A STUDY OF THE EFFECTS OF FOUR EDDY POSlTlONS ON THE VESUAL PERCEPTION OF VERTICALITY IN THE APP‘ARENT FRONTAL PLANE Thesls {or the Degree 0* M. A. MICHIGAN STATE UNIVERSETY Richard F. Haines 1962 LIBRARY Michigan State University ABSTRACT A STUDY OF THE EETBCTS OF 3003 BODY POSITIONS 0] THE VISUAL PERCEPTION O! VERTICALITY I! THE AIPABBNT IROHTAL PLAN] by Richard I. Raines Previous studies have dealt with the lateral or side-to-side tilt of the body while administering a perceptual test of vertical- ity. This body tilt seened to effect the position of an illuminated red when adjusted to the subjective vertical. The present study tilted 8s forward and backward while administering the perceptual test. The test consisted of placing the rod so that both ends of it were equal distance from S. The four body tilts that were used were: (1) that tilt giving vision 10' below the horizontal, (2) that tilt giving herisental line of sight, (3) that tilt giving vision 10' abeve the horisontal and, (4) that tilt giving vision 45° above the horizontal. Another variable was studied in this experinent. This was the effect of the initial position of the stimulus red on its final position. This variable was systematically studied by adnin- istering different initial red positions te 8 while he reclined in a contour chair that could be tilted into the above four degrees ef tilt. Twenty two ef the twenty feur 8s were volunteers from begin- ning psychology classes at H. 8. 0. during the sun-er tern. Twelve were nale and twelve fenale. Bash 3 adjusted the rod to the sub- Jective vertical himself. The true vertical was a plush line, j Richard I. Haines when 8' was in the 9' tilt position, and at other tilts- was defined as that position where both ends of the rod were equal distance ‘1'“ 8e An analysis of variance was perforned on the data obtained fron this experinent. The three Iain effects were chair position, red position, and sex of 8. None of these was found to be stat- istically significant at the .05 level. Three two factor inter- actions were also calculated and one was significant at the .05 level. This was the interaction of chair position with sex. The three factor interaction of chair-rod-sex was not found significant. These results were portrayed in graph forn and in a sohenatic dia- gran (rigure II) as well as being discussed. Although the analysis was not statistically significant the data did suggest that nale 3s showed a larger deviation fron a per- fect rod setting than did fenale Se in all tilts except the +45. tilt. In this tilt fenale Ss tended to show narhed deviations fren perfect settings. This finding was explained as being at least par- tially due to the experincntal conditions. Studies were discussed that related to this study and a sunnary included. Approved §;1飣L-’S?>onZLL;L Major rrcfessor Date WM 5‘. «on» A STUDY 01 T33 EFFECTS OF FOUR BODY POSITIONS 0! TH] VISUAL PERCEPTION O! VERTICALITY II THE AIPARBNT IROHTAL PLAN] By Richard I. Haines A THISIS. Subnitted to hichignn State University in partial fulfillnent of the requirenents for the degree ef' MAST]! O! ARTS Departnent of Psychology 1952 ACKNOWLEDGMENT The author wishes to express his thanks to Dr. 8. n. Bartley for his assistance and supervision as chairnan of the Master's connittee. He helped to nahe this an inter- esting and invaluable learning experience. Further, to Dr. Terrence Allen and Dr. hens Toch, go the author's sincere appreciation for their constructive suggestions and guidance during the preparation of this thesis. vror his help in the construction of the apparatus goes.ny thanks to Aaron Kolb of the Psychology Shop. Finally to ny wife goes the deep: gratitude she deserves for all of her help throughout the preparation and completion of this research. R. I. H. 11 II in IV INTRODUCTION NETHOD . . RESULTS . DISCUSSION . SUMMARY . REIERIIGBB . ATTENDIX . Table of Contents 111 17 23 30 Graph _ II List of Graphs rage lean scores obtained for nale fcnale 8s and initial rod position of 9 . . . . 19 Mean scores obtained for nale and f‘Iale 8s and initial rod positions of +20 and‘ -20 e e e e e e e e e e e e e 0 20 iv Figure II III List of rigures rage Results obtained by Schubert and Brecher for five different body pesitions . . . 7 Photograph of Apparatus . . . . . . 13 Results obtained in the present exper- inent for four different body positions . 21 Table I List of Tables r-Tests for the significance of difference of scores among experinental test conditions e e e e e e e e e e 71 Inge 18 Appendix List of Appendices Inge-I Sehenatie diagnan ef the plane of rotation for the stinnlus red and other planes for-descriptive purposes . . . . . . . 31 lean scores obtained for all test conditions (sale and fenale Ss separately) . . . . . 