Free
Article  |   July 2015
Eye dominance in binocular viewing conditions
Author Affiliations
  • Jan Johansson
    Department of Clinical Neuroscience Karolinska Institutet, Stockholm, Sweden
    jan.johansson.1@ki.se
  • Gustaf Öqvist Seimyr
    Department of Clinical Neuroscience Karolinska Institutet, Stockholm, Sweden
    gustaf.oqvist.seimyr@ki.se
  • Tony Pansell
    Department of Clinical Neuroscience Karolinska Institutet, Stockholm, Sweden
    tony.pansell@ki.se
Journal of Vision July 2015, Vol.15, 21. doi:10.1167/15.9.21
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Jan Johansson, Gustaf Öqvist Seimyr, Tony Pansell; Eye dominance in binocular viewing conditions. Journal of Vision 2015;15(9):21. doi: 10.1167/15.9.21.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Different degrees of eye dominance may need to be considered when comparing monocular and binocular performance or estimating binocular summation effects. The purpose of this study was to explore eye dominance under binocular viewing conditions and observe gradual differences in preference. Two binocular eye-dominance tests were evaluated and compared to the hole-in-the-card sighting test. The first experiment was based on a binocular sighting test. The second originated from the variable-angle mirror test, utilizing physiological diplopia to determine which eye dominates the binocular percept. The participants were 32 healthy subject with normal sight. For both experiments there were plausible effects of different degrees of eye dominance affecting how the subjects positioned or perceived the scene. The outcomes were supported by a statistically significant correlation with an interocular difference in sensitivity to degraded visibility. A statistically significant correlation was found between the hole-in-the-card test and the variable-angle mirror test (r = 0.527, p < 0.01). The experiments confirm the plausible occurrence of forms of eye dominance under binocular viewing conditions. The correlation with the hole-in-the-card test was weak to moderate, due to factors that appear to be related to testing and viewing conditions. Interocular differences in sensitivity to blur appear to be a useful factor in further grading of eye dominance.

