Recognition in a visuospatial memory task: The effect of presentation

The influence of the mode of presentation (simultaneous vs. sequential) on accuracy and latency of visuospatial recognition was explored in three experiments. In Experiment 1, two groups were administered a visuospatial task in which a number of positions were shown either simultaneously or sequentially (in a random order); memory was tested using a recognition procedure of visuospatial patterns, either identical or different (a single cell displaced). The results showed that (1) performance was higher in the simultaneous than in the sequential presentation, and (2) decision time increased with complexity in the sequential presentation but not in the simultaneous presentation. In Experiment 2, the same task was used in three conditions of presentation, simultaneous, random sequential, and ordered sequential; at test, a single location, rather than a pattern, was presented for recognition. The results showed that (1) performance was higher in the simultaneous and in the ordered sequential presentations than in the random sequential one, and (2) decision time increased with complexity. In Experiment 3, the same task was used in the same three conditions of presentation, simultaneous, random sequential, and ordered sequential; at test either an identical or a “displaced” pattern was presented for recognition. The results showed that (1) performance was equivalent in the three types of presentation, and (2) decision time increased with complexity for “hit” items; different patterns of linear relations were observed for “correct rejections” items. The results are interpreted in terms of the organisation of visuospatial working memory, and three types of encoding—extrafigural spatial encoding, visual pattern encoding, and spatial path encoding—were proposed.     

Visual comparison of rotated and reflected random-dot patterns as a function of their positional symmetry and separation in the field

Subjects viewed pairs of random-dot patterns which were presented in a number of arrangements varying in the transformations applied to the patterns, in the distance between the patterns, and in the symmetry of the pattern positions with respect to the point of fixation. The task was to judge whether the patterns were “the same” taking into account possible rotations or reflections, or “different”. It was found that correct judgements for identical patterns were most affected by the distance between the two patterns, whereas correct judgements for patterns where one had been rotated through 180° or reflected were most affected by the symmetry of the pattern positions. A scheme modelling the visual recognition of transformed patterns, sufficient to explain the results, is presented.     

Visual search for positional relationships between pattern elements

The purpose of these experiments was to illustrate how a search for positional relationships between the elements in a stimulus pattern is effortful and serial, whereas a search for pattern elements (without requiring any positional information) takes place rapidly and in parallel. In all three experiments of this study, the target and distractor patterns consisted of two vertical line segments. In the experiment concerning the search for positional relationships between pattern elements, one of the line segments in a line pair had a gap, and the observer's task was to indicate whether the positional order of the two lines was identical in all the line pairs of the display, or whether the line segments were in a mirror-image order in one line pair (the target). In the two experiments concerning the search for pattern elements (no positional information was required), the observer's task was to look for a line pair with one gap line (the target) among line pairs containing two broken lines or among line pairs with unbroken lines. In all three tasks, the reaction time for a correct target detection was measured. The results showed that the search in the first task was highly serial, and in the second and third tasks of "feature" searches, the search time was nearly independent of the number of distractor pairs. It is suggested that this dissociation may be interpreted in the context of the quality of information processing in extrafoveal vision, i. e. the elements of a stimulus pattern can be clearly visible outside the fovea, but it is not possible to perceive accurately the positional relationships between them.     

Discrepancy and nondiscrepancy methods of assessing visual-auditory interaction

Much of our information about interaction among the sensory modalities has come from studies in which discrepancy between the modalities has been imposed experimentally. The question arises whether the conclusions generated from discrepancy research are affected by the use of the nonnatural experimental situation: Would the same conclusions be reached if discrepancy were not used? In addition to using the index of interaction derived from discrepancy, as in prior research, the present study assessed two nonexperimentally imposed indices to explore the interaction between vision and audition in a spatial localization task. The additional indices, or “tags,” were standard deviations (SDs) and confidence ratings (CRs) of localization responses. The SD tag showed a pattern of interaction between vision and audition that was identical to that found with spatial discrepancy. The CR tag produced a similar pattern but did not show the same regular variation with the independent variable. Both SD and CR tags showed identical patterns of results when an experimentally imposed discrepancy was and was not used, suggesting that the use of moderate degrees of experimental discrepancy does not produce misleading conclusions about intersensory interaction.     

The differentiation of rhythmic structure

Listeners judged whether two five-tone nonmetric rhythms were the same or different. Each rhythm was presented one, two, or four times to study the process of perceptual differentiation. The results indicated that the listeners perceived these rhythms in-terms.of the grouping of the tones, and not in terms of the timing between the groups. Two rhythms that had the same perceptual grouping were judged as being identical, even if the timing between the groups was different. The perception of the groupings of tones developed gradually. If each rhythm was presented only once, then the listeners had only a global percept, focused on groups (runs) of three elements, and often judged two different rhythms as being identical. If the rhythms were presented two or four times, then the grouping of the tones became more differentiated and the listeners were less likely to judge different patterns as being identical. Thus, perception of auditory rhythmic structure appears to follow the same developmental process as the perception of visual spatial structure.     