32 vii INTRODUCTION The subjective vertical can be defined as an individual's perception of the true vertical. The subjective vertical has been studied by nany investigators and it has been found that with a snall sidewise tilt of the body a lunineseent line in the dark will be perceived as vertical when it is tilted an appropriate nunbor of degrees in the opposite direction of the body tilt. This phenouenon has been called the B—phencnonon (after G. 3. Miller). In.nost of these studies 8 was tilted in the lateral plane, i.o., frcn side to side, and the effect of this body tilt has been noted upon his visual perception of the vertical. Killer (6) refined an earlier study by Aubert (3) and found that when the head was tilted very little the apparent tilt of the stinulus rod was in the cane direct- ion as the head tilt; but when the head tilt was greater the sttuulus rod was placed in direction opposite to the tilt of the head. Other studies have exauined the subjective vertical under nany different kinds of conditions. In cone of those studies (GE, 12, 17) the force of gravity upon 8 was altered by using a centri- fuge while in other studies (2, 4, 3, 9, 10, 13, 15, 39) the subject .was tilted. Several other experinents (1, s, 16, so) have studied the effects of tilting the whole environnent in which 8 sat during the perceptual test. An experhuont by Werner and Vapnor (14) dealt with the effect of the starting position of the stinulus rod 1 on the subjective vertical. 'This variable was studied in order to deternine if factors other than body tilt would effect the subject- ive vertical. In all of those studies 8 was required to adjust an illuninated rod fron sone starting position to the position that he felt was vertical. This final position was culled the subjective vertical. The basis upon which.nost of these studies were perforned seeuod to be an inplicit assunption concerning the various eleuents nahing up perception of the subjective vertical. hecdworth and Sehlosberg (20, pg. 525) nake this assunption ncre explicit when they write, '...your visual perception of the vertical is a eonpro- nise between that dictated by gravity and that shown by the visual field." This passage illustrates the physiologically oriented perceptionist's assumptions; for under the two headings of gravity and vision the other factors can be arranged. Sons of those researchers assuned that the behavior of adjusting a red to the subjective vertical involved the integration of nany physiological processes. The series of export-eats perforned by Werner, Vapnor, and Chandler on the sensory-tonic field theory of perception all dealt with visual perception of the subjective vertical. These experi- nenters perforned an oxperiuent (13) that attenpted to reveal seuo factors involved in perceiving the subjective vertical other than the visual factor. It dealt with the effect of supported and un- supported body tilt cn the visual perception of the vertical. The results of their oxperiuent indicated that under identical angles 3; of body tilt, the subjective vertical was displaced to a greater degree when the body was unsupported than when it was supported. Under conditions of being fully supported, 8 sat in a chair that was provided with a headrest, sides and back; thus the body was held firnly in place. Under relatively unsupported conditions 8 sat on a stool, and was required to naintain his body in the de- sired incline by his own nuscular efforts. Since the experinental task required S to naintain a given degree of tilt by using his own nuscles in the unsupported condition the oxperiuentors concluded that the greater degree of displace-ant of the rod frcu vertical was due to greater nuscular inyolvenent. They felt that the degree cf’nuscular involvcuent of S was nero in- portant than the degree of body tilt'pg£.gg. in causing the subject-— ive vertical to be displaced frcn a perfect setting. In this study So were tilted in the lateral plane, i.e., frcu side to side. So far as the writer has been able to doternine, only one study directly concerned with body tilt in the forward-backward direction is represented in the literature. This was a study done by Schubert and Brecher (10). These two researchers attcnpted to doternine the effect on one's visual perception of the vertical if he were tilted in the nodial plane (forward and backward) rather than in the lateral plane as had always been done before. They ad- uinistered the perceptual test at certain fixed degrees of body tilt. S's task was to adjust the stinulus rod to the subjective vertical so that both ends of the rod were equal distance fron S. This position was defined as the subjective vertical by Schubert and Brecher. In contrast to the body support used by Werner, 'hpncr, and Chandler: Schubert and Brechcr used a huge franc within which 8 stood. The franc had body fitting supports at the shoulder, waist, and ankle levels. The entire apparatus was nounted on an axle which pernittcd it to be rotated through a full 360.. The 8 within this franc could then be placed in any degree of tilt with respect to gravity while adjusting the stinulus red to the subjective vertical. The only places that S's body weight was supported were at the three body braces which fit the