Introduction
From time to time, the question turns up of which eye is the dominant, or preferred, eye. This may be in the context of presbyopic correction, such as monovision, or when binocular balancing of refraction is difficult due to anisometropia. In research experiment, one eye may need to be chosen due to experimental limitations or the expectation that eye dominance could influence visual performance. 
In a previous study (Johansson, Pansell, Ygge, & Seimyr, 2014), monocular and binocular reading performance was compared in subjects with normal binocular vision. The study failed to find a significant correlation between reading performance and sighting-eye dominance. There was, however, a certain trend demonstrating that the greater the interocular difference in reading performance, the more likely the faster reading eye agreed with the sighting dominant eye. A similar observation was made in a study that evaluated reading performance in subjects with a pronounced eye preference (Spache, 1944). These observations led us to hypothesize that an asymmetry in monocular reading performance may be related to different degrees of eye dominance. Accordingly, the degree of eye dominance may be necessary to consider when estimating the binocular component in visual performance. 
Some studies that have made functional comparisons between the dominant and the nondominant eye have found indications of superior performance for the dominant eye. The literature, however, is not concurrent, and studies on different grades of eye dominance versus performance are scarce. 
One study compared monocular and binocular performance in a number of occupational-type tasks (Sheedy, Bailey, Buri, & Bass, 1986). These ranged from visual-manual tasks (e.g., pointers and straws, needle threading) to reading and letter counting. Eye dominance was determined with a sighting task and without any grading of the dominance. In the monocular comparison, only four of the 13 subjects showed a statistically significant better performance when using one eye over the other. In these four subject, the agreement between performance superiority and sighting eye dominance was 75%. However, for the group as a whole the agreement was only 46%. These findings suggest that sighting dominance tests may predict performance superiority when there is a pronounced performance difference. 
Another study examined conjugate eye movements at reading distance (40 cm; Oishi, Tobimatsu, Arakawa, Taniwaki, & Kira, 2005). The study included 20 subjects with a dominant right eye, determined with a sighting task (near–far alignment). It was found that horizontal saccade velocity was higher in the dominant eye for both adducting and abducting saccades. Since no subjects with left-eye dominance were included, no direct comparisons were made of performance at left-eye dominance. Vergilino-Perez et al. (2012) investigated horizontal saccade parameters in 45 subjects with an approximate 50-50 split between right-eye (RE) and left-eye (LE) dominance. The dominance was determined with a sighting task. The viewing condition was binocular, but only the movements of the dominant eye were recorded. Larger amplitudes and faster peak velocity were found in saccades made towards the ipsilateral side of the dominant eye. It was distinguished that for some of the subjects the differences could be explained by eye dominance, while for some it could be explained by adduction–abduction differences. It was suggested that this difference could in part be explained by different degrees of eye dominance. 
Shneor and Hochstein (2006) applied nonrivaling dichoptic presentation in a visual search task for 21 subjects. Thirteen subjects were RE dominant and eight were LE dominant, according to a sighting task (hole-in-the-card). It was found that the dominant eye consistently performed better in terms of detection. Subsequently, it was suggested that the dominant eye has priority in visual processing. Another visual search experiment (Porac & Coren, 1979) involved eight subjects with RE dominance and eight subjects with LE dominance. Eye dominance was determined with two sighting tests (Miles ABC and point test). In this study it was also found that at dichoptic presentation, the dominant eye performed better in terms of correct detection. However, when comparing pure monocular performance (one eye occluded) in the same experiment, the eyes' performances were equal. These findings suggest that the viewing condition is an important factor in estimating performance related to eye dominance. 
Two studies have investigated the effect of different degrees of eye dominance on binocular function. One study (Nitta, Shimizu, & Niida, 2007) measured the degree of sensory eye dominance (binocular rivalry method) in 10 subjects. Binocular visual function was then compared as a contact-lens reading add was added to the dominant or the nondominant eye. The subjects were divided into two groups based on the strength of eye dominance. It was found that in subjects with strong sensory eye dominance, the binocular summation decreased at low spatial frequencies. The other study compared binocular summation of contrast sensitivity as optical defocus was induced to the dominant and the nondominant eye (Handa, Shimizu, Mukuno, Kawamorita, & Uozato, 2005). The degree of sensory eye dominance was determined in 20 subjects. The subjects were then split into one group of 12 with weak dominance and one group of eight with strong dominance. It was found that a summation effect in contrast sensitivity was only maintained when defocus was induced to the nondominant eye. In the group with strong eye dominance, there was a statistically significant difference in contrast sensitivity when defocus was induced to the dominant and the nondominant eye. 
From the research described here, it appears that the degree of eye dominance, along with differences due to viewing conditions, needs to be considered when estimating the effects of eye dominance. 
There is limited recent research on methods to quantify eye dominance under binocular viewing conditions. The studies have predominantly used dichoptic presentation of stimuli to estimate the imbalance in contribution of either eye. The research has included tasks such as form discrimination (Yang, Blake, & McDonald, 2010), motion coherence (Li et al., 2010), target detection (Valle-Inclán, Blanco, Soto, & Leirós, 2008), and binocular rivalry (Handa et al., 2004; Handa, Shimizu, Uozato, Shoji, & Ishikawa, 2012). In one of the studies by Handa et al. (2004), the contrast of rivaling stimuli was manipulated to estimate the interocular difference in duration of exclusive visibility. A large variation in magnitude of interocular imbalance was observed between subjects. However, the sighting dominant eye (hole-in-the-card test) was found to consistently agree with the eye showing longer duration of visibility. In two studies that used nonrivaling stimuli (Li et al., 2010; Yang et al., 2010), 38%–39% of the subjects were classified as having strong eye dominance. Both studies found the correlation between sensory eye dominance and sighting dominance (hole-in-the-card) to be weak. On the other hand, one of the studies (Li et al., 2010) found a greater consistency with the sighting test in the group with strong eye dominance. In the study that used target detection to quantify eye dominance (Valle-Inclán et al., 2008), a stream of alphanumeric characters was rapidly (200 ms per character) and simultaneously presented to either eye. The subjects' task was to identify a target letter that was presented to only one of the eyes. The interocular difference in target detection probability was found to be normally distributed and typically agreed with sighting eye dominance (Miles ABC). From the data over target detection proportions in this study, it appears that approximately 35%–40% of the subjects showed almost complete LE or RE dominance, while the rest showed different degrees of dominance. 
In terms of using binocular sighting as a method to quantify eye dominance, there is even more limited recent research. One study (Purves & White, 1994) investigated monocular preferences in binocular viewing in a task where the subjects (n = 97) were to observe a target through a fenestrated screen. The screen allowed foveation of the target in only one eye at a time. Even though the purpose of the study was not to quantify eye dominance, it reported some interesting findings. The task was repeated 50 times for each subject, and the number of times either eye was used for foveation was recorded. Only four subjects exclusively used one of the eyes, while the majority of subjects instead alternated between the eyes to different extents. Approximately half of the subjects were classified as either moderately dominant or ambiocular, based on how often either eye was used for fixation. These results indicate a difference in behavior from the hole-in-the-card test, where subjects tend to consistently choose the same eye (Ehrenstein, Arnold-Schulz-Gahmen, & Jaschinski, 2005; Li et al., 2010; Rice, Leske, Smestad, & Holmes, 2008; Zeri, De Luca, Spinelli, & Zoccolotti, 2011). In the experiment by Purves and White (1994), the partially occluding fenestrated screen was positioned 2 m from the eyes of the observer. This differs from many traditional sighting tasks, where the occluding medium/object is positioned at maximum arm's length. The proximity of the nearer object in the hole-in-the-card test, and the prevention of visual feedback to the eye not used for alignment, may be factors explaining this difference in behavior. 
From the experiment by Purves and White (1994), and as suggested by other researchers (Kommerell, Schmitt, Kromeier, & Bach, 2003), it appears that when faced with a task that requires the subject to choose between the view of either eye, some subjects value one of the monocular views higher. On the other hand, a significant share of subjects seem to value the monocular views more or less equally. This makes the decision of which eye to use more difficult, and hence an alternating behavior occurs. The present study deals with the question of whether the weighting of the monocular views can be quantified during binocular sighting and by monitoring the binocular percept. This has in part been addressed in previous research. 
In one experiment, the hole-in-the-card test and point test were modified to prevent visual feedback (Barbeito, 1981). The majority of subjects then aimed from a reference point approximately midway between the eyes. This reference point coincided with the measured location of the egocenter as determined with a dichoptic alignment method. When visual feedback was allowed, the subjects' reference point shifted sideways to coincide with one eye. The reference points with and without visual feedback were found to consistently be located on the same side of the midline. It was therefore suggested that sighting dominance is a residual effect of the egocenter being slightly shifted towards either eye. 
In another study (Kommerell et al., 2003), sighting eye dominance (Parson's monoptoscope) was compared with ocular prevalence. Ocular prevalence refers to the subject's perceived alignment of stereoscopically presented near and far targets, imaged within Panum's area. The extension of a theoretical straight line drawn through the perceived position of the near and far target crosses the interocular bridge at a point that provides a measure of ocular prevalence. A crossing point located at the midpoint between the eyes indicates no prevalence (0%)—that is, both eyes contribute equally to the percept. The measure can range up to 100% LE or 100% RE. On average, the subjects showed an ocular prevalence of 24.1%–32.0%, with the prevalence exceeding 10% in 13 of 20 subjects. The laterality of ocular prevalence and sighting eye dominance agreed in 75% of the subjects. The experiment was extended by measuring the effect of a neutral density filter held in front of the prevalent eye. On average, the prevalence then shifted by 43 percentage points to the opposite side of the midline. It was suggested that ocular prevalence is the result of partial suppression of one eye, in order to help disregard double images at stereo disparities close to the borders of Panum's area. 
Two studies have used visual alignment tasks under true binocular viewing conditions in order to assess the weighting of monocular views (Charnwood, 1949; Francis & Harwood, 1951). It was demonstrated that, during this viewing condition, the subject aims along a visual axis that projects from a position located between the eyes (projection center), as opposed to monocular sighting tasks, where the projection center coincides with the visual axis of the fixating eye (Charnwood, 1949). As dioptric blur (Charnwood, 1949) or luminance reduction (Francis & Harwood, 1951) was induced to one eye, it was observed that the projection center shifted towards the nonfiltered eye. However, in relating these observations to sighting dominance, both studies found that the agreement between the projection center and sighting dominance was inconsistent. In one of the studies (Francis & Harwood, 1951), however, it was observed that the pattern for how the projection center changed was different depending on whether blur was induced to the ipsilateral or contralateral side of the projection center. In plotting the repositioning due to induced blur, the slope was greater when blur was induced to the dominant side. This slope difference was sufficiently repeatable over time and considered to be a demonstration that the effects of eye dominance exist under true binocular conditions. To our knowledge this observation has not been replicated. 
Another interesting study utilized physiological diplopia to determine whether the sensory input from either eye dominates the binocular percept (Bjork, 1980). This test, the variable-angle mirror test (VAMT), appears to belong in the sensory category of eye-dominance tests. However, what differentiates it from other sensory tests is that it does not apply dichoptic presentation or retinal rivalry, which are otherwise common (Evans, 2007). The test consists of two mirrors placed at a right angle so that subjects who look into the mirrors see a nonreversed image of their face. The border between the mirrors appears in physiological diplopia, and, depending on the interocular weighting of monocular views, one of the diplopic images will interfere more strongly with one eye's view. A good correlation between the VAMT and the sighting test was found, but another important finding was that a significant number of subjects did not show any eye preference. To our knowledge this test has not been further evaluated for the occurrence of different grades of eye dominance. 
The purpose of the current study was to evaluate whether different degrees of eye dominance can be identified under binocular viewing conditions using two principles: (a) how the subject positions during binocular sighting and (b) how the subject's binocular percept of a scene is affected by the subject's individual weighting of monocular views. The intention was to keep test conditions at natural binocular viewing conditions where the subject must deal with physiological diplopia. The first objective was to assess the subjects' baseline response—i.e., without any manipulation of visibility in either eye. Hypothetically, this result will give an indication of how the subject habitually values the views of either eye. The second objective was to complement the baseline result, by assessing if there was an interocular difference in tolerance to degraded visibility. The hypothesis was that if the baseline result is due to partial suppression of either eye, then this will show as different sensitivity to degraded visibility. Accordingly, this will affect the magnitude of change from baseline. The third objective was to assess how the results of binocular eye dominance in these two tasks agreed with a sighting dominance test (hole-in-the-card). 
Methods
The full procedure took place in an evenly lit room (mean illumination of 550 lux). The total participation time was approximately 1 hr at one occasion. After informed consent, all subjects went through an examination to ensure that they met the inclusion criteria. The order of the experiments was first the binocular sighting test (BST), followed by the VAMT, and finally the hole-in-the-card sighting test (HICT). The HICT was saved for last to reduce any effects of awareness of sighting dominance during the experiments. 
Subjects
A total of 32 healthy subjects (age 23.5 ± 2.8 years, 26 female and six male) were included (Table 1). Another nine subjects participated but were excluded from analysis due to not meeting the inclusion criteria. The inclusion criteria were monocular visual acuity at distance of 0.10 LogMAR or better (Visual Acuity Chart 2000 “ETDRS” 4 m; precision-vision.com), not more than one line of visual-acuity difference between the eyes, a stereo vision acuity of 60 arcsec or better (TNO random dot test), and no suppression at test distances (Worth four-dot test). The research adhered to the tenets of the Declaration of Helsinki and was approved by the regional ethics review board in Stockholm. 
Table 1
 