The differentiation of rhythmic structure.

Listeners judged whether two five-tone nonmetric rhythms were the same or different. Each rhythm was presented one, two, or four times to study the process of perceptual differentiation. The results indicated that the listeners perceived these rhythms in terms of the grouping of the tones, and not in terms of the timing between the groups. Two rhythms that had the same perceptual grouping were judged as being identical, even if the timing between the groups was different. The perception of the groupings of tones developed gradually. If each rhythm was presented only once, then the listeners had only a global percept, focused on groups (runs) of three elements, and often judged two different rhythms as being identical. If the rhythms were presented two or four times, then the grouping of the tones became more differentiated and the listeners were less likely to judge different patterns as being identical. Thus, perception of auditory rhythmic structure appears to follow the same developmental process as the perception of visual spatial structure.     

On the perception of symmetrical and repeated patterns

Four experiments investigated rapid perceptual judgments about tachistoscopically presented patterns that were either symmetrical about or repeated across a vertical axis. The same patterns were presented under two different instructional conditions; some Ss were to judge the two halves of each pattern “same” or “mirror,” others were to judge each pattern as a whole “symmetrical” or “asymmetrical.” With dot patterns, RTs were faster for symmetrical than for repeated patterns when the two halves were close together, but not when they were separated, regardless of instructions. With simpler patterns made up of arrowheads and C-shapes, however, “same” RTs were faster than “mirror,” but “asymmetrical” RTs were marginally slower than “symmetrical,” regardless of spatial separation. The advantage of “same” over “mirror” did not seem to be simply a labeling effect. The results suggest that left-right symmetry is perceptually more salient than left-right repetition when the patterns are perceived holistically. By contrast, distinct patterns can be matched more rapidly when they are the same than when they are left-right mirror images.     

Binocular vision with two stabilized retinal images

Experiments are described in which each eye is presented with a target whose image remains on the same part of the retina when the eye moves. The patterns presented to each eye may be similar and may be placed on corresponding parts of the retina or may be placed in non-corresponding positions: alternatively, different targets may be presented to the two eyes. Each pattern fades intermittently. Sometimes both are seen together and sometimes both fields are dark at once. There is a small negative correlation between the times of clear vision with the two eyes. When corresponding areas of the two retinas are illuminated with red and green light respectively, the composite colour (yellow) is never perceived with steady illumination. When two similar patterns are in nearly corresponding positions there may be subjective fusion. With two different targets there is sometimes a subjective impression that the two patterns move with respect to one another even though their positions on the retina are fixed.     

Spatial modulation of repetition blindness and repetition deafness

When two identical visual items are presented in rapid succession, people often fail to report the second instance when trying to recall both (e.g., Kanwisher, 1987). We investigated whether this temporal processing deficit is modulated by the spatial separation between the repeated stimuli within both audition and vision. In Experiment 1, lists of one to three digits were rapidly presented from loudspeaker cones arranged in a semicircle around the participant. Recall accuracy was lower when repeated digits were presented from different positions rather than from the same position, as compared to unrepeated control pairs, demonstrating that auditory repetition deafness (RD) is modulated by the spatial displacement between repeated items. A similar spatial modulation of visual repetition blindness (RB) was reported when pairs of masked letters were presented visually from either the same or different positions arranged on a semicircle around fixation (Experiment 2). These results cannot easily be accounted for by the token individuation hypothesis of RB (Kanwisher, 1987; Park & Kanwisher, 1994) and instead support a recognition failure account (Hochhaus & Johnston, 1996; Luo & Caramazza, 1995, 1996).     

Investigating the anticipatory nature of pattern perception in sport

The aim of the present study was to examine the anticipatory nature of pattern perception in sport by using static and moving basketball patterns across three different display types. Participants of differing skill levels were included in order to determine whether the effects would be moderated by the knowledge and experience of the observer in the same manner reported previously for simple images. The results from a pattern recognition task showed that both expert and recreational participants were more likely to anticipate the next likely state of a pattern when it was presented as a moving video, but only the experts appeared to have the depth of understanding required to elicit the same anticipatory encoding for patterns presented as schematic images. The results extend those reported in previous research and provide further evidence of an anticipatory encoding in pattern perception for images containing complex, interrelated patterns.     