contour of his body. One exception to this was the upright tilt when 8 could stand in the franc. When 8 was not in the standing position he was in a relatively unsupported condition. The results of this oxperiuent showed that in the cans of a forward body tilt the stinulus red, in order to appear vertical, had to be tilted by a certain anount toward the body. In the case of a backward body tilt, it had to be tilted away. These findings are sinilsr in direction of rod setting to those found in cxpcrincnts applying lateral tilt of the body (5, c, 12, 13) even though this oxperiuent used nedial tilt of the body. Figure I schcnatically illustrates the results obtained by Schubert and Breoher for tilts of 90° backward (vision that was straight up), 45. backward, O. (vision that was horizontal). 45. forward, and 90. forward (vision that was straight down). Note that in general as the anount of body tilt increased frcn standing position the angular deviation of the subjective vertical also increased frcn the true vertical. 0. body tilt produced the snallest angular deviation of the rod frcn a per- 5 feet rod setting. Schubert and Brechcr's original study consisted of a different body tilts through tho entire 360° of rotation tho, first being the upright position and tho rest at each 45' interval. Only the above five positions were included in Figure I for clarity and relevance to the present study. Interpretation of these results nust take into account the kind of body support used by Schubert and Brochcr. Iran the pre- viously nentioned study (13) concerning supported and unsupported body tilt one would expect that Schubert and Brecher'e apparatus would here a large effect on the results. 3 ncrcly stood in the huge franc in tho 0' tilt and nude his red adjustnents. In the 45' forward and backward tilts, on the other hand, 8 had to adjust the rod while his body was pressing down on three scni-contoured forns that extended across his shoulders, across his waist, and across his ankles. This bodily strain also occurred in all of the other tilts that were adninistered. As long as 8 was not let out of the apparatus the strain would continue. It was felt that Schubert and Dresher's experinent was an in, portant one haying relevance to situations encountered in everyday life and that it should be at least partially replicated with bet- ter control of the body support. In order to do this a chair was built and two studies using it were carried out. The first of those was a prelininary study to deternine if the chair would provide ad- equate body support throughout the test period. This study was also an attanpt to dotennine if there would be an effect on S's percept- ien of the vertical when he was tilted in four different degrees of tilt. The second study that was perforned is that study reported in the present thesis. The second study differed frcn the first in that in the second study the stinulus rod rotated in the nedial plane (forward and backward) while in tho first study it rotatod in the frontal plane (fron side to side). Appendix A presents a schcnatic diagran.of the plane of rotation for the stinulus rod as well as other planes for descriptive purposes. The results of the prelininary study carried out by the writer showed that if S's eyes were kept closed he could not tell what do- greo of tilt he was in, especially when he was tilted very slowly into the position. The results also showed that the chair did not cause So to feel tired throughout the entire test. The degree of body tilt did effect the position of the subjective vertical but in an inconsistent nanner. This result would soon to warrant further study. The second study, presented in the present thesis, dealt with body tilt in the forward-backward plane as did Schubert and Brechor's. The difference between the two studies is twofoldp first, the present study tilted So in only four different degrees of tilt while Schubert and Brochcr tilted their Ss through the entire 360. in 45. intervals; and second, the present study provided full support for the body in each of the four tilts while Schubert and Brecher did not. It is a valid question to ask if body tilt in one plane has anything in connon with body tilt in another plane. Likewise when a variable is found that effects visual perception of the vertical in one plane it is a valid question to ask if that variable night 7 Figure I Results Obtained by Schubert and Brocher for five different body positions * Noun Setting of the Tilt Position Diagran of S's Position rod fron true vertical T.V. I Toward 8 Away fron 9o’ Backwand ct:::::3 o8 12.5 T.Vo o 45 Backward 1d? T.V. --n- o 0 4.5° ToVo O \ 45 Porward \\ 22.40 r.v. o 90 Porward i 15.5° 9° 1 ‘ (range of - uncertainty) Key Synbel prlanation = o o o o o “3‘1 tud 1“!» b0” “1. ---r- . . . . . Direction of 8's vision p,v, . . . . . True Vertical ( Rod's ends equal distance fron S) also effect visual perception of the vertical in another plane. Just such a variable was found. A previous study (14) found