Subjects.
Table 1
 
Subjects.
Filters
When the retinal images in either eye are of different clarity, a regional suppression mechanism occurs, where blurred parts of the retinal image in one eye are suppressed by corresponding nonblurred image parts in the other eye (Schor, Landsman, & Erickson, 1987; Simpson, 1991). Fusion can, however, still occur to a certain degree (Larson & Lachance, 1983; Lovasik & Szymkiw, 1985) due to similar features in the stimuli, for instance certain spatial frequencies. Differences in monocular image quality have also been shown to influence the computation of binocular visual direction (Charnwood, 1949; Francis & Harwood, 1951; Mansfield & Legge, 1996), including for diplopic images (Rose & Blake, 1988). 
In the present study, a combination of Bangerter filters (Ryser Ophthalmologie, St. Gallen, Switzerland) and clear tape was used to degrade visibility. The intention was to induce a partial stepwise occlusion while maintaining binocular vision. Different options for inducing visual degradation were considered. Optical blur was declined due to the possibility of small residual hyperopic refractive errors neutralizing part of the optical blur. Also, the plus lens induces a myopic refractive state to the eye, which may affect the outcome of eye-dominance assessment (Charnwood, 1949). Neutral density filters are another viable option that, along with dioptric blur, has been shown as a method for inducing partial occlusion while preserving binocular vision (Li et al., 2012). It has, however, been shown that in order to achieve a reduction in visual acuity corresponding to Bangerter filters or optical defocus, a neutral density filter strength of 1.86 log units or more must be applied (Li et al., 2012). This corresponds to a light transmission of less than 2%. We considered this extensive reduction in light transmission a possible risk of inducing rivalry and blank-outs. This would make the task more difficult for the subjects and consequently cause the test to be less sensitive. 
The Bangerter filters consist of densely packed microelements that produce localized image distortions (Perez, Archer, & Artal, 2010). This has effects on visual acuity, vernier acuity, and contrast sensitivity (Odell, Leske, Hatt, Adams, & Holmes, 2008; Perez et al., 2010; Rutstein et al., 2011). In terms of effects on binocular function, the Bangerter filters have been shown to cause partial suppression up to the 0.4 strength, with a disruption of binocularity at the 0.2 strength (Li et al., 2012). 
During the initial pilot tests with Bangerter filters, it was noted that the specification and the actual degradation of visual acuity did not agree. It was particularly challenging to identify a filter that would produce the desired first step of a one-line reduction in acuity. Similar observations have been made in previous research involving subjects with normal vision (Li et al., 2012; Odell et al., 2008; Rutstein et al., 2011). Through further pilot testing, the filtering strengths were established by measuring the actual visual acuity on a logarithmic Visual Acuity Chart 2000 “ETDRS” 4 m (Table 2). To achieve the desired effect of filter 1 (F1), the Bangerter filter was replaced with clear tape (Scotch Classic tape, 3M, St. Paul, MN). Similar translucent tapes are in some cases used clinically to eliminate diplopia. The clear tape does not have as occlusive an effect, but its fine structure resembles the effect of the diffusing microelements in Bangerter filters. The median reduction in LogMAR visual acuity was 0.18 between filter 0 (F0) and F1, 0.13 between F1 and filter 2 (F2), and 0.08 between F2 and filter 3 (F3). 
Table 2
 
Visual acuity (LogMAR) with final established filters measured in six eyes.
Table 2
 
Visual acuity (LogMAR) with final established filters measured in six eyes.
The filters were mounted in segments (30 × 30 mm) on bars of 2-mm glass (filter bar). During the experiments, the filter bars were held approximately 15 mm in front of the subject's eye. The first segment (F0) on the filter bar was plane glass with no occlusion foil attached. The next three segments (F1–F3) were covered with filters with an increasing level of degrading effect. In the BST, the different filters were mounted on separate bars, and the test order of filtered eye and filter strength were counterbalanced. In the VAMT, the filters were mounted on one stick to allow successive increases of filtering without the subject having to alter head position. 
Binocular sighting test (BST)
The BST was derived from the experiments by Charnwood (1949) and Francis and Harwood (1951). The experimental setup resembles the traditional point or Porta test for determining eye dominance, where the subject points the finger to a given target. Since the finger will appear in diplopia, the subject needs to choose which diplopic image of the finger to align with the target. However, instead of pointing, the current task is to align a near and far object suspended from the ceiling and positioned farther away from the eyes of the subject (Figure 1). Any effects of manual habit (hand, arm) are thus minimized, and the greater distance from the subject and between objects reduces the disparity of the diplopic images. The subject is instructed to fixate the more distant bead and to position so that the bead is seen through the closer positioned ring. A pointer above the head, not seen by the subject, indicates the subject's head position (i.e., projection center) as the subject positions him- or herself to align the objects. This position is then recorded by taking a photograph of the face and pointer with a digital camera that is aligned with the other objects. The experimental procedure for each subject consisted of two conditions—the filter bar held in front of the left eye or the right eye—and four different filter steps. In summary, each subject did eight trials, positioning him- or herself to align the objects with four different filter grades, held in front of the left and the right eye. Each subject did the experimental procedure once. 
Figure 1
 
A camera (a), a bead (b) with diameter 15 mm, a ring (c) with diameter 15 mm, and a pointer (d) were aligned on a straight line (stimulus line). The distance between bead and ring (c − b) was 45 cm, and between ring and eye (d − c), 155 cm.
Figure 1
 
A camera (a), a bead (b) with diameter 15 mm, a ring (c) with diameter 15 mm, and a pointer (d) were aligned on a straight line (stimulus line). The distance between bead and ring (c − b) was 45 cm, and between ring and eye (d − c), 155 cm.
In the procedure, the subject looked down, put the filter in front of one eye, and was then instructed to look up with the filter in front of the eye and to fixate the bead through the opening of the nearer ring. A photograph was taken the moment the subject reported alignment of the bead and ring. The filter in front of the right eye was held with the right hand, and the left with the left. Before each change of filter, the subject looked down and repeated the procedure. 
A total of eight photographs were taken for each subject. In the postprocessing, a scale was inserted using photo-editing software (Inkscape 0.48.2, http://www.inkscape.org). The scale ranged from the center of the right pupil (0) to the center of the left pupil (100). The pointer's position along this scale indicated the projection center. A value less than 50 meant that the projection center was located towards the subject's right eye. A value greater than 50 meant that the projection center was located to the subject's left eye. A value equal to 50 meant that the projection center was midway between the eyes. The baseline response (baseline projection center, BPC) was based on the calculated mean of position when plane glass was held in front of the left and the right eye. 
The experiment was conducted with one dependent variable—the position of the head relative to the pointer—and two independent variables: eye (two levels, right or left) and filter (four levels). The measurements were repeated within subjects and counterbalanced for filter and eye. 
Variable-angle mirror test (VAMT)
This experiment is based on the VAMT (Bjork, 1980). In this test, two plane-glass mirrors, attached by a black hinge, are arranged at a right angle (Figure 2). The subject is positioned 50 cm away, centered in front of the hinge, and looks straight ahead at his or her face. As can be seen from the ray trace, the subject will be looking at a nonreversed image of his or her face in the mirror. Midway between the subject and the mirror image, the hinge will appear as two vertical lines due to physiological crossed diplopia. The right eye will see the left double image of the hinge, which will appear to be superimposed on the right eye in the mirror image, and vice versa. The theory behind the experiment is that the subject's percept of the physiological double images, hypothetically influenced by eye dominance, will decide which eye the image of the hinge will be perceived over. The stronger the eye dominance, the more likely the subject will suppress one double image, and the remaining image will appear to be superimposed on the dominant eye. In the case of less pronounced eye preference, the subject may see both double images, one covering each eye. In this case, depending on monocular weighting, one of the images of the hinge may appear darker than the other. 
Figure 2
 
