whitehills and Pinecrest schools in East Lansing, Michigan comprised the subjects for this study. The age range of the group at the beginning of the study was from 10 years 2 months to 12 years A months. The seven tests were placed in a specific order to restrict boredom and limit the development of fatigue or impair- ment in the subjects. This series of tests was administered on three occasions at intervals of approximately one calendar month beginning in mid-June 1968. The subjects were tested in small groups ranging from two to a maximum 21 22 of four boys. Each subject was randomly assigned to one of the tests in the series such that a slower test- ing subject would cause as little delay as possible to a faster testing subject. Each test was repeated suffi- cient times to determine the possible presence of a learning curve. The repetition of the series of tests at monthly intervals was to determine (a) the relia- bility of test repetition and (b) the possible training or maturational effect through time. Tests and Test Administration The test series was administered in the following order: Weight Discrimination, Single Foot Balance, Parallel Blocks, Distance Perception Walk, Thickness Discrimination, Weight Shifting, and Dot Repetition. Each subject was blindfolded for every test or where it was required within the test procedure (35). At no time during the tests were the subjects allowed to know the results of their performance. Weight Discrimination This test was first used by Weber (5) to deter- mine the smallest discriminable difference between two weights. He discovered this just noticeable difference to be a constant fraction of the weights themselves, the so—called Weber Fraction. For weights, this fraction is approximately one thirtieth. The test is used 23 clinically to measure a patient's ability to discrimi- nate but rarely has it been used to compare different populations. Since impulses from the Golgi tendon organs and the muscle spindles make it possible to discriminate variations in weight or resistance, and since they are both receptors of the proprioceptive system, the author used the Weight Discrimination.test to measure the abi— lity of children to discriminate between fine variations in muscle tension. Test Description and Procedure.—-Each subject was required to compare six different weights with a standard weight. The weights ,ranged from 60 grams to 90 grams at intervals of 5 grams, the standard weight being 75 grams. The weights consisted of small glass bottles 8 c.m. high by 3 c.m. in diameter containing lead shot and packed with cotton. Each of the six comparison weights of 60, 65, 70, 80, 85 and 90 grams were pre- sented to the subject alternately with the standard weight of 75 grams. The group of comparison weights were compared with the standard weight in ten trials. Within each trial the order of the comparison weights were randomly assigned. Thus the total number of com- parisons the subject had to make was sixty. The blind- folded subject was seated so that he could lift the weights with his elbow resting on the table and a .-- 2n rotating board 33 c.m.‘in diameter was used for the con— venience of presenting the weights, as shown in Plate 1. Between each comparison the subject was asked to rest his forearm on the table in order to reduce the effect of fatigue. Each subject was asked to use his dominant hand to lift the weights. With each comparison weight the subject was asked to say whether it was "heavier" than, "lighter" than,or "the same" as, the standard weight and the subject's reply was recorded. This test was scored by summing the incorrect answers, each of which was classed as being one error. These errors were also weighted in terms of their distance from the standard. For example, an error of 5 grams from the standard was weighted as 1 error, an error of 10 grams from the standard was weighted as 2 errors and an error of 15 grams from the standard was weighted as 3 errors. Thickness Discrimination Various tests designed to measure the perception of size or width have been used in the past (16, 17, 3A). However, the authors of these tests used them to measure variation within individuals and not between them. Since the position of a limb is perceived by inter- preting impulses received from the joint angles, the measurement of joint angle perception may therefore be determined by measuring the perception of joint angle variation. The Thickness Discrimination test is a means ~'. x-}'~-";‘W “~17“.- ’ ‘ T’s-fight??? .’ ' 1‘ ". .; I ~“ 25 Plate 1 Weight Discrimination Test ‘ . 