evi- dence that the initial position of the stinulus rod effected the resultant position. The subjective vertical, in general, was found to depend upon the position at which the rod was placed at the beginning of a trial. The question was raised whether or not this phenonenon would occur when the stinulus rod rotated in the forward- backward plane rather than in the frontal plane as it did in the hbove nentioned study. All three of the initial rod positions used? in the study gays confirnation of the statenent that, "the apparent vertical is always closest to the position in which the rod was set initially for that trial." (pg. 70) These three initial positions were all in the apparent frontal plane of 8. he nattcr what degree or direction of tilt S was placed in the stinulus rod always ro- tated in a piano that was 90’ to his line of sight. Three initial rod positions were used in the present thesis study although the rod now rotated in the nedial plane (forward and backward) so that the rod's top swung toward S and its bottcn swung away or vice versa. NETNOD Sxpcrinental Design: The four degrees of body tilt usediin this oxperiuent were labelled according to the nunber of degrees above or below hori- zontal vision extended when S fixated on the center of the stinulus nod. If vision was below horizontal it was arbitrarily called ninus (-). If it was above horizontal it was called plus (+). The ' four degrees of body tilt woro; -10“, 0', +10', and +459. Three initial positions of the stinulus rod were presented to S in each of the four body tilts. These initial positions were; 20. fron the position defined as true vertical toward 8 (so that the rod's top cone toward 8), 0., and 20. fron the position defined as true vertical away fron 8 (so that the rod's top receded fron S). The stinulus red was considered to be in the true vertical position when it was positioned at 90‘ to S’s line of sight. whoa S was tilted into cone degree of tilt and he fixated on the niddlo of the rod the true vertical position was taken as that position where both ends of the rod were equidistant fron hin. This was identical to the procedure used by Schubert and Brechor (10). This procedure was also used in order to keep the visual conditions as close as possible to those encountered in horizontal vision. Thus if S were looking horizontally the red would be truly vertical with 1O respect to gravity when it was 90. to his line of sight. The ex- perinent was not neant to deternine if S could estinatc the tilt he was placed in but to study the effect of body tilt upon the visual perception of the rod“s position had 8 been looking horizontally. By having 8 always try to pdace the rod in the cane positionIundor different body tilts an assessnent could be nadc as to the ancunt of angular deviation being produced by the tilt. Since five ninutes were given each 3 to adapt to a new degree of tilt it was prohibitive to randonly present all exporinental conditions of red position and chair tilt. It was for this-reason that the three initial rod positions within each degree of tilt were randenized and the order of presentation of the four tilts was rand- onized separately. rash of tho throo initial rod positions was assigned a nunber fron one through three and each tilt a nunber fron one through four. The table of randcn nunbers (11, pg. 484) was entered at a point chosen by putting consecutive nunbers fron 00 to 60 on slips of paper and picking one (without sight) for the line to betentered. This was repeated using nunbers fron 00 to 14 for the celunn to be entered. The order of presentation for the three: rod positions was deternined by selecting the nunbers one, two, or three in the order they appeared in the table and recording than directly onto the score sheets. The order of presentation of body tilt was deternined sinilarly. Stinulus Object: The stinulus object enployed in all of the test conditions was 11 an illuninated rod 20 inches long and 1?16 of an inch in disneter. The width of the rod subtended a visual angle of 39.65 ninutes of are. When the rod was placed in the apparent frontal plane and viewed for length it subtended a visual angle of 11.95“. The stinn ulus rod pivoted about its center on a shaft. The rod was painted white and was illuninated by a 7i'watt rod light situated 4 feet away and just below S's line of sight towards the rod. The rod looked white with a slight pink tinge. Apparatus: The apparatus consisted of a stinulus rod assanbly and a chair with controls. The stinulus rod assenbly will be described first. The stinulus red assenbly consisted of the illuninated rod, a dull black circular background, a protractor on the axis of rotation of the rod, and pulleys necessary for the rotary control of the red by S. This entire assenbly was supported by a rigid franework which was attached to the chair so that as the chair was tilted into various tilts the whole franowerk and stinulus assenbly tilted with it. This was suppose to keep the eyeball novenents snall if