The aim of this study was to test the basis of the horizontal-vertical illusion. As part of a voluntary course activity, 68 female and 42 male students from Western Sydney University were asked to distinguish whether the horizontal or vertical line in a plus sign (+) is longer. Using the software program, Open Sesame, a stimulus was presented on the screen for 500ms and participates were required to record their response with the keyboard. The point of subjective equality (PSE) was determined through line orientation, which was measured on two levels (horizontal and vertical) and contrast polarity of the stimuli, measured using three conditions (black, white, black and white). Results indicate that vertical lines do appear longer than horizontal lines when presented together in a plus configuration, supporting the hypothesis. Contrary to expectations, it was concluded that the black and white stimulus provided a larger illusion of the three conditions.
The Effect of Stimulus Context and Contrast Polarity on the Horizontal-Vertical Illusion
Despite being one of the most elementary visual effects, the horizontal-vertical illusion remains poorly understood (Mamassian & de Montalembert, 2010). The horizontal-vertical illusion is the effect that a vertical line is perceived as longer than a horizontal line of the same physical length (Zhu & Ma, 2017). This phenomenon is among several illusions where a central aspect of a simple line image, such as the length, straightness, or parallelism of lines appears distorted by other elements of the image, such as the background, foreground line, or other intersecting shapes. The strength of the horizontal-vertical illusion is affected by the shape of the visual field, body orientation and, stimulus context. Armstrong and Marks (1997) found that psychophysical judgments are greatly influenced by stimulus context, which can be highly sensitive to such factors as the range of the stimuli, how often it is presented, and the sequence of presentation.
Several studies have shown that the spatial arrangement of the horizontal and vertical components, or the stimulus configuration, is an important determinant of the illusion magnitude (Cormack & Cormack, 1974). The most common representation of the illusion is the inverted T figure, that combines two different effects: the vertical line differs not only in its orientation, but it also bisects the horizontal line, which is known to magnify the apparent size difference (Li & Durgin, 1017). To the contrary, Cormack and Cormack (1974) report that rotating the T figure by 90° would give a negative illusion, that is, the horizontal line would appear longer than the vertical. Many alternative methods of manipulation can modify the magnitude of the illusion, for example, simply separating the lines so that they no longer form a unit tends to produce a slightly larger effect. Marks (1992) claimed that differential context effects are not universal but depend on the stimulus dimensions. For example, shifting the physical lengths of lines presented in different colours had no differential effect on judgments of relative perceived length, whereas shifting the physical lengths of lines presented in different spatial orientations did.
Based on the results and recommendations of previous research, the aim of the present study was to examine the relationship between contrast polarity and distinguishing line lengths as part of the horizontal-vertical illusion. To achieve clear and accurate results, the study measured line orientation (independent variable) on two levels: horizontal and vertical. To extend the research of Spehar (2002) contrast polarity of the stimuli (independent variable) was examined under three conditions: black, white, black and white. It was therefore hypothesised that 1. Vertical lines will appear longer than horizontal lines when presented together in a plus configuration; and 2. The size of the illusion will depend on the contrast polarity of the stimuli. All black and all white stimuli will show a larger illusion than mixed black and white stimuli.
The results of the current study support the hypothesises that vertical lines will appear longer than horizontal lines; a small but significant difference was evident. However, the hypothesis that the size of the illusion will depend on the contrast polarity of the stimuli was only partially supported. Contrary to expectations, the results indicate that the black and white stimulus provided the largest illusion of the three conditions.
According to Armstrong & Marks (1997), increasing the physical length of the horizontal line over the vertical leads to an increase in the perceived length of the vertical, suggesting that the perceived length of lines presented horizontally and vertically depend on the values of the stimuli presented in the two orientations. The overestimation of the vertical segment relative to the horizontal one is consistent with the findings of Mamassian and de Montalembert (2010). Cormack and Cormack (1974) reveal that the inverted T figure yields a larger illusion magnitude than the plus configuration, which may be attributed to the bisection illusion. In the plus configuration, both lines are bisected, perhaps cancelling or attenuating the bisection effect.
Several limitations have influenced the findings of the present study. For example, participants completed the horizontal-vertical illusion experiment within their own time, on their personal devices, in an uncontrolled and unsupervised environment. Factors such as the angle of the screen, brightness, and distance from the screen were not measured when collecting the data. A lack of prior research in the field of contrast polarity restricted the study because the results were unable to be compared.
Future research within this field should further establish the relationship between the horizontal-vertical illusion and contrast polarity. More work in needed to examine the precise characteristics effecting the magnitude of the illusion, and to what degree they are relatable to everyday life.
In sum, the aim of the present study was to examine the relationship between contrast polarity and distinguishing line lengths as part of the horizontal-vertical illusion. The current study has contributed to the understanding of a complex phenomenon.