The variable-angle mirror test. The subject is looking at a nonreversed image of his or her face, and the hinge (a) between the mirrors is seen in crossed physiological diplopia.
Figure 2
 
The variable-angle mirror test. The subject is looking at a nonreversed image of his or her face, and the hinge (a) between the mirrors is seen in crossed physiological diplopia.
In the procedure, the subject was positioned centered in front of the mirror and asked to make small adjustments of the mirror angle until the face could be seen in full. The experimenter then held the filter bar in front of the subject's left or right eye. For each subsequent increase of filter level, the subject was instructed to look at each eye in turn in the mirror image and report which eye the image of the hinge appeared to be in front of. It was explained that it may be perceived as a dark, slightly blurred vertical line. To avoid misunderstandings and unintended influence on the response, a piece of tape was stuck to the subject's cheeks: black tape on the right cheek and white tape on the left. Subsequently, the subject and experimenter referred to the black or the white side instead of the right or the left. In case two lines were perceived, the subject was asked to judge which line was darker. If the subject could not differentiate this, it was recorded as no eye preference. Each subject did the experimental procedure once. The experimental procedure consisted of two conditions—the filter bar held in front of the left eye or the right eye—and four different filter steps. In summary, each subject did eight trials. The responses when plane glass was held in front of the left and the right eye were recorded as the baseline response. 
The experiment was conducted with one dependent variable—the percept by the subject—and two independent variables: eye (two levels, right or left) and filter (four levels). The measurements were repeated within subjects and counterbalanced for order of filtered eye. The subject's response was evaluated according to a relative score. If the hinge was perceived to cover the right eye, it was scored 1; if the hinge was perceived to cover the left eye, it was scored −1. In the event the subject perceived two hinges of equal saliency, a score of 0 was recorded. 
Hole-in-the-card test (HICT)
In this test the subject held a card (20.0 × 12.8 cm) with both hands. In the center of the card was a 3-cm diameter hole. The subject was instructed to always look with both eyes and then, in one movement, raise the arms in order to visually align the hole in the card with a target at 2 m distance. The eyes were then covered one at a time and the eye that kept the alignment was recorded as the dominant eye. The procedure was repeated three times. 
Statistics
IBM SPSS Statistics version 22 was used for analysis. Quantitative data were analyzed with a two-way repeated-measures ANOVA and Pearson correlation analysis. Spearman correlation was used to analyze qualitative data. 
Results
Hole-in-the-card test (HICT)
The HICT was repeated three times and each subject gave consistent responses. Twenty-one subjects (65.6%) were RE dominant and 11 subjects (34.4%) were LE dominant (Table 1). 
Binocular sighting test (BST)
Baseline projection center (BPC)
For most of the subjects, the BPC was achieved following a two-step positioning. For a moment, the subject aimed from a point midway between the eyes. Shortly after this, a shift in head position occurred that positioned the projection center towards either eye. 
The first objective of the experiment was to assess the location of the BPC. The readings of the actual projection center with plane glass in front of the RE and the LE were averaged to give the BPC value. Based on this, 18 subjects (56.3%) had a BPC located on the right-hand side of the midline, 13 subjects (40.6%) had a BPC located on the left-hand side, and one subject (3.1%) had a BPC located on the midline (Figure 3). The correlation between BPC and HICT was weak (r = 0.18, not significant). One third of subjects with RE dominance according to the HICT, and 45.5% of subjects with LE dominance, showed the opposite behavior for BPC (Figure 4). 
Figure 3
 
Distribution of the subjects' baseline responses.
Figure 3
 
Distribution of the subjects' baseline responses.
Figure 4
 
BPC versus sighting dominance.
Figure 4
 
BPC versus sighting dominance.
Effects of induced blur
The second objective of the experiment was to assess the interocular difference in sensitivity to degraded visibility and its effect on BPC. First, the actual projection points were studied. An overview analysis of the subjects' positioning showed some characteristic response patterns. These patterns are illustrated by three sample subjects in Figure 5a through c
Figure 5
 
(a) Type I response. (b) Type II response. (c) Type III response. The x-axis shows filter steps starting with F0 (plane glass), and the y-axis shows the recorded projection center. The scale starts with 0 (center of right-eye pupil) and ends with 100 (center of left-eye pupil). The squares show the actual projection center when the filter is induced to the right eye, and the circles show the actual projection center when filtering is induced to the left eye.
Figure 5
 
(a) Type I response. (b) Type II response. (c) Type III response. The x-axis shows filter steps starting with F0 (plane glass), and the y-axis shows the recorded projection center. The scale starts with 0 (center of right-eye pupil) and ends with 100 (center of left-eye pupil). The squares show the actual projection center when the filter is induced to the right eye, and the circles show the actual projection center when filtering is induced to the left eye.
The first type of response pattern (type I) was found in 22 subjects (62.5%). These subjects showed a clear preference at baseline (F0), with the actual projection points located closer to one of the eyes (Figure 5a). The projection point remained more or less unaffected as blur was stepwise induced to the contralateral eye. On the other hand, the first step of blur induced to the ipsilateral eye of the projection point caused a major change. 
The second type of response pattern (type II) was found in four subjects (12.5%). In these subjects, the actual projection point immediately changed to the opposite side as step 1 of the filter bar (F0) was held in front of one eye (Figure 5b). That is, plane glass alone in front of either eye shifted the projection point towards the opposite eye. The altered projection point then stayed in this position as further blur was added. The same pattern occurred regardless of which eye was tested. 
The third type of response pattern (type III) was found in six subjects (18.8%). They initially positioned themselves so the actual projection point was on the same side as the eye in front of which the bar was held (Figure 5c). As blur was induced to one eye, the subjects made a stepwise shift of the projection point towards the opposite eye. 
A fourth pattern (not shown in the figure) was observed in two subjects (6.2%). The projection point in these subjects was located close to one particular eye, regardless of which eye the blur was induced to. 
In the next step of the analysis, the magnitude of blur-induced absolute change from BPC (LE and RE actual projection points at F0 averaged) was analyzed. A pattern emerged in which the BPC seemed to be a factor for predicting the magnitude of absolute change. The further the BPC was located towards either eye, the greater the change in projection center that occurred when inducing blur to the ipsilateral eye, and the lesser the change when inducing blur to the contralateral eye (Figure 6a through c). A two-way repeated-measures ANOVA was conducted with eye and filter as independent variables and the absolute change, in scale steps, as the dependent variable. A statistically significant interaction effect was found between the effects of eye and filter (df = 2), F = 6.69, p < 0.01. 
Figure 6
 