S fixr ated on the center of the stinulus red. In all cases 8 controlled the novenent of the red by scans of a series of pulleys and cord. The control knob was within easy reach of 8's right hand. hcasurenent of the ancunt of angular deviation of the rod was achieved by using a pretractor. It was situated on the axis of re- tation. A pointer was attached to the axis which was in turn 12 attached to the stinulus rod thus.giving a direct neasurenent of the ancunt of angular deviation fron the position of the rod that was 90° to S‘s line of sight. The pretractor, which was period- ically calibrated with a plunb line, was read to an accuracy of'} of one degree. Description of the chair is easiest by referring to the photo- graph found on the following page. The chair was constructed of two sides forning the contour and braced for torsion. The chair, stinulus rod, stinulus rod assenbly, and controls were supported on a stand which allowed the entire apparatus to tilt forward and baekh ward. Braces held the chair rigidly in each of the four degrees of tilt as required. An aircraft type seat belt was installed in such a position to help hold 8 in the chair when it was tilted in the sore extrcnc tilts (-10° and +45°). A conpletely adjustable headrest was attached to the chair. This headrest was padded for eonfort and also to help reduce noises during the oxperiuent. The head rest provided a.neans of holding S's head within a snall range of lateral novenent. It was re- adjustcd at the beginning of each new chair tilt to help relieve any neck tension due to shift of body center of gravity when the chair was tilted into a new position. A heavy weight canvas was laced to the sides of the contour chair. In the section of the seat where nest of the body weight was situated an elastic cord was used to lace the canvas to the sides, in other areas cotton cord was used. This elastic shock cord provided a "softer“ section of the canvas to help hold 3 into the 1'5 u .‘u'l ‘I.’ esvchsamd me .3535 AH sauna 14 chair in the norc cxtrcne tilts. In order to elininate all roon illunination neans were taken to block the little light caning in around the window shades. Over the entire length of the chair and stinulus red franeuork was stretched a black cloth. It hung down on both sides to forn a tun- nel. At its end was the stinulus assenbly. A reduction screen of black cloth was also used. This screen was in the frontal plane at a distance of 4 feet fron S's eyes. The circular hole pernitted only a view of the black circular background and stinulus rod (when illuninated). So of differing heights had to be adjusted in tho seat so that the black background was centered in the hole of the reduction screen. This arrangenent of cloth elininated essentially all of the light caning into the exporinental area. Subjects: Twenty four Se, 12 sale and 12 fenale, woro tested in this ex- perinent. Their ages ranged fron 18 to 26 years with an average of 21.4 yearsa host of these Ss were naive. All Ss were volunteers fron a beginning psychology class who received credit for their tine. Procedure: The S was not at the door of the exporinental roan by the experinenter and taken to a chair with desk. Each 8 was provided with his own score sheet upon which the experinenter recorded the following infornation as well as S's red settings: (1) nanc, (2) ago, (3) sex, (4) university najor, (59 date, (6) tine, (7) handed- 15 nose and (8) an apprexinate Snollen rating of 8's visual acuity. S was then seated in the apparatus, the seat belt passed over his lap: and secured snugly to the other side of the chair's franc. S's head was then centered in the headrest. When 8 said he was confor- table in the chair the roan lights were extinguished and a flash- light was uoed by the experinenter to point out S's control knob and light switch. During the ensuing dark adaptation period the following instructions were given verbally to 8. "You are about to take part in an experhnent con- cerning visual perception. As you saw, there is a white red in front of you. In your left hand is a light switch and in your right hand a knob to adjust the white rod. This knob will rotate the rod to any position. I want you to wait until you hear no say all right. Then you should turn on your light switch and open your eyes. ‘Iou should then innndiately begin adjusting the white red to .the position that nakes both ends of the red seen the sane distance fran you. Take your tine. Iou will have fran 10 to 13 seconds to nake every adjustnent. The inn portant thing to ranenbor is that I want a careful set- ting of the rod. Again, there is no hurry. Please keep: your eyes shut except when you turn your light on. When you fool you are done adjusting the rod all you have to do is sinply turn off your light and shut your eyes. When you do this it tells no that I can record your rod set- ting and reset the rod to a new starting position. 