The relationship perceived between the real body and the mirror image

We analyse here people's perception of their reflections in mirrors placed in different positions. In two experiments, participants looked at their mirror image, in a third experiment they looked at another person's image. In both cases they were asked to answer a series of questions about how the virtual body appeared relative to the real body, focusing on different aspects. In experiment 1, they were asked to decide whether the reflections were identical, similar, different, or opposite in terms of the global relationship, orientation, and lateralisation (left right arm). In experiment 2 they were instructed to make simple gestures and to evaluate if the gestures in the reflection were identical, opposite, similar, or different from theirs. Results show that 'identity' was preferred when the mirror was in front, and 'opposition' was preferred when the mirror was below. When opposition was experienced, it was attributed mainly to the exocentric frame of reference. Egocentric left-right reversal was not a common experience, although it was reported more frequently when the mirror was in front. The different roles of the exocentric and egocentric frames of reference were further tested in experiment 3, in which the condition of an observer looking at another person's reflection was studied. Contrary to the emphasis on the egocentric frame of reference in the literature on the 'mirror question', results presented in this paper demonstrate the importance of the exocentric frame of reference in influencing how observers react to their reflections.     

Effects of pattern orientation and number of symmetry axes on the detection of mirror symmetry in dot and solid patterns

It has been postulated that as the number of axes of symmetry in a pattern increases, so pattern 'goodness' increases. Recently, a distinction was made between two different theoretical accounts of regularity or 'goodness' in relation to patterns with mirror symmetry: the 'transformational' and the 'holographic' models. It was argued that the former predicts a 'goodness' ordering of four > three > two > one whereas the latter predicts four > two > three > one, where '>' means greater regularity or goodness. In three experiments, we have tested these predictions. In experiment 1, we measured percentage correct and reaction time to dot patterns which had one, two, three, or four axes of symmetry and were flashed for 150 ms. Experiment 2 was identical except that patterns were presented for 2000 ms. In experiment 3, dot patterns were replaced by solid shapes which also had one, two, three, or four axes of symmetry. Although it was found that stimuli with four axes clearly allowed superior performance to that of stimuli with one axis, results obtained with stimuli with two and three axes were almost identical and in between those obtained with one and four axes. The data thus support the suggestion that extra axes add 'goodness' to symmetrical patterns but not in a monotonic fashion.     

The control of amplitude and direction in a bimanual coordination task

Bimanual coordination requires task-specific control of the spatial and temporal characteristics of the movements of both hands. The present study focused on the spatial relationship between hand movements when their amplitude and direction were manipulated. In the experiment in question, participants were instructed to draw two lines simultaneously. These two lines were instructed to be drawn in mirror symmetric or perpendicular directions of each other while the length was instructed to be the same or different. The coordinative quality of amplitude control was compared when the task required symmetric and asymmetric bimanual spatial coordination patterns. Results showed that the amplitude accuracy decreased when different amplitudes and/or directions had to be generated simultaneously. The coordinative quality of direction was also compared when the task required symmetric and asymmetric bimanual spatial coordination patterns. Unlike amplitude, the direction accuracy was largely independent of coordination symmetry/asymmetry of direction or amplitude. The results suggest that the coordinative quality of amplitude control does not only interfere with amplitude asymmetry, but it also interferes with direction asymmetry. Moreover, in bimanual coordination amplitude control is more vulnerable to the influence of direction control demands than vice versa.     

Rule learning in rats: serial tracking in interleaved patterns

Humans have the ability to chunk together information from nonadjacent serial positions in sequential patterns. For example, human subjects can extrapolate the pattern, A-M-B-N-C-O-D-P-E-..., to find the missing element, Q, by sorting pattern elements into two component interleaved subpatterns: A-B-C-D-E and M-N-O-P-... Two experiments investigated the ability of rats to reorganize pattern elements from nonadjacent serial positions into chunks not presented by the experimenter. Rats learned either a structured or unstructured sequence interleaved with elements of a repeating sequence (experiment 1) or an alternation sequence (experiment 2). In both experiments, rats learned the interleaved subpatterns at different rates. Acquisition rate was correlated with the structural properties of component subpatterns and the nature of the rules required to describe the interleaved subpatterns. The results indicate that rats are sensitive to the organization of nonadjacent elements in serial patterns and that they can detect and sort structural relationships in interleaved patterns. Received: 7 September 1998 / Accepted after revision: 27 January 1999     

Children's performance in mental rotation tasks: orientation-free features flatten the slope