(a) Filter grade 1. (b) Filter grade 2. (c) Filter grade 3. Scatter plot by filter grade (F1−F3) of baseline projection center (x-axis) versus the absolute change (y-axis) from the baseline projection center. The plots in the upper panels refer to the change as blur is induced to the right eye, and the lower panels show the corresponding plot as blur is induced to the left eye.
Figure 6
 
(a) Filter grade 1. (b) Filter grade 2. (c) Filter grade 3. Scatter plot by filter grade (F1−F3) of baseline projection center (x-axis) versus the absolute change (y-axis) from the baseline projection center. The plots in the upper panels refer to the change as blur is induced to the right eye, and the lower panels show the corresponding plot as blur is induced to the left eye.
A Pearson correlation analysis of BPC versus absolute change in projection center showed a statistically significant correlation for all three filter grades—F1: r = −0.513, p < 0.01; F2: r = −0.538, p < 0.01; and F3: r = −0.535, p < 0.01. 
Variable-angle mirror test (VAMT)
Baseline response
When the first step of the filter bar (plane glass) was applied, some subjects responded differently for the LE and the RE. Some responded with no preference for one eye but an LE or RE preference for the other eye. The response was scored as −1 if it was LE preference at both trials, 1 if it was RE preference at both trials, and ±0.5 if it was no preference at one trial and righter or LE preference at the other. In the case of no preference at either trial, the score was set to 0. According to this classification, 18 subjects (56.3%) showed RE preference, five (15.6%) showed LE preference, and nine (28.1%) showed no preference. The correlation with the HICT was moderate (r = 0.527, p < 0.01). Twenty subjects (62.5%) showed an agreement with the HICT, nine (28.1%) showed no preference, and three (9.5%) showed opposite dominance (Figure 7). 
Figure 7
 
Variable-angle mirror test versus hole-in-the-card test.
Figure 7
 
Variable-angle mirror test versus hole-in-the-card test.
Effects of induced blur
The sample subject in Figure 8 shows an example of pronounced RE dominance. The subject's baseline response was RE dominance. Two steps of induced blur were required on the RE to cause a release of the partial suppression of the LE, meaning that the subject reported seeing two faint images of the hinge, one covering each eye. As one more step of blur was induced to the RE, partial suppression occurred to the RE, meaning that the subject perceived the hinge to cover the LE. The LE, on the other hand, was partially suppressed from the beginning. Thirteen subjects (40.6%) showed a similar RE preference pattern, and two (6.2%) showed the corresponding LE preference pattern. In subjects with less pronounced dominance (11 subjects, 34.4%; Figure 9), a slight relative difference in tendency to suppress blur could still be observed. Subjects with no preference at baseline (six subjects, 18.8%; Figure 10) showed a symmetric LE and RE pattern of blur suppression. 
Figure 8
 
Pronounced right-eye preference. The upper panel shows the result of inducing blur to the right eye, and the lower panel of inducing blur to the left eye. The value 1.0 on the y-axis means that the subject perceived that the right eye was covered by the hinge, while the value −1.0 means that hinge was perceived to cover the left eye. A value of 0 means the subject perceived two images of the hinge, one covering each eye.
Figure 8
 
Pronounced right-eye preference. The upper panel shows the result of inducing blur to the right eye, and the lower panel of inducing blur to the left eye. The value 1.0 on the y-axis means that the subject perceived that the right eye was covered by the hinge, while the value −1.0 means that hinge was perceived to cover the left eye. A value of 0 means the subject perceived two images of the hinge, one covering each eye.
Figure 9
 
Less pronounced eye preference with a slight weighting towards the left eye. A relative difference in sensitivity to blur can be observed where two increments of blur to the left eye versus one step to the right eye were required to transfer the percept.
Figure 9
 
Less pronounced eye preference with a slight weighting towards the left eye. A relative difference in sensitivity to blur can be observed where two increments of blur to the left eye versus one step to the right eye were required to transfer the percept.
Figure 10
 
No eye preference at baseline and a similar sensitivity to blur for each eye.
Figure 10
 
