The oblique effect has long been used to help explain some of the wiring of the retina and overall visual mechanics explanation. This effect, proposing that humans are more capable in viewing gravitational vertical and horizontal lines, has been studied under numerous conditions and applications. As the past and current literature has discovered, the presence of vertical and horizontal line receptors in the lower and higher centers of the visual system are in greater quantity than are those receptors detecting oblique lines or lines in a tilted-angle position - eliciting to this oblique effect (Campbell et al. 1966). DeValois et al. (1982) reveals that this increased quantity of vertical and horizontal line detectors or receptors implies that there is an intense sensitivity to orientation (e.g., vertical lines and horizontal lines). Furmanski and Engel reported higher fMRI activity in the visual cortex of subjects who viewed cardinal axes lines, rather than oblique lines (2000). Humans are essentially designed to ambulate in and respond to upright and purely gravitational conditions (Coppola et al. 1998).
The study at hand attempted to replicate a portion of the Marion Luyat and Edouard Gentaz (2002) study which also found the oblique effect to be influential and pervasive. Much of the study attempted to determine the amount of error (mean error) of the subjects' visual perception under two situations: determining the subjective vertical (subjective 0 degrees) in the Rod and Frame Apparatus for body orientations of vertical(0 degrees), tilted left (-45 degrees), and tilted right (45 degrees) and reproducing a series of angles preset by the experimenter which were 0 degrees, subjective vertical for each body orientation +45 degrees, - 45 degrees and - 90 degrees. The differences in this study were determined by the physical limitations of the rod and frame apparatus. The box orientations or the positions the Rod and Frame Apparatus box were placed relative to a gravitational vertical, were 28.5 degrees in either direction of gravitational vertical to represent a tilted right and tilted left condition. The reproduction of the various angles of the adjustment line also were different (i.e., 18 degrees in either direction instead of 45 and 90 degrees). Additionally, the biggest difference in this replication study was that instead of actual body orientation being altered the box of the rod and frame apparatus was altered (to simulate an oblique or skewed visual environment).
It was our hypothesis that as a result of this known oblique effect, the subjective vertical task and angle reproduction task would have contained more error (in regards to finding exact gravitational zero or finding the exact angle to replicate) in an oblique or tilted environment. Therefore, the subjects should have had more success determining the subjective vertical and duplicating angles at a box of a vertical condition than of 28.5 degrees left or right of the vertical. Finally, the amount of error that occured in the tilted conditions should have correlated to the specific angle of tilt (i.e., A subjective vertical in a box right condition should have been inaccurate with a measure of x number of degrees to the right).
Methods
Subjects:Participants consisted of 18 undergraduate students of Alma College of various majors or areas of study. These students consisted of both males and females and were between the ages of 20 and 22 years.
Materials: A Polymetric Company Rod and Frame Apparatus; Portable Model was used: serial number V-1260-AR; constructed in Hoboken, New Jersey. Its box portion measured 12 inches wide, 12 inches in depth, and 24 inches long. This device consisted of a steel, rectangular frame encased with translucent, white plastic sides. This structure was open on one end to allow the participant to view into it, the opposite and enclosed end. This end was characterized with an adjustable piece of the white, translucent plastic with a painted black line as its center, as it was also encased in a metal frame. The mid-section of the device was adjustable 28.5 degrees in either direction, as to give the appearance of a tilted or plumb environment viewpoint. Another frame of steel with circular plates on the long ends of the device (to indicate degrees of tilt and line position) supported the box.
Design: The subject was asked to determine their subjective vertical from a preset tilt or gravitationally vertical position. They accomplished this by turning the line-adjust knob clockwise or counter clockwise until a point of subjective vertical was reached (when they felt the line was truly straight up and down). This was completed for each position that was preset on the box of the apparatus. Box Left represented a -28.5 degree tilt relative to vertical, box Right represented a 28.5 degree tilt relative to the vertical, and Vertical box represented a vertical condition, 0 degrees, of the apparatus. Subjective verticals were determined by how close or far off of gravitational zero the subject places the line under each of the testing conditions (mean error). It was these subjective verticals that were used as one of the preset angles for the reproducibility portion of the test.
The second portion of the test was the reproducibility portion: we presented an angle of choice or standard orientations on the adjust line of -18 degrees, 18 degrees, 0 degrees, and Subjective Vertical under the different testing conditions Left (box left), Right (box right), and Vertical (vertical box) and asked the participant to view the specific line placement (angle) for as long as preferred. At this point the subject looked away from the stimulus and the standard orientation was altered positive or negative 10 degrees. The participant was signalled to reopen his/her eyes and to reproduce the angle. The mean error was found: the amount of error in degrees the subject veered from reproducing the exact angle. Again, each testing condition and standard orientation were repeated for 2 trials and the average error of each condition and standard orientation was determined as the mean error for the subject. Averages of the total mean error were calculated for each situation for the subjective vertical task and the reproducibility task.