'Do: you have any questions?" Measures Inpleyed: The angular displacenent of the stinulus rod that 8 undo was 16 neasured in degrees fron the rod position that was 90. to S’s line of sight. In recording the red settings a plus (+) was used to indicate that the top of the rod was away fran S and a.ninus (-) to indicate that the rod's top was toward 8. Within each degree of chair tilt each initial rod position was adninistered twice to each 8. The arithnetio uses of these two settings was used as the neasure of S's perfornancc. RESULTS Table I presents the results of the statistical analysis perforned on the data fran this exporinent. The variables of sex, chair position, and rod position were analyzed. The interactions of these nain effects were also analyzed. lone of the three nain effehts was found to be statistically significant at the .05 level. Three two factor interactions were calculated and one was found to be statistically significant at the .05 level. This was the interaction of chair position with sex. The three factor interaction of chair-rod-sex was not found to be statistically significant. hean scores for designated chair tilts and all initial rod positions are presented in Graphs I and II. The ordinate rep- resents the anount of angular deviation of 8's setting fron a perfect setting. The abscissa represents the degree of chair tilt. hale settings are graphed separately fron fenale settings. Generally nale Ss set the top of the stinulus rod closer to thenselvos than did wonen Ss. One exception to this occurred in tho +45' body tilt where fcnalo Se tended to set tho top of tho stinulus rod two degrees closer to thenselves fron a perfect set- ting than did nale 8s. Thus fenale Sordid a better job or locating the required position than did nale So when they were not being tilted extrenely. 11 18 rdno I P-Tests for the significance of difference of scores anong experimental test conditions Source of Variance SS. df MS P P* SCI 30.40 1 BOe‘G e07 Subjects within Sexes 8732.24 22 396.92 chair Position 343.3e 5 114.46 .ee Chairs Set 1018o20 3 359o40 2392* ChairttSubjects within Sexes; 7688.53 66 116.49 10‘ Position 477e3° 2 238.65: 1 on ‘0‘. N 301 35°o47 2 175o23 1 co, Rod % Subjects within Sexes: 7077.88 444 160.88 Chairs Rod 107.96 6 17.99 .25 - Chair as Red x Sex 260.33 6 43.42 .61" Chairs led a Subjects within Sexes: 9631.71 132 72.96 Total 35,71e.oo 2e? 1 Significant at .05 level When tiltod forward 10' both son and wcnen tended to sot tho top of tho stinulus rod toward then and when tilted back 45' thoy tended to set the top away fron then. This finding is sinilar to that found by Schubert and Drooher (10). Conparison of Graph I with Graph II shows that tho o° initial rod position load to rod settings (top of rod) closer to S than did initial rod positions of +ao' or -2o'. It is obvious that both nale and fenale So did not nake perfect rod settings when a perfect setting was defined as that 19 Graph I lean scores obtained for nale and fe'ale 8’s and initial red position of 0 led; Position J 5o 1. Male Settings Top \ 2' * toward 8 1. J. y / oe Perfect Setting, 1° , Top o 1 Panalc Settings away 2 ” fron 3o , ‘ . . s -1o’ o' +1o‘ +45.“ Chair Position 20 Graph II hean scores obtained for nale and fenalo 8's and initial rod position of +20 and -20 Rodi Position o T ~20 halos; -20° Penalos OP 3% ’ j // toward Perfect Setting /,/’ 4/ ----- -m- u 1' . “W' ,‘ ' // 1'? 2. +209 Penales' 3, w ‘ ay o halos fron 3e . 8‘ -1o' 0‘ +1 o‘ ’ +45} Chair Position 21 Figure III Results obtained in the present oxperiuent for four different body positions Tilt Position Mean Setting of the ' Diagran “f 3 ' r"1#1°‘ rod fron true vertical 45. Backward T. o Toward 8 Away fron S halcs=-.52° Penales +.1o /\ ToVo ,_,s 1 halos +1.3“ 10? Backward xq& Pcnales +1.2. T.V. e halos 4.06. 0 Panales:+.25t’ i 1 T.V. 10. Porward [9]--.‘ halos -1.72 Penales +1.2. Key Smnbol Explanation [2:22:22] . . . . . Longitudinal body axis ---o— . . . . .' Direction of 8’s vision r.v. . . . . . True Vertical ( nod's ends equal distance fron 8) 22 red position where both ends of the rod were equal distance fron 8. Both graphs show the ancunt of deviation of 8's settings in degrees fran a perfect setting. hale Sc tended to nake less erratic or wandering settings fron a perfect setting than did fanale Sn. This' was neat clearly illustrated in Graph II. Intrapncx settings were nere consistent than inter-sex settings. Pigure II on the preceding page presents the noun rod settings obtained in the present oxperiuent in a fern shnilar to that of Schubert and Brecher's results (presented in Pigure I). The di- rectional trends of 8's rod settings are the cane although the size of the angular deviations is snaller in Pigure II. DISCUSSION The present study dealt with perception of the vertical and sane of the possible factors related to its distortion. Per the present study distortion of the vertical was taken to nean any red setting that was not in the position where both ends of the rod were equal distance fran 8. At each of four discrete body tilts in the forward-backward plane 3 had to adjust the