No eye preference at baseline and a similar sensitivity to blur for each eye.
In the next step of the analysis, the interocular difference in number of filter steps required to alter the baseline response was calculated. This was done by subtracting the number of steps required for the LE from the number of steps required for the RE. The result was 0 in the case of no difference. In the case of the RE requiring more steps of blur, the value was positive in the range of 1–3. In case of the LE requiring more steps, the value was negative in the same range. A Spearman correlation analysis of the baseline response versus the interocular difference in blur required to transfer suppression showed a significant correlation (r = 0.75, p < 0.01). 
Relation between tests
The correlation between BST and VAMT on total group level was weak (r = 0.16, not significant). A total of 22 subjects (68.8%) showed an eye preference in both tests. These 22 subjects were divided into a group of 14 subjects who showed an agreement between BST and VAMT and a group of eight who did not show an agreement. For the group of 14 there was a significant correlation between baseline responses of BST and VAMT (n = 14, r = −0.61, p = 0.02). There was also an almost complete agreement (92.9%) with the HICT. For the group of eight there was no significant correlation between baseline responses of BST and VAMT. The results of the HICT agreed better with the VAMT in this group (75.0%). 
A total of 10 subjects showed no eye preference in one or both of the BST and VAMT. One subject (10.0%) showed an agreement between the HICT and VAMT, and four subjects (40.0%) showed an agreement between the HICT and BST. 
Discussion
These experiments were done to explore eye dominance under true binocular conditions, where both eyes have the same preconditions to contribute. In the HICT, the subjects showed a consistent and typical distribution between RE and LE dominance (Ehrenstein et al., 2005; Rice et al., 2008; Shneor & Hochstein, 2006; Yang et al., 2010). On the other hand, in the binocular dominance test the distribution got less obvious, with up to 28% of the subjects showing no dominance in the VAMT. This was reflected in our observations during the experiments where all subjects gave clear-cut responses in the HICT, while their responses in many cases were less straightforward during the binocular tests. 
The experimental setups may be considered to have limitations, due to the inevitable physiological crossed diplopia that occurred. Our intention, however, was to keep the experimental setup simple and to see how the subjects responded to this naturally occurring feature. Another aspect is the use of clear tape to achieve the desired filtering effect of the lowest filter grade (F1). This may obviously affect the reproducibility of the study, due to availability of the product or to any variability in the occluding effect between brands. We did, however, carefully pilot-test before and are confident about the occluding effect in the current experiment. 
Binocular sighting test (BST)
The main finding of this experiment was that a majority of subjects positioned to aim from a reference point, the projection center, that coincided not with one eye but rather with a point somewhere between the eyes. Furthermore, this behavior was correlated to an interocular difference in sensitivity to degraded visibility. 
It was observed that the subjects initially positioned themselves to aim from a projection center approximately midway between the eyes. Then, shortly after that, they translated their positioning sideways to take a final position where the projection center was repositioned towards either eye. It appears that the initial positioning was based on directional judgement from the subjects' center of visual direction. However, in accordance with the observations in a previous study (Barbeito, 1981), the subjects soon made an adjustment due to visual feedback. Five subjects took a final position that made the line of sight of one eye coincide with the stimulus line, and thus appeared to disregard the input from one eye. A similar pattern has been found when performing the test with one eye occluded (unpublished observation). In the study of monocular preferences in binocular viewing (Purves & White, 1994), it was also found that a minority of subjects exclusively used one of the eyes. 
The remaining subjects did, however, aim from a point somewhere between the eyes, as was also observed in the original experiments (Charnwood, 1949; Francis & Harwood, 1951). It may be considered that this behavior is due to a change of position of the subjects' center of visual direction, or cyclopean eye, from which they judge direction relative to themselves. Previous research has, however, emphasized that the cyclopean eye has a fixed position approximately midway between the eyes (Banks, Van Ee, & Backus, 1997; Mapp & Ono, 1999). It is also important to make a distinction about the experimental task in the current study. The task was about relative direction—that is, the subjects were to position themselves to make the objects appear aligned. There is evidence that the position of cyclopean eye is connected with absolute direction and thus not related to results of relative alignment tasks (Mapp & Ono, 1999). It can, however, not be excluded that the cyclopean eye may have had a part in the alignment behavior for some subjects. In the event that the cyclopean eye had a noncentral position, it may have affected the decision on how to align the objects. This may then also explain the two-step positioning also observed previously (Barbeito, 1981). A continuous video recording of positioning would have been required to confirm this theory and clarify if there was a relation between the initial and final positioning. 
This experiment provided a challenge to the subjects due to the physiological diplopia. However, only a few subjects made a remark about diplopia, and they were in those cases instructed to use the most salient image. The mechanisms behind the weighted aiming behavior in the majority of subjects largely remain to be understood, but a possible explanation may be found in the perceived direction of the objects. It is generally viewed that a fused stimulus is perceived in a visual direction resulting from an averaging process of the oculocentric visual direction seen from either eye (Sheedy & Fry, 1979). However, this process may be modified through weighting of the monocular directional signal, due to differences in contrast (Banks et al., 1997; Mansfield & Legge, 1996), luminance, image clarity, or forms of eye dominance (Charnwood, 1949; Francis & Harwood, 1951). The experiments by Rose and Blake (1988) found indications that a similar effect occurs to physiological diplopic images. It was suggested that, due to a perceptual mechanism that seeks to match and assign similar directions, the diplopic images are perceived to be closer than they are based on a purely retinotopic local sign. 
A further important finding in the current experiment was a poor overall agreement between sighting dominance and the BPC. This is contrary to the findings in a previous study (Barbeito, 1981), where sighting dominance and the position of the egocenter, or cyclopean eye, mostly agreed. Based on this, it was suggested that sighting dominance was a residual effect of the egocenter being located to one side of the midline. When forced to choose an eye, the subjects shifted towards the eye that required the least adjustment. Another study (Kommerell et al., 2003) that explored the concept of ocular prevalence found that the sighting dominant eye and laterality of ocular prevalence agreed in 75% of subjects. Even though these studies refer to effects of different concepts, egocenter and ocular prevalence, the principles of their test methods appear to resemble each other. Both methods appear to measure relative visual direction under stereoscopic or dichoptic conditions. This may suggest different mechanisms behind the behavior in a real scene, as in the current experiment, versus a stereoscopically presented scene. 
As blur was induced to one eye at a time, different patterns could be observed. The stronger the weighting of the BPC towards one eye, the more monocular blur could be induced to the contralateral eye without affecting positioning. The finding of a significant correlation between BPC and positioning when blur was induced may suggest that a form of dominance affected the behavior. Francis and Harwood (1951), who measured the effect of neutral density filters in front of either eye, also found an approximate proportional relationship between reduced sensation in either eye and change of projection center. Another important finding in that study was that of interocular differences in the slope of the change in projection center. It was suggested that this may be a sign of eye-dominance effects. The present study confirmed this finding. Furthermore, the results appear to support the hypothesis that the location of the projection center during binocular alignment is the result of partial suppression in one eye. 
Some experimental limitations need to be considered. The bead subtended 26 arcmin at 2 m distance, and the inner ring subtended 22 arcmin at 1.55 m distance. This difference, along with effects of defocus blur on the nearer ring, means that there was an allowance for alignment error. Another limitation is uncertainty in the measurement. Care was taken to photograph the positioning immediately after the subject reported aligning the objects. However, it cannot be excluded that small movements by the subject may have affected the measurement. In terms of the filters, one may suspect a ceiling effect, since several subjects made a major repositioning already at the first filtering level, followed by only minor adjustments. For the purpose of a more sensitive grading of eye dominance, even more subtle increments of blur may be necessary. 
Variable-angle mirror test (VAMT)
The main finding of this experiment was the occurrence of a plausible form of eye dominance. It showed both in the baseline response and in a correlation to interocular differences in blur suppression. Several subjects reported initially finding it difficult to notice the image(s) of the hinge. In some cases it was perceived to be opaque, and thus occluding one eye, but in many cases it was perceived as a semitransparent, grayish line through which the eye could still be seen. 
At the baseline response, about two thirds of the subjects showed an obvious preference. The line appeared more salient over one particular eye, regardless of which eye was manipulated. For another 12 subjects the percept of the scene appeared less evident. They responded differently depending on which eye the plane glass was held in front of. In these subjects, no preference was noted when one particular eye was manipulated—that is, two equally salient lines were perceived. When the other eye was manipulated, the subjects responded that the most salient line was seen by the nonmanipulated eye. These observations, we think, suggest the presence of subtle interocular differences in sensitivity to manipulation. 
A moderate but statistically significant correlation was found between the HICT and the VAMT. Just like in the original experiment (Bjork, 1980), we observed that quite a few subjects (28%) did not show a preferred eye. This observation in the original experiment made the author question the differentiating capacity of the method. However, it may in fact reflect the nature of eye dominance in subjects with normal sight: Eye preference in many subjects may be a transient concept that reflects the flexibility of the visual system to adapt to visual conditions. 
As blur was induced, we observed different response patterns. A correlation was found between the baseline response and the interocular difference in amount of blur required to transfer the percept of the most salient line from one eye to the other. So similar to the BST, there appeared to be a relation between weighting of monocular input and interocular blur suppression. Apart from providing an idea of the degree of eye dominance, it helped to further differentiate subtle tendencies for eye dominance. At baseline, 28% of the subjects did not show a preference; however, when the interocular difference in blur suppression was measured, differences emerged, and the proportion of subjects showing no preference was reduced to 19%. 