49A”? -2: A “x'mfl.i _ , .. Q» hv F». Re a: Q.» fi‘r‘: J._‘ 26 of measuring the perception of joint angle variation. The joint of the fingers and thumb are involved in the discrimination of width. Therefore, since much of the sensory cortex is devoted to these appendages the dis- crimination of width is a measure of fine joint angle perception. In the test developed by the author the subject was required to use only his dominant hand for it has been shown by McPherson and Renfrew (3A) that objects of equal size held simultaneously in each hand tend to be judged to be unequal. Test Description and Procedure.--The apparatus as shown in Figure l was placed in front of the subject so that his dominant hand could be used in the discrimi- nation. A standard block 18 c.m. thick was permanently placed in the right hand groove of the apparatus. The group of comparison blocks 15, l6, l7, 19, 20 and 21 m.m. thick were compared with the block of standard‘ thickness in ten trials. Within each trial the order of the comparison thicknesses were randomly assigned. Thus the total number of comparisons each subject had to make was sixty. The subject, blindfolded, was seated in front of the apparatus as shown in Plate 2. He was required to feel the thickness between his fingers and thumb with the hand either in the pronate or supinate position being allowed one attempt at each 28 Plate 2 Thickness Discrimination Test \st 29 comparison within each trial. He was then asked to say whether the comparison thickness was "thicker" than, "thinner" than or the "same" as the standard thickness. Each reply was recorded and the test was scored by sum- ming the incorrect answers, each of which was classed as being one error. These errors were also weighted in terms of their distance from the standard. For example, an error of 1 mm. from the standard was weighted as 1 error, an error of 2 mm. from the standard was weighted as 2 errors and an error of 3 mm. from the standard was weighted as 3 errors. Dot Repetition The measurement of a more gross joint angle per- ception than that of the joints of the fingers and thumb is necessary since gross movements are as much a func- tion of daily life as fine movements. This test was developed to measure gross joint angle perception of the shoulders and elbow joints. Test Description and Procedure.-—The blindfolded subject was seated at a table holding a sharp pencil, close to the point, in his dominant hand. He was re- quired to make a dot on a piece of paper in front of him without his arm resting on the table as shown in Plate 3. He was asked to remember his arm position, straighten his arm laterally, as shown in Plate A and then attempt to duplicate the arm position making 30 Plate 3 Dot Repetition Test Plate A Dot Repetition Test -'" m“ ‘u (I) Em”.- 31 another dot in the same position as the first. This pro- cess was repeated for twenty trials. Each pair of dots were circled and numbered in the order of performance, one through twenty trials. The distance between the first and second dot of each trial was measured in mil— limeters and recorded. Parallel Blocks It is constantly necessary to integrate movements and limb positions from both sides of the body. Measure- ment of this integration has in the past been in a very gross form, usually through the performance of some spe— cific sports skill where many other variables come into play. It has long been necessary to isolate and measure the most important factor which is joint angle percep— tion. This test was therefore developed by the author to measure the bilateral integration of joint angle perception. In this test the importance of the subject main— taining the correct head and neck position must be emphasized since Wyke (A7) determined that when the head is rotated, the direction of pointing error is inversely related to the direction of rotation. Thus at all times during testing the subject's head should remain facing forwards. Test Description and Procedure.--The apparatus, as shown in Figure 2, was placed at right angles to the -vv - 33 subject such that the mid-line of the apparatus was opposite the mid—line of the subject. The subject was seated close to the table so that rotation and lateral movement was as limited as possible. The subject was required to keep his head facing forwards throughout the test. Blindfolded, the subject was required to hold the small wooden blocks between the thumb and forefinger of each hand and slide them in their grooves from a per- determined starting position until he believed they were opposite, as shown in Plates 5 and 6. The subject was allowed to move the blocks back and forth past each other several times if he wished, in order to determine when they were opposite. Twenty trials were given in half of which the left block was furthest away from the subject,and in half of which the right block was furthest away. The order in which the predetermined positions were tested was selected randomly. The distance between the blocks, using the nearest edges to the tester, was recorded in centimeters. At no time during each trial was the subject allowed to place his arms on the table. However, he could rest them between trials. Single Foot Balance It is impossible to measure, bya.simple test, the perception of labyrinthine function without also measuring postural reflexes. Since the functioning 3A Plate 5 Parallel Blocks Test u‘ltvl Plate 6 Parallel Blocks Test 35 of the labyrinth and the postural reflexes are important elements in proprioception the author developed a Single Foot Balance test to measure the refinement of their integration and control. Test Description and Procedure.