Testing Procedure: Actual testing consisted of the subject determining the subjective verticals for Left, Right, and Vertical as two trials had been performed of each condition. The reproduction task followed after the subject had a chance to look out of the Rod and Frame Apparatus. The testing recommenced with the reproduction task: the subject replicated specific angles of -18, 18, 0 degrees and the subjective vertical averages from the previous task under the Left, Right, and Vertical Box conditions. After each series or set of angles replicated at a given box orientation, the subject was offered another break to look away from the box. The subject was complete after 42 trials with the rod and frame apparatus: 6 for the subjective vertical task and 36 for the reproducibility task.
Results
The vertical box orientation elicited a mean error of 0.223 degrees away from the gravitational vertical (leaning slightly towards a right dominance) The average mean error for the box right condition was 2.00 degrees away from the vertical (leaning in the direction of the box orientation - rightward); and the error for the box left condition was approximately -1.80 degrees away from the vertical (leaning in the direction of the box orientation - left). The average results for all subjects tested are displayed in figure 1. The box left orientation exhibited the most error (especially at 0 degrees = > -1 degree error; 18 degrees = 0.6 degree error). The 0 degree standard orientation did produce significant error across the board (the error of reproduction here was consistently inaccurate to a degree of greater than -1. At every other standard and orientation and box condition there was at least one person who had performed the reproduction task with no error (e.g., subject had returned the standard orientation without any noticeable error). The box right condition and the vertical box condition almost paralleled their amount of error in degrees from all box orientations and adjust line placements or orientations. Both of them varied by an amount that was significant (at least 0.6 of a degree) at negative -18 degrees. [The error for the negative 18 degree corresponds fairly equal throughout all of the box orientation conditions.] It was only at 18 degrees of the standard adjust line orientation that the two box orientations varied (and then it was only less than 0.1 degree). This standard orientation was elicited the highest error of all the standard orientations. A graph of the mean error for all standard orientations and box orientations is found on figure 2. The subjective vertical reproduction seemed to elicit the lowest amount of error (or greatest degree of accuracy) under the reproduction task with values of all three box orientations wavering between slightly greater than -0.1 to just less than 0 degrees.
Discussion
It seemed as though the oblique effect came true for this study, though the hypotheses were not completely supported. The subjective vertical task showed a considerably higher amount of error during the tilted conditions (i.e., Right and Left) than it did during the Vertical condition. Furthermore, the degree of error was related to the direction of the tilt. A left box tilt exhibited a negative error, whereas a right box tilt elicited a positive error. According to these methods, these differences between the two boxes are almost negligible. Under our system of measurement in this study and in the Luyat and Gentaz (2002) study, this seems to indicate that the specific orientation pulled the subject's judgment of accurate gravitational vertical line placement to that orientation. The reproduction task exhibited slightly different and surprising results. As noted, the right box orientation did not elicit much different results (in regards to proper return to standard orientation placement of the adjust line) than did the vertical box condition. Perhaps, the subjects did not consider the right box condition to have been too difficult an environment to disregard its cues, as the testing condition was not a totally controlled laboratory design. Again, both the vertical box and the box right condition were the most significant in error when the standard orientation adjust line was set to negative 18 degrees (18 degrees to the left). Perhaps, it was just the method in which the experimenters conducted their test that led a biased towards relatively equal results for two very different box orientations. The high amount of error for the box left at 0 and 18 degrees may have been related to the fact that most of the subjects also found it difficult to reproduce a negative 18 degree angle, though the exact mechanics are unknown. The error for the negative 18 degree corresponds fairly equal throughout all of the box orientation conditions, so it seems as though this was not an individualized error. The positive 18 degree angle seems to have received a high proportion of error across the board as well.
The relative degree of equal error in the box right and vertical box could also possibly due to the fact that as Shimamura and Prinzmetal (1999) suggests a 15 to 20 degree angle or line will elicit the greatest or strongest oblique effect, the inability accurately judge oblique or tilted lines (e.g., reproduce given lines of these standard orientations, finding subjective vertical lines under different conditions of box orientation). Additionally, Shimamura and Prinzmetal (1999) stated that certain height illusions at the Mystery Spot are experienced with the orientation or background frame is 18 degrees from the vertical in either direction. This finding is partially why two of the standard orientations of this study where set to positive and negative 18 degrees, with a 10 degree adjustment for the reproduction task. [Additionally, they are proportionate to the orientations of 45 degrees on either side of the vertical (one of the Luyat and Gentaz, 1999, orientations), with a 25 degree adjustment. The ratio for adjust degree to standard orientation line in both settings, the current study and for Luyat and Gentaz (1999) is .556]
As similar findings of Luyat and Gentaz indicate, that for a majority of the reproduction tasks, the oblique line replication was more difficult and having less accuracy in all orientations (besides the 0 and 18 degrees for Box Left). The trend may suggest that orientation has a much smaller part of the individual success of than the oblique effect may indicate. However, this finding may also seem to contain a certain degree of common sense since the subjects were not asked to locate an orientation independent of their visual environment. The reference norm or environment in which a subject could judge the angle could have been the proponent to leveling the playing field for all box conditions (e.g., as no one condition dominated in success of this task over the other). The neural network of line detectors of the visual cortex may have undergone a certain amount of adaptation under this condition. In other words, oblique line detection should be difficult in no matter what surrounding a subject is presented.