Some subjects reported that the image of the hinge was not covering the eye but was located between the nose and the eye. The mechanisms behind this remain to be understood but could in part be explained by the apparent displacement that may occur to diplopic nonfused objects (Rose & Blake, 1988). There may also be an aspect related to eye position. The subjects were initially instructed to make small adjustments of the mirror angle until they could readily see the image of their face clear and single. In the event that a slight residual mirror adjustment remained, which was compensated for by sensory fusion, the computation of visual direction may have been affected by different monocular oculocentric (secondary) direction. 
Some limitations of the experiment need to be considered. As observed, the test was quite challenging to some subjects, and it is therefore important to consider that neutral responses are to be expected. Secondly, the result of the test relies solely on the verbal response from the subject. Without the possibility for the experimenter to verify the response, there may be some uncertainty in the measures. 
Relation between tests
BST versus VAMT
The correlation between the BST and the VAMT on the total-group level was weak. Interestingly though, a group of 14 subjects (44%) showed complete agreement between the tests. Furthermore, there was almost complete agreement with the HICT for this group. It appears that these subjects had pronounced eye dominance that remained despite the differences in test conditions. While we acknowledge the known challenges of comparing results between different tests, it still appears that this share of subjects fits within the range (35%–50%) that have been categorized as having a strong eye preference in previous research (Handa et al., 2005; Li et al., 2010; Purves & White, 1994; Valle-Inclán et al., 2008; Yang et al., 2010). 
BST versus HICT
The poorest agreement was found between the HICT and the BST. This was somewhat surprising, since one may expect these tests to be more closely related. However, Purves and White (1994) also found only a moderate correlation between the Miles ABC test and monocular preference under binocular viewing. They also observed that only a handful subjects exclusively selected one eye in the task, while the vast majority used both eyes to a lesser or greater extent. 
The challenge in the BST of dealing with physiological diplopia may have led to a certain influence of chance on the results. Most subjects did not report seeing a double image, but a decision still had to be made about which image to use for alignment. If the experiment had been repeated several times, to monitor the consistency of the responses, it might have helped to further differentiate patterns in behaviors related to dominance. On the other hand, many subjects (69%, type I) were quite consistent in their response, suggesting the occurrence of a form of dominance consistent with blur suppression but less consistent with monocular sighting (HICT). 
VAMT versus HICT
The VAMT showed a moderate agreement with the HICT, which in a way supports the idea of the sighting test as a sensory test (Evans, 2007). On the other hand, previous research using sensory tests implies that tests utilizing rivaling stimuli (Handa et al., 2004; Valle-Inclán et al., 2008) agree more strongly with sighting tests. The VAMT presents not a rivaling stimulus but a rather strong stimulus, the image of the face. As noted in the original experiment (Bjork, 1980), the face compels stronger attention than other objects might. This may explain the difficulty experienced by many subjects in initially noticing the images of the hinge. If this is the case, this test may be more effective in identifying subjects with pronounced eye dominance and thus explain the stronger agreement with the HICT as compared to the BST. A possible clinical application for the VAMT may then be to identify habitual fixation behavior in subjects with pronounced eye dominance due to binocular imbalance. 
Conclusions
To conclude, we think these experiments confirm that forms of graded eye dominance occur under binocular viewing conditions. The observation that many subjects did not show pronounced eye dominance during these experimental conditions agrees with previous research. The effects of eye dominance appear to become less obvious when binocular visual feedback is allowed and effects of manual habit are minimized. This suggests that in subjects with normal sight, eye dominance may be a transient feature depending on visual conditions. 
The low to moderate consistency with a traditional sighting test like the HICT further emphasizes what previous research also implies: Different tests will give different answers. It may therefore be necessary to consider the application before choosing a method to assess eye dominance. The HICT still appears to provide a reasonable guideline, particularly in pronounced eye dominance, but an awareness of its limitations may be necessary. Further research will be required to better understand the effects of eye dominance in different applications, e.g., by comparing the outcome of different eye-dominance tests with functional measures of visual performance. 
Acknowledgments
The authors wish to thank Hanna Jonsson and Åsa Söderberg for their excellent contribution in data collection and input to the analysis. 
Commercial relationships: none. 
Corresponding author: Jan Johansson. 
Email: jan.johansson.1@ki.se. 
Address: Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden. 
References
Banks, M. S., Van Ee R., Backus B. T. (1997). The computation of binocular visual direction: A re-examination of Mansfield and Legge (1996). Vision Research, 37 (12), 1605–1613.
Barbeito R. (1981). Sighting dominance: An explanation based on the processing of visual direction in tests of sighting dominance. Vision Research, 21 (6), 855–860.
Bjork A. (1980). The variable angle-mirror, a new tool for the study of ocular dominance and eye fixation. Acta Ophthalmologica (Copenhagen), 58 (2), 202–209.
Charnwood L. (1949). Observations on ocular dominance. Optician, 118, 85–96.
Ehrenstein W. H., Arnold-Schulz-Gahmen B. E., Jaschinski W. (2005). Eye preference within the context of binocular functions. Graefes Archive for Clinical and Experimental Ophthalmology, 243 (9), 926–932, doi:10.1007/s00417-005-1128-7.
Evans B. J. (2007). Monovision: A review. Ophthalmic & Physiological Optics, 27 (5), 417–439, doi:10.1111/j.1475-1313.2007.00488.x.
Francis J. L., Harwood K. A. (1951). The variation of the projection centre with differential stimulus and its relation to ocular dominance. Paper presented at the Transactions of the International Optical Congress, British Optical Association, London.
Handa T., Mukuno K., Uozato H., Niida T., Shoji N., Shimizu K. (2004). Effects of dominant and nondominant eyes in binocular rivalry. Optometry and Vision Science , 81 (5), 377–383.
Handa T., Shimizu K., Mukuno K., Kawamorita T., Uozato H. (2005). Effects of ocular dominance on binocular summation after monocular reading adds. Journal of Cataract and Refractive Surgery, 31 (8), 1588–1592, doi:10.1016/j.jcrs.2005.01.015.
Handa T., Shimizu K., Uozato H., Shoji N., Ishikawa H. (2012). A new method for quantifying ocular dominance using the balancing technique. The American Orthoptic Journal, 62, 77–86, doi:10.3368/aoj.62.1.77.
Johansson J., Pansell T., Ygge J., Seimyr G. O. (2014). Monocular and binocular reading performance in subjects with normal binocular vision. Clinical & Experimental Optometry, 97 (4), 341–348, doi:10.1111/cxo.12137.
Kommerell G., Schmitt C., Kromeier M., Bach M. (2003). Ocular prevalence versus ocular dominance. Vision Research, 43 (12), 1397–1403.
Larson W. L., Lachance A. (1983). Stereoscopic acuity with induced refractive errors. American Journal of Optometry and Physiological Optics , 60 (6), 509–513.
Li J., Lam C. S., Yu M., Hess R. F., Chan L. Y., Maehara G., Thompson B. (2010). Quantifying sensory eye dominance in the normal visual system: a new technique and insights into variation across traditional tests. Investigative Ophthalmology & Visual Science , 51 (12), 6875–6881. [PubMed] [Article]
Li J., Thompson B., Ding Z., Chan L. Y., Chen X., Yu M., Hess R. F. (2012). Does partial occlusion promote normal binocular function? Investigative Ophthalmology & Visual Science , 53 (11), 6818–6827. [PubMed] [Article]
Lovasik J. V., Szymkiw M. (1985). Effects of aniseikonia, anisometropia, accommodation, retinal illuminance, and pupil size on stereopsis. Investigative Ophthalmology & Visual Science, 26 (5), 741–750. [PubMed] [Article]
Mansfield J. S., Legge G. E. (1996). The binocular computation of visual direction. Vision Research, 36 (1), 27–41.
Mapp A. P., Ono H. (1999). Wondering about the wandering cyclopean eye. Vision Research, 39 (14), 2381–2386.
Nitta M., Shimizu K., Niida T. (2007). The influence of ocular dominance on monovision--the influence of strength of ocular dominance on visual functions. Nihon Ganka Gakkai Zasshi, 111 (6), 441–446.
Odell N. V., Leske D. A., Hatt S. R., Adams W. E., Holmes J. M. (2008). The effect of Bangerter filters on optotype acuity, vernier acuity, and contrast sensitivity. Journal of AAPOS, 12 (6), 555–559, doi:10.1016/j.jaapos.2008.04.012.
Oishi A., Tobimatsu S., Arakawa K., Taniwaki T., Kira J. (2005). Ocular dominancy in conjugate eye movements at reading distance. Neuroscience Research, 52 (3), 263–268, doi:10.1016/j.neures.2005.03.013.
Perez G. M., Archer S. M., Artal P. (2010). Optical characterization of Bangerter foils. Investigative Ophthalmology & Visual Science , 51 (1), 609–613. [PubMed] [Article]
Porac C., Coren S. (1979). Monocular asymmetries in recognition after an eye movement: Sighting dominance and dextrality. Perception & Psychophysics, 25 (1), 55–59.
Purves D., White L. E. (1994). Monocular preferences in binocular viewing. Proceedings of the National Academy of Sciences, USA, 91 (18), 8339–8342.
Rice M. L., Leske D. A., Smestad C. E., Holmes J. M. (2008). Results of ocular dominance testing depend on assessment method. Journal of AAPOS, 12 (4), 365–369, doi:10.1016/j.jaapos.2008.01.017.
Rose D., Blake R. (1988). Mislocalization of diplopic images. Journal of the Optical Society of America A, 5 (9), 1512–1521.
Rutstein R. P., Foster N. C., Cotter S. A., Kraker R. T., Lee D. H., Melia M., . (2011). Visual acuity through Bangerter filters in nonamblyopic eyes. Journal of AAPOS, 15 (2), 131–134, doi:10.1016/j.jaapos.2010.11.015.
Schor C., Landsman L., Erickson P. (1987). Ocular dominance and the interocular suppression of blur in monovision. American Journal of Optometry and Physiological Optics, 64 (10), 723–730.
Sheedy J. E., Bailey I. L., Buri M., Bass E. (1986). Binocular vs. monocular task performance. American Journal of Optometry and Physiological Optics , 63 (10), 839–846.
Sheedy J. E., Fry G. A. (1979). The perceived direction of the binocular image. Vision Research, 19 (2), 201–211.
Shneor E., Hochstein S. (2006). Eye dominance effects in feature search. Vision Research, 46 (25), 4258–4269, doi:10.1016/j.visres.2006.08.006.
Simpson T. (1991). The suppression effect of simulated anisometropia. Ophthalmic & Physiological Optics , 11 (4), 350–358.
Spache G. (1944). One-eyed and two-eyed reading. The Journal of Educational Research, 37 (8), 616–618.
Valle-Inclán F., Blanco M. J., Soto D., Leirós L. (2008). A new method to assess eye dominance. Psicológica, 29, 55–64.
Vergilino-Perez D., Fayel A., Lemoine C., Senot P., Vergne J., Dore-Mazars K. (2012). Are there any left-right asymmetries in saccade parameters? Examination of latency, gain, and peak velocity. Investigative Ophthalmology & Visual Science , 53 (7), 3340–3348. [PubMed] [Article]
Yang E., Blake R., McDonald J. E.,2nd. (2010). A new interocular suppression technique for measuring sensory eye dominance. Investigative Ophthalmology & Visual Science, 51 (1), 588–593, doi:10.1167/iovs.08-3076. [PubMed] [Article]
Zeri F., De Luca M., Spinelli D., Zoccolotti P. (2011). Ocular dominance stability and reading skill: A controversial relationship. Optometry and Vision Science, 88 (11), 1353–1362, doi:10.1097/OPX.0b013e318229635a.
Figure 1
 