—-The subject, blind- folded and barefoot, was required to balance on one foot for as long as possible, being timed to the nearest sec— ond as shown in Plate 7. The right and left foot were tested alternately for three trials each. The subject was allowed two practice trials on each foot. Timing commenced immediately the non-weight—bearing foot was removed from the floor and it terminated as soon as either the weight—bearing foot moved or the other foot touched the floor. At no time was the subject-allowed to rest his legs one against the other. The time of each balance was recorded to the nearest second, and the results were calculated in relation to dominant and non- dominant weight—bearing limbs. Weight Shifting Proprioception is the perception of one's body in space and therefore the ability to equally distribute one's weight between the two weight-bearing limbs should show an ability to perceive one's body position in space. This perception involves feedback from both the muscles and the joint receptors and is therefore a measurement of the ability to integrate that feedback. 36 Plate 7 Single Foot Balance Test 37 Test Description and Procedure.—-The apparatus con— sisted of two similar bathroom scales which were placed on a wet towel to prevent them slipping, between 8 and 10 inches apart, depending upon the inside leg measurement of the boy. An attempt was made to standardize the angle between the legs and the pelvic girdle at 20 degrees, while standing with equal weight on each of the two scales. The half weight of the subject was determined. Then, blindfolded, the subject was required to move from stand- ing on one scale, as shown in Plate 8, so that his weight was equally distributed between the two scales as shown in Plate 9. The difference between the half weight and the weight shown on the right foot scales was the error, recorded in pounds. Each subject was allowed twenty trials, ten moving from left to right and ten moving from right to left. The sequence of the direction of weight shift was assigned at random. Distance Perception Walk Proprioceptive feedback is constantly being used in total body movement. Measuring the perception of static positions is possible through the measurement of the perception of specific joint angles. However, in the perception of movement, feedback from the muscle and joints enables the individual to perceive duration and rate. In order to measure this perception it is not so vital to measure the perception 38 Plate 8 Weight Shifting Test Plate 9 Weight Shifting Test 39 of specific joint angles or muscle tensions but to measure their integration in the total movement. This is only possible by using movement which involves a longer duration and a repetition of specific joint angles and muscle tensions than is available in the movement of a limb through a single arc. For this reason the author developed the Distance Perception Walk Test. Test Description and Procedure.--The subject, bare- foot, was required to look at a line twelve meters away, replace his blindfold and then walk to the line attempt- ing to stand as close to the line as possible. The sub- ject was then led back to the starting position by the experimenter and asked to repeat the attempt. This pro- cedure was followed for ten trials. Measurement was made from the toe nearest to the line if the subject stopped short of the line and from the front of the toe furthest from the line if the subject went beyond the line. All measurements were to the nearest centimeter and all measurements were treated as being positive errors. Analysis of Data The Pearson product-moment coefficient of cor- relation was calculated for each test to determine its reliability. The mean of each trial in each test was plotted to form a frequency polygon in order to determine whether learning occurs within the performance of the test. The frequency polygons of each test were also _‘ affira” “ AIVA" - UJVQV AO compared to show test variation from one month to the next. Bar graphs were used to illustrate the distri- bution of error in the weight discrimination and thick- ness discrimination tests. CHAPTER IV RESULTS AND PRESENTATION OF DATA Introduction The purpose of this investigation was to develop tests which would measure both fine and gross