Studies of the development of mental rotation have yielded conflicting results, apparently because different mental rotation tasks draw on different cognitive abilities. Children may compare two stimuli at different orientations without mental rotation if the stimuli contain orientation-free features. Two groups of children (78 6-year-olds and 92 8-year-olds) participated in an experiment investigating development of the ability to mentally rotate and the ability to recognize and use orientation-free features. Children compared two stimuli, one upright and one rotated, and responded as quickly as possible indicating whether the stimuli were the same or different. The stimuli were either two panda bears or two ice-cream cones with three scoops of ice-cream of different colors. The panda bears were either identical or mirror images. The cones were either identical, mirror images, or non-mirror images. Response times increased linearly as a function of the angle of orientation when stimuli were the same and when the stimuli were mirror images. But response times were much less dependent on angle of orientation for non-mirror image stimuli. Children as young as 6 years recognized orientation-free stimulus features and responded without mentally rotating when the task permitted this strategy.     

Pattern detection by mongol and non-mongol subnormals.

Ten mongols and ten clinically heterogeneous subnormals matched on chronological age, mental age and digit span took part in an experiment in which tape-recorded supra-span digit sequences with different patterns were presented. There were six patterns: random, mirror (e.g. 583385); same-digit pairs (e.g. 558833), same-digit throughout (e.g. 333333), couplet repetition (e.g. 585858) and triplet repetition (583583). The numbers of digits correctly recalled in any order by the mongols in the various conditions ranked from least to most were: random, mirror, same-digit pairs, same-digit messages, triplet repetition and couplet repetition. The rank order for the non-mongols was the same except that the positions of couplet and triplet repetition were reversed. Mongols had significantly poorer recall for random, mirror and same-digit pair messages than non-mongols but were their equals in other conditions. The mongols' performance was more sensitive to pattern than the performance of the other subjects. There was some evidence that, in the messages with same-digit pairs and the same digit throughout, all subjects (but mongols in particular) tended to insert new digits into the response sequence and that the digit introduced was the next one in simple arithmetic progression. It would appear that the hypothesis about poor auditory-vocal channelling capacities of mongols needs qualification.     

The effect of prolonged viewing on the recognition of global and local levels of hierarchically constructed patterns

Two experiments were carried out to examine delays in the recognition of test patterns after prolonged (25 s) viewing of an adaptation pattern which was composed of either the same or different local components and global structures. In the case of geometric patterns that contained a few relatively large elements, there were significant delays of about 70 ms in the global decision task when the test and adaptation stimuli (TS and AS) were of the same pattern, or when they had the same structure but different components. In contrast, in the case of geometric patterns that were composed of many relatively small elements, such a figural relation between the AS and the TS did not affect the recognition speed of the TSs either in the global or in the local decision task. These results suggest that prolonged viewing reduces the efficiency of global processing and that such a disruptive effect occurs depending upon the number and the relative size of the elements that make up the whole pattern.     

The effect of location on the discrimination of spatial vibrotactile patterns

The present study examined whether the locations of patterns on the skin affected the ability to process information about their shapes. In Experiment 1, pairs of spatial vibrotactile patterns, using the array from the Optacon, were presented sequentially to subjects’ left index fingerpads. The location of each pattern in a pair was varied randomly among four locations on the skin. The subjects responded “same” or “different” on the basis of the shapes of patterns, regardless of their locations. Discrimination accuracy was highest and response time fastest when patterns occupied identical locations (ILs), and performance suffered with increasing distance between patterns. In Experiment 2, pairs were presented to corresponding points or to noncorresponding points on separate fingerpads. When patterns occupied corresponding points on separate fingers, accuracy was lower than when patterns occupied ILs on a single finger, but higher than when patterns occupied noncorresponding points on separate fingers. The results suggested that discriminability declined partly because patterns did not occupy ILs, and partly because separate locations had different densities of innervation.     

Visual memory and the perception of a stable visual environment

The visual world appears stable and continuous despite eye movements. One hypothesis about how this perception is achieved is that the contents of successive fixations are fused in memory according to environmental coordinates. Two experiments failed to support this hypothesis; they showed that one's ability to detect a grating presented after a saccade is unaffected by the presentation of a grating with the same spatial frequency in the same spatial location before the saccade. A third experiment tested an alternative explanation of perceptual stability that claims that the contents of successive fixations are compared, rather than fused, across saccades, allowing one to determine whether the world has remained stable. This hypothesis was supported: Experienced subjects could accurately determine whether two patterns viewed in successive fixations were identical or different, even when the two patterns appeared in different spatial positions across the saccade. Taken together, these results suggest that perceptual stability and information integration across saccades rely on memory for the relative positions of objects in the environment, rather than on the spatiotopic fusion of visual information from successive fixations.