stinulus red as that both its ends were equal distance fron hin. This study at- tanpted to deternine if any distortion would accrue fran the body tilt. The results of the present study did not lend support to the thesis that different degrees of body tilt will effect the posit- ion at which an illuninated rod is seen as having both its ends the cane distance fron S. The analysis did find a significant two- factcr interaction however. The nain effect of rod position was~ also not significant. Although the nain effect of chair position was not found to effect average rod setting to a statistically significant extent, it did produce a significant interaction with sex of subject. It appears that nale Ss showed a larger deviation fron a perfect set- ting thanrdid fenalo Se, except when tilted back into the +45 tilt, where fenales' rod settings show a.narked deviation fron a perfect setting. In the +439 tilt fenalo Ss night have becone acre 23 24 distracted fran the task since this degree of tilt is unusual for a wanan wearing a skirt. Although neither the effect of the rod's initial position or the rod by sex interaction were statistically significant, the data suggest that if the initial rod position is 0., which is already a perfect setting, nale Ss set the top of the rod closer to thenselvcs than if the rod were begun either closer or farther away fron the 0¢ position. Penale Ss tended to do dust the opposite of what the solo So did, i.e., they consistently set the o' initial positioned rod farther fran thanselvcs than if the initial position had not been 0'. This would seen to suggest that sales react differently to the starting position of the rod than do fenalo Ss. The nale settings are nore regular than are fenalo settings and the degree of body tilt does not effect sales, and thus their red settings, as such as it does fenalo Ss. Werner and Wapner's study (14) also ob- tained sinilar sex differences. Graph II illustrates the finding that red settings are con- sistent within each sex and less consistent between sexes. This again suggests certain differences between nale and fenalo Ss. It scans to be a consistent finding of this oxperiuent that Ss nust move the stinulus rod in any direction before they will accept one position as being the required position. 'Witkin (15, 17) found narked sex differences. One of these studies uncovered a "field de- pendence” in wonen when they were laterally tilted within a tilted environnent. hale Ss relied nore upon the "physical aspects” of the situation when in the cane degrees of tilt. 25 The writer takes the position put forth by Wapner, ct al. (13, pg. 349) when they write, "the crucial postural factor (in causing distortion of the vertical) is not position per se but de- gree of nuscular involvenent. It is this factor which...contributes to changes in perception." The body support used in the present oxperiuent should have kept nuscular involvenent to a.nininun when compared to the body support used by Schubert and Brecher. As far as the writer can determine the other expernnental conditions of the two studies are the cane. Thus if nuscular involvenent is a factor in effecting one's visual perception of the vertical the snaller angular deviations fron a perfect rod setting found in this study could be accounted for by the snaller ancunt of nuscular involvenent. Better techniques are needed with which to study this phenomenon before any definite statenents can be nade. The previously nentioned study by Wapner, et al. (13) showed that the degree of deviation fron a perfect rod setting is greater for the unsupported than the supported condition. This study also showed that the displaoencnt increases (fron a perfect setting) if the body tilt is increased fron an upright position. These findings would soon to support the above contention that the snaller angular deviations found in the present study, as conpared to the findings of Schubert and Brochcr (10), are due to a snaller ancunt of nuscular involvenent in nain- taining one's posture during the experincnt. SUNNANT The present study was designed to investigate further the effects of postural changes upon the visual perception of subjective verticality in the apparent frontal plane. Differing fron nost of the previous studies, this study tilted the body in the nedial plane, i.e., forward and backward. Pour different degrees of tilt were used in this plans, they were: -1o‘, o‘, +1o', and +45'. The stinp ulus rod rotated in the sane plane as the body tilted in, i.e., the nedial plane. It was adjusted by S to that position naking both ends of the rod look equal distance frcn hin in each of the four tilts. In addition to noting the body tilt effect the variable of initial rod position was exanined. Three different initial rod pos- itions were administered in each of the four tilt positions. Twenty four Se, 12 nale and 12 fenalo were tcsted. They sat in a full length canvas contour chair in order to support the body without the involvenent of any najor muscles. The nain results showed that there was no significant effect of body tilt on the visual perception of subjective verticality in the apparent frontal plane. The position at which the stinulus rod was set initially did not significantly effect tho setting of tho rod. The two factor interaction of body position and sex appeared signif- icant at the .05 level. 