preprio- ception in children. Twenty one boys from the fifth and sixth grades of two East Lansing elementary schools were given a series of seven tests of proprioception on three occasions at intervals of one month. They were tested in small groups of two to four boys, each boy being randomly assigned to a test in the series then following the order of the series from that point. The tests developed and the order of the series was Weight Discrimination, Single Foot Balance, Parallel Blocks, Distance Perception Walk, Thickness Discrimination, Weight Shifting and Dot Repetition. The test designed by the author were the Parallel Blocks, Thickness Discrimination, Dot Repetition, and the Distance Perception Walk. The Weight Discrimination test was designed on the basis of Weber's work (5) and-other psychological studies (25, A3). The Single Foot Balance was modified from a test designed by Katsuta (28) and Al M—qu 1,...“ A2 the Weight Shifting test was developed from one used by Scott (Al). The subjects were blindfolded for each test and at no time were they informed of their performance while they were being tested. The Weight Discrimination test, with six variables, ten trials and using twenty one sub- jects, gave a total of 3,600 trials per month. Thus for the three series of tests Weight Discrimination demanded 10,800 trials to be made. The same number of trials was also necessary in measuring Thickness Discrimination. The Parallel Blocks, Dot Repetition and Weight Shifting tests, each having a single variable measured, with twenty trials and using twenty one subjects, each demanded 1,260 trials for the three series of tests. The Distance Perception Walk test, with ten trials and one variable, demanded 630 trials for the three series. The Single Foot Balance test with three trials and two variables demanded 360 trials. Therefore, in the entire study it was necessary to give 26,370 trials of the propriocep- tive tests. It was impossible to control temperature, humidity, barometric pressure, feeding habits, prior physical activity or the possibility that some subjects might practice between the series of three tests. It was noted that the environmental conditions were not con- sistent from one test series to the next, or even within each test series. J- p) ( ('f 3*: vlv pk ~V‘ Vs A3 Before the last test series one subject was forced to drop oug being confined to bed for six weeks. Treatment and Presentation of Data The data were analysed using the Pearson product moment coefficient of correlation to determine the reliability of the tests. The mean scores of each sub— ject were used in the calculation of the correlations in all the tests with the exception of Weight Discrimination and Thickness Discrimination. Total errors and weighted errors were correlated for these two tests. In addition to the mean scores, the best raw scores of each subject were correlated for the Single Foot Balance test. The data used for the correlations together with the standard error of the mean for each set of data, are tabulated in the appendices. The mean errors per trial are illustrated graphi- cally, in Figures 3, 5, 6, 7, 8, and 10, for each test to show the presence of learning during the test perform- ance. The curves for each test have been plotted on the same graph to illustrate any learning or maturational effect from one month to the next. For two of the tests, Weight Discrimination and Thickness Discrimination, bar graphs are presented, in Figures 2 and 9, to illustrate the distribution of error around the standard weight or thickness. AA Results The results of each test are presented and dis- cussed separately. The reliability of each test was determined by correlating the results of June with July, and of July with August. Weight Discrimination This test proved to be most reliable irrespective of whether errors or weighted errors were correlated (See Table I). Figure 3 shows the presence of a learning curve during each month's administration of this test. This curve levels off after five trials on each variable. The last four trials of this test shows a marked increase in errors which may be due to fatigue or some other fac— tor. No month to month variation is evident. Figure A shows the distribution of errors and weighted errors around the standard weight. It is evi- dent that more errors were made in comparing lighter weights with the standard. The lighter weights being more frequently judged to be heavier, than heavier weights were judged to be lighter. Errors of the first five trials of this test were also correlated and high correlations were ob- tained (See Table II). A5 Hmm.m mom.s mam. mes. Hmw.m msm.m mos. mos. mm.m mm.m mo.m o:.H :m.H m©.H om.ma ms.ma sm.Hm homhm oopnwfimz mm.mH wm.mH zm.:a gophm .w:< maze pmsws< masw \sfiss \mcss mpmefipmm coapmamhnoo mo hoppm mo opmocmpm pcmfioamacoo comnmmm pmdwsa mHSh snow new: on» mo poshm ommocmpm um5w3< mHSh coda I‘M! .QOHpmcHsHsomflo unmamzau.H mamas 8 7 6 5 McOccd 00:50:35 C003. TL nLOLLU coo: Mean Weighted Errors Mean Errors A6 WEIGHT DISCRIMINATION weighted Errors June ----- July Errors _-_- August 6 .. 