Participants
An opportunity sample of 48 undergraduate psychology students participated as part of their course. Consent was obtained verbally by the experimenter.
Materials
Measure of visual strategy.- The stimulus was based on the “random angular shapes” used by Cooper (1976, p.435). The shapes varied in the number of edges or points as a measure of complexity (Cooper, 1976). The Five standard shapes, composed of 6, 8, 12, 16, and 24 edges (see, Appendix.1 for details). Associated with each, were six distractors, which were random disturbances of the standard shape, and all varied in similarity to the standard shape by reflection or perturbation (see Cooper, 1976).
Block Design Task.- The stimulus arrays were presented at a size of 63 x 63 pixels; each individual block in a four-block pattern was 32 x 32 pixels.
Patterns varied in complexity representing three levels of: 1,2 & 3 internal edges. Eight different patterns for each level of complexity made a total of 40 altogether (see Appendix. 2).
Procedure
An Apple Macintosh computer was used for both tasks. All stimulus arrays were presented at a resolution of 640 x 480 pixels. Participants all sat at a distance of 50cm from the monitor. In both tasks participants responded by pressing the ‘L’ button to indicate whether the test form was the same or the ‘S’ button to indicate a difference. The corresponding keys were reversed for those who were left-handed. Participants were also asked to respond to all trials as rapidly as possible without sacrificing accuracy for speed.
Visual comparison task.-
Subjects viewed one of five standard visual forms. Following the disappearance of the standard shape, a blank field was presented for the duration 500 msec. Immediately following the offset of the blank field, a test form was presented and a response of same or different was made. Three sets of trials were completed, the first consisting of 60 practice trials and two blocks of 60 experimental trials. Second and third sets of 10 practice trials and two blocks of 60 experimental trials. The test form was displayed in the same orientation as the standard form in all trials, it could either be the same or different through experimental manipulation (see Cooper, 1976). In all trials there was an equal probability of a same or different response. In addition a trial replacement procedure was also used, where incorrect trials were added to the trial set along with a balancing trial.
Block Design Task.-
A stimulus pattern was presented for a fixed duration of 2sec, following the offset, a fixation point was presented centrally in the form of a cross for 3sec. Immediately following the fixation point, a response pattern was presented. Only one of the four blocks composing the pattern was displayed however the remaining three patterns were left blank. Participants had to decide whether the block presented in that position was the same or different to that which appeared previously.
Two trial sets, the first, 10 practice trials and 4 blocks of 40 experimental trials; the second, 5 practice trials and 4 x 40 experimental trials. Each pattern was presented on a “same” trial 4 times (40 x 4) = 160 “same” trials. Different patterns were determined by a change in one of the four block positions. Each pattern was changed in each of the four block positions creating four versions of each pattern (40 x 4) = 160 “different” trials. Each trial set of 160 contained equal numbers of same and different trials and equal numbers of patterns of each level of complexity.
Results
According to the pattern of performance observed in the visual comparison task, participants were allocated to one of two groups of visual strategy (Analytic or Holistic). Twenty-five Participants were classified Analytic and