2E5 27 The results of the present study were discussnd in the light of previous studies that had tilted Ss while noting the effect on their visual perception. The significant sex by body position inter- action was discussed as well as the other non significant findings. Two graphs and an illustration (Piguro II) were included to por- tray the findings...which suggested that it wasn't the degree of body tilt itself that caused the distortion of the subjective vert- ical but the degree of'nuscular involvenent during the experinent. 1o 2o 3. 4. Th 6. 7. 9. BIBLIOGRAPHY Asch, S. B. and Witkin, H. A., Studies in Space Orientap tion. I. Perception of the Upright with Displaced Visual ’101d.e Jo mo Psycholo, 1’48, 38’ 325' 337. Asch, S. I. and Witkin, H. A. Studies in Space Orienta- tion. II. Perception of the Upright with Displaced Visual fields and with Body Tilted. J. Exp. Psychel., 1948, 38, 455-477% Aubert, H. line scheinbare bedeutende Drchung von Objecten bei Reigung des Kopfes, Virchew's Arch., 1861’ 2°, 581-3’30 Pischer, H. H. hessend Untersuchungen Uber die Gegonrol- lung dor Augcn und die Lokalisation der scheinbaren Vertikalen.bei seitlichct Icigung des Gosantkorpers bis an 560 . Graefes Arch. Opthal., 1950, 123, 476-508 o hann, C. W. and Berry, I. h. The Perception of the Post- ural Vertical: II. Visual Pastors, Joint Report lo. 5, Tulane University and U. S. I. School of Aviation hedicine, Pensacola, Pla., 1949. hdller, G. B. Uber das.Aubert'sche Phfinonen. z. Psychel., 1916, II: 49, 109-244. Noble, C. B. The Perception of the Vertical: III. The Visual Vertical as a Punction of Centrifugal and Gravitational Perces. J. Exp. Psychel., 1949, 39, 839—830. Passcy, G. I. Perception of the Vertical: IV. Adjustnent to the Vertical with Vernal and Tilted Visual Prunes of Reference. J. Exp. Psychel., 1950, Pearson, R. G. and lauty, G. I. Proprioceptive Percep- tion of Verticality. J. Bxp. Psychel., 1960, 59, 425. 28 10. 11o 12o 15. 14. 15. 16. 17o 18o 19o 20o 29 Tschernak-Scysenegg, Arnianon, Introduction to Physio- logical Optics. Springfield: Charlcs C. Thanaa Publisher, 1952, Pp. 210-212. Walker, I. h. and Lev, J. Statistical Inference. lew- Iork: hoary Belt and Co., 1933, Ppa 462-463. Wapner, 8., Werner, H., and horant, n. Bxperinents on Sensory-Tonic Pield Theory of Perception: III. Effect of Body Rotation on the Visual Perception of Verticality. J. pr. Psychol., 1951, 42, 351-357e , and Chandler, K. A. prorinents on Sensory-Tonic Pield Theory of Perception: II. Effect of Supported and Unsupported Tilt of the Body on Visual Perception of Verticality. J. Expo Psychol., 1’51, 42, 3““350e Werner, h. and Wapner, S. lxperinents on Sensory-Tonic Pield Theory of Perception: IV. lffect of Initial Position of a Red on Apparent Verticality. Jo II’o IIIOhOIo, 1,52, 43, ‘8-74o Witkin, h. A. Purther Studies of Perception of the Up- right when the Porce acting on the Body is changed. J. Exp. Psychol., 1952, 43, 9-20. ' - . a a - c ...... , Perception of Body Position and of the Position of the Visual Pield. Psychol. Monogro, 19", 63, Nae 7p (VhOIO Nd. 302)o , Perception of the Upright when the Porce acting on the Body is changed. J. lxp. Psychol., 1950, 40, 95-106. , and Asch, S. 3. Studios in Space Orienta- tion: III. Perception of the Upright in the Absence of a Visual Pield. J. Exp. Phychol., 1948, 38, ‘93-‘14e ' , and , Studies in Space Orienta- tion: IV. Purther Bxperincnts on Perception of the Upright with Displaced Visual Pields. J. Expo Psychol., 1948, 38, 762-782. - Woodworth, h. S. and Schlossbcrg, 1., lxperinental Psychology (Revised Bd. ). New York: Holt, 1954. ’- APPENDICES 31 Appendix A Schenatic diagran of the plane of rotation for the stinulus rod and other planes for descriptive purposes tinulus Red Horizontal Plane Line of Sight Lateral Plane £{;::ial Plano Appendix A illustrates the stinulus red, line of sight, and three najor planes. S's eyes are considered to be at the inter- section ef the three planes (Point A). The stinulus rod rotated about the axis shown above so that its top cane toward S while its botten receded fron S or vice versa. The true vertical was that position of the rod where both of its ends were equal distance fran 32 Appendix B hean scores obtained for all test conditions (nale and fenalo Ss: separately) Male Chair Position -1o' 0' +1o +45 .20. 9" o8? +0e1’ 9'2 o82 9°e74 Rod 0' -2.41 -o.76 -o.74 -1.mu Position +2o' +0.96 +o.59 +1.99 +0.92 Polale Chair Position -1o' o' +19. +45? -2o° +1.2e -o.37 +1.11 ~2td4 Red o' +2.15 +0.64 +1.73 o2.02 Position +2o° +0.16 +o.49 +0.75 ~1.26 s FM. A no u .m .- #11... H, .. 1t .9 "WIT {1111111111111 flfifllflfifllflililjlfliflilfifil'il'Es