5 h- 4 .. 1} 1 1 1 1 1 1 1 1 t 1 I 2 3 4 5 6 7 8 9 IO Trials 1100!! 3. Couperisos of three, monthly series of tests using errors end veighted errors. late the learning curve present eech month end thet there is little venetian tro- one month to the next. 33% 32.165. 88 93 .332. 833a 35 .268 as.» . 3d: 3.3 .323 g hereon on on 35¢» 33 :33. .930 P3P? egos and 3c. .233 283». .5 e592 :8». 32.33 as. n25 no 83255.8 3.... .e ESE ESQ: . EZSES .. 7////<\\\\\\§ .- ESE: .. ////<\\\\\\\\\ .- e. 7//<\\\\\\\_ WM///<\\\\\\\\\\ 9- _ _ O. m 0 m 0_ omom< om> LO I /////<\\\\\\\\ m - ‘33_I>- 4303(IN— "DDZLIJ <§ m . ?/<\\\\\\ m - 9. mg m_. o. . <\\\& o_- m . ////<\\\\\\\\\ m - m... Rm 9. o. . /<\\\\& o_- .. éw/KSSES .. omOm< mmommm ZO_._.:mom_Q .rIw_m>> omdozflrm Bomwm A8 TABLE II.--Weight Discrimination--First Five Trials. W Pearson Coefficient Standard Error of Correlation of Estimate June/July July/August July August f Errors .777 .609 2.50 2.59 Weighted Errors .81A .597 A.O2 A.85 V Discussion.--It is evident from these results that the Weight Discrimination test is most reliable and can be used effectively to determine the perception of fine muscle tension. It must be noted that at least five trials must be given on each variable in order to over— come the effect of the learning curve. It is interesting to note that more errors were made comparing lighter weights with the standard than were made comparing the heavier weights with the standard. Dot Repetition This test did not prove to be very reliable, its highest correlation being .556 as shown in Table III. Figure 5 shows the presence of a learning curve within the first four trials. It also illustrates the most erratic performance of this test from one trial to the next. There is also a possible test familiarity developed from June to July and August shown by the lower curve of the latter two months. A9 .EE mmo.m Hmm.m mmm. mma. .58 m:.: mm.m mo.m .EE mH.mH :m.mH om.om pmsms¢ maze pm5w3< maze mussfismm eo 909mm Uhmvcmpm \thh \m::% scepmflmmhoo mo pcmsoacumoo nonsmom pmsms< maze meow com: on» mo nommm ommocwum pmsws< masw mood mcmoz Coapfipmamm pOQII.HHH mam¢B 50 In correlating the first ten trials of this test, which takes into account the initial learning~curve, much lower correlations were calculated as shown in Table IV. TABLE IV.—-Dot Repetition-—First Ten Trials Pearson Coefficient Standard Error of Correlation of Estimate June/July July/August July August .052 .398 A.8. , A.A9 Discussion.--The results show that the Dot Repe- tition test is not reliable enough for use in measuring joint angle perception of the elbow and shoulder joints in children. It is evident from Figure 5 and the standard error of estimate in Table III that too much variation occurs from one trial to the next. The author would therefore not recommend the use of this test as a measure of proprioception. Distance Perception Walk The reliability of this test was found to be quite low as shown by the correlations in Table V. Figure 6 shows the presence of a learning curve which levels off after four trials. The curves for each month show little or no variation from one another. 51 .8313: 33.3 Bios 3253 B» e5 13.3 s the .5 n33: 58: 88 e888 :02: wags-S 3a: and: .5 3o- ..fioo no 3r:- hfiaoa .83» no socuonsoo .n E82 32:. ON w. w. v. N. O. m m .v N — — q a d _ u A _ _ + 1ou u! $10113 ZOFEMnEK HOG 52 .Eo mam.:m :om.:: :52. Hmm. .Eo mm.ma mm.oa o:.ma .Eo w.wm m.mm z.mma pmsws< hadh pmdwzd mHSh pmzwd< mafia o25h pmsw5< mash roam \sfiss \mcse mpweflpmm soapwammpoo new: map mo mcmmz mo Loppm mo gonna Uhmocmpm Uhnpcmum pcofiofimmmoo compmom .xfiaz eonpamosma moampmfianr.> mamas Error in Cm. Error m cm. 53 DISTANCE PERCEPTION WALK BOO '- _ June ZOCe- ----- JUW ---—- August 200 - ISO _ IOO - 50 - I l I l l I I I I 1 TriaIs FIGURE 6. Comparison of three, monthly series of tests. Note the initial learning curve present each month and the lack of between month variation. PARALLEL BLOCKS 5 I— June 4 ”' ‘‘‘‘‘ duly —-—- August 3- 2 ._ Trials FIGURE 7. Comparison of three, monthly series of tests. Note the initial learning curve and the lack of between month variation. 5A The first five trials of this test were correlated to determine reliability, which was found to be low, as shown in Table VI. TABLE VI.--Distance Perception Walk--First Five Trials. m Pearson Coefficient Standard Error of Correlation of Estimate June/July, July/Aug. July Aug. .330 .3A9 59.8 59.3 Discussion.