A camera (a), a bead (b) with diameter 15 mm, a ring (c) with diameter 15 mm, and a pointer (d) were aligned on a straight line (stimulus line). The distance between bead and ring (c − b) was 45 cm, and between ring and eye (d − c), 155 cm.
Figure 1
 
A camera (a), a bead (b) with diameter 15 mm, a ring (c) with diameter 15 mm, and a pointer (d) were aligned on a straight line (stimulus line). The distance between bead and ring (c − b) was 45 cm, and between ring and eye (d − c), 155 cm.
Figure 2
 
The variable-angle mirror test. The subject is looking at a nonreversed image of his or her face, and the hinge (a) between the mirrors is seen in crossed physiological diplopia.
Figure 2
 
The variable-angle mirror test. The subject is looking at a nonreversed image of his or her face, and the hinge (a) between the mirrors is seen in crossed physiological diplopia.
Figure 3
 
Distribution of the subjects' baseline responses.
Figure 3
 
Distribution of the subjects' baseline responses.
Figure 4
 
BPC versus sighting dominance.
Figure 4
 
BPC versus sighting dominance.
Figure 5
 
(a) Type I response. (b) Type II response. (c) Type III response. The x-axis shows filter steps starting with F0 (plane glass), and the y-axis shows the recorded projection center. The scale starts with 0 (center of right-eye pupil) and ends with 100 (center of left-eye pupil). The squares show the actual projection center when the filter is induced to the right eye, and the circles show the actual projection center when filtering is induced to the left eye.
Figure 5
 
(a) Type I response. (b) Type II response. (c) Type III response. The x-axis shows filter steps starting with F0 (plane glass), and the y-axis shows the recorded projection center. The scale starts with 0 (center of right-eye pupil) and ends with 100 (center of left-eye pupil). The squares show the actual projection center when the filter is induced to the right eye, and the circles show the actual projection center when filtering is induced to the left eye.
Figure 6
 
(a) Filter grade 1. (b) Filter grade 2. (c) Filter grade 3. Scatter plot by filter grade (F1−F3) of baseline projection center (x-axis) versus the absolute change (y-axis) from the baseline projection center. The plots in the upper panels refer to the change as blur is induced to the right eye, and the lower panels show the corresponding plot as blur is induced to the left eye.
Figure 6
 
(a) Filter grade 1. (b) Filter grade 2. (c) Filter grade 3. Scatter plot by filter grade (F1−F3) of baseline projection center (x-axis) versus the absolute change (y-axis) from the baseline projection center. The plots in the upper panels refer to the change as blur is induced to the right eye, and the lower panels show the corresponding plot as blur is induced to the left eye.
Figure 7
 
Variable-angle mirror test versus hole-in-the-card test.
Figure 7
 
Variable-angle mirror test versus hole-in-the-card test.
Figure 8
 
Pronounced right-eye preference. The upper panel shows the result of inducing blur to the right eye, and the lower panel of inducing blur to the left eye. The value 1.0 on the y-axis means that the subject perceived that the right eye was covered by the hinge, while the value −1.0 means that hinge was perceived to cover the left eye. A value of 0 means the subject perceived two images of the hinge, one covering each eye.
Figure 8
 
Pronounced right-eye preference. The upper panel shows the result of inducing blur to the right eye, and the lower panel of inducing blur to the left eye. The value 1.0 on the y-axis means that the subject perceived that the right eye was covered by the hinge, while the value −1.0 means that hinge was perceived to cover the left eye. A value of 0 means the subject perceived two images of the hinge, one covering each eye.
Figure 9
 
Less pronounced eye preference with a slight weighting towards the left eye. A relative difference in sensitivity to blur can be observed where two increments of blur to the left eye versus one step to the right eye were required to transfer the percept.
Figure 9
 
Less pronounced eye preference with a slight weighting towards the left eye. A relative difference in sensitivity to blur can be observed where two increments of blur to the left eye versus one step to the right eye were required to transfer the percept.
Figure 10
 
No eye preference at baseline and a similar sensitivity to blur for each eye.
Figure 10
 
No eye preference at baseline and a similar sensitivity to blur for each eye.
Table 1
 
Subjects.
Table 1
 
Subjects.
Table 2
 
Visual acuity (LogMAR) with final established filters measured in six eyes.
Table 2
 
Visual acuity (LogMAR) with final established filters measured in six eyes.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×