As Luyat and Gentaz (1999) discuss a finding made by Sauvan and Peterhans (1999), only 40 percent of the visual orientation receptors and/or neurons are hypothesized to have been preprogrammed or wired in at early developmental stages of a rhesus monkey. More receptors subjectively learn over time to be sensitive to gravitational axes. A similar hypothesis can be made to humans: as developing stages proceed there is probably more crossover of 'gravitational duties' to those visual orientation receptors. It seems to further be enforced that the oblique effect should override our ability to make conscious and accurate decisions of where a gravitationally vertical line is located under tilted or skewed conditions. It seems that humans are mechanically wired to be gravitational organisms. Though we are programmed to find a vertical line or horizontal line, our inability to detect oblique lines (because of a simple lack of neural components or lack of practice) contradicts and makes difficult the performance of the Rod and Frame task at those oblique conditions. What is more, the subjective vertical in the reproduction task seemed to be the key of the reference frame for the majority of the subjects. In other words, the gravitational / cardinal axes nor the oblique frame seemed to influence this result. Perhaps, the subject's gravitational, pre-programmed determination to prefer a gravitationally vertical position was influential. Finally, this study did seem to justify the use of the this Rod and Frame Apparatus in comparison to the Luayat and Gentaz method of having the subject physically alter head/body orientations to the right or the left. Though several differences or contradicting findings were found, it seems that the environmental cues or conditions of the procedure seem similar in both cases: The support for the oblique effect is maintained. However, it would also seem that a repeat of this study, using a greater quantity of participants and greater control of the testing procedure, may be necessary to find more of a concise pattern in the specific analysis of the oblique effect according to the boundaries of the Rod and Frame Apparatus.
References
Campbell et al. (1966). "The Effect of Orientation on the Visual Resolution of Gratings." Journal of Physiology (Lond.), Vol. 187, 427-436; cited in Goldstein, Bruce E. Sensation and Perception. (2002). 6th Edition. Pacific Grove, CA: Wadsworth Group. 89
Coppola, David M. et al. (1998) "The Distribution of Oriented Contours in the Real World." Proceedings of the National Academy of Sciences. Vol. 95. Issue 7, 4002-4006.
DeValois, et al. (1982) "The Orientation and Direction Selectivity of Cells in Macaque Visual Cortex." Vision Research, Vol. 22, 531-544; cited in Goldstein, Bruce E. Sensation and Perception. (2002). 6th Edition. Pacific Grove, CA: Wadsworth Group. 89.
Furmanski, Christopher and Engel, Stephen. (2000). "An Oblique Effect in Human Primary Visual Cortex." Nature Neuroscience, Vol. 3, 535-536.cited in Goldstein, Bruce E. Sensation and Perception. (2002). 6th Edition. Pacific Grove, CA: Wadsworth Group. 89.
Luyat, Marion.; Gentaz, Edouard. (2002) "Body Tilt Effect on the Reproduction of Orientations: Studies on the Visual Oblique and Subjective Orientations."Journal of Experimental Psychology: Human Perception and Performance. Vol. 28, No. 4, 1002-1011. Can be downloaded from http://www.univ-lille3.fr/Recherche/ureca/luyat/JEPHPP.pdf click to download file
Shimamura, Arthur P.; Prinmetal, William. (1999) "The Mystery Spot Illusion and Its Relation to Other Visual Illusions." Psychological Science. Vol. 10, No. 6. 501- 507. Can be downloaded from http://www.geocities.com/artsphotos/MystSpot.pdf click to download file
Suavan, X. M.; Peterhans, E. (1999). "Orientation constancy in neurons of monkey Visual Cortex." Visual Cognition. Vol. 6, 43 Ð 54; cited in Luyat, Marion.; Gentaz, Edouard. (2002) "Body Tilt Effect on the Reproduction of Orientations: Studies on the Visual Oblique and Subjective Orientations." Journal of Experimental Psychology: Human Perception and Performance. Vol. 28, No. 4, 1002-1011.
More Posters
Web Posters from other courses and projects
Alma College Psychology Department