—-The results show that the Distance Perception Walk test is not a reliable instrument for measuring the perception of movement. However, the low correlation calculated for the first five trials may account for the low correlations of the total errors. This test would not be recommended for use in any test series of proprioception at this stage. Parallel Blocks The reliability of this test was found to be con- sistently high, as shown in Table VII, irrespective of the position of the hands during pre test stage. Figure 7 shows a brief learning curve which levels out after three trials however the curves do not become stable until after the seventh trial. There was no variation from one month to the next. 55 .Eomsm. moo.a wmo. saw. .80 mm. on. mm. .80 om.a mo.a mo.m Ummzuom ocmm pcmcHEooIcoz .anms. mam. Hep. owe. .50 mm. mm. mm. .60 wm.a mm.H ~>.H oumzmom ocmm unmeano .anmo. owe. mam. New. .50 um. om. :m. .50 mo.H mw.H Hm.a mnonmm Hmuoe «msms< maze .ws< >H3h pwsws< haze mosh .m3< maze oczm \ssss \mcss cpmefiumm coapmamapoo new: on» mo mama: mo Lopsm no nopnm unaccepm onmocmpm pcmaoammmoo comhmom .mxooam H0HHMhmmlt.HH> mqm<8 56 When the first ten trials were correlated, fairly high reliability was discovered, as shown in Table VIII. TABLE VIII.--Parallel Blocks--First Ten Trials. Pearson Coefficient Standard Error of Correlation of Estimate June/July July/Aug. July Aug. .706 .800 1.0A .76A Discussion.—-The results show that the Parallel Blocks test is a most reliable instrument for the measure- ment of bilateral integration of joint angle perception in children. As in the previous tests it is necessary to test beyond the learning curve. The author would recommend at least eight trials when this test is used in a series to measure proprioception. Thickness Discrimination This test proved to be quite reliable with fairly high correlations as shown in Table IX. Figure 8 illustrates the absence of a learning curve in this test, whether error or weighted error were used. However, some test familiarity seems to have occured as shown by the lower curves of July and August as compared to the June curve. 57 mmm.w Hoo.> mam. mow. sa.m >H.m om.m mm.Hm :H.Hm mo.mm nomhm capzwamz :m:.m mmo.m sz. mmw. mm. mm.H mm.H mm.ma mm.ma :H.ma monhm .m:< mash .w:< mama .ms< mH3h anew .w:< haze anon \sfisn \mssm mpwefipmm coapmHmnhoo new: on» no mama: mo sounm mo noanm ommocmpm pudendum pcoHOfimmmoo QOmmmom .COHmeHEHhomHQ mmmCXOHQBII.NH mqm< // \//\V \ / y’\¢/ 7 - \ I ‘ ‘ 3 /~ ‘ / \ / \ I ’ \ / \ / ‘1, 6 - ‘ \ / ‘ t " I I I I I I I I I I l 2 3 4 5 6 7 8 9 IO June Errors ————— July 7 .. -—-——- August 6 ... \\ / “\ \ N.“ 5 I— \.\~J \”(‘;+‘\-/‘ // \/ V 4 .. 3 .. I I I I I I I I I 2 3 4 5 6 7 8 9 IO ate Trials FIGURE 8. Comparison of three, monthly series of tests. Note the improvement of test funiliarity from June to July and August, also the lack of any learning curve. 59 In Figure 9 it is evident that more errors were made in Judging the thicker thicknesses than were made in Judging the thinner thicknesses. Discussion.-—The results show that Thickness Dis- crimination test is a reliable instrument for measuring the fine Joint angle perception of the fingers and thumb of the dominant hand. Since no learning curve exists fewer trials are necessary and the author would recommend three or four for each variable only. However it must be noted that between June and July some test learning or familiarity caused the July curve to be lower. The July and August curves do not appear to be significantly dif- ferent. It is interesting to note that the thicker thicknesses were more frequently Judged to be thinner than their true value, than were the thinner Judged to be thicker. Weight Shifting Only moderate correlations were calculated in deter- mining the reliability of this test, as shown in Table X. Figure 10 illustrates the wide variation in per— formance from one trial to the next. No learning curve is present and there is no consistent month to month variation. Discussion.——The results show that even though there are moderate correlations for the Weight Shifting test, the wide range of variability from one trial to 6O .mnfizsm.a mmm.a mum. ems. .mnH om. mm. mm. .mndmm.m mm.m mm.m .ws¢ mHSh .w:< zHSh .ws< zHSW mcsm .m:< mfizh mash \sash \mcsw mumsfipmm soapmaohpoo cmmz on» ma madmz mo nonnm no nonnm vnmucmpm upmasmpm seafloaummoo cammmom .mcfinnfinm unwamzrn.x mamas 61 9 + .81» 8.3 hinge; on 3 venue... page 98! end See-noun» henna an» .3. e5 gage .530 Bake euol can» evo- .eeenuoa 3. e5 3:93 cup—he c3332. 6!. 38.3 no Sugars-3 29 .0 38¢: m. m 0 m 0. . 92$ 7/////Zr\\\\§ //////////. \\\\\\\\N. ID I 7/3/34 SS m7 0.- ////////////<\\\\\\\ m- ////////, \\\\ 0.- 5/9/32 SEES, ///////// \\\\\\\ m - _ O. m 0 m _ O. omdozflrm m>0m< 01402.55 304mm mmommm m.m.—.1995 ‘DD_J>- 43030)}- ”JDZLIJ 0. m 0 m 0. _ . _ _ p . a??? SEES. .. o: g///<\ _ o_- . 233/2 \\\\\\\\ . -1 m... g& 2- o: 7/////<\\\b o_- m .7////////// \\\\\\\\L_ O. n O m 0. 05324.5 m>om< omdozflhm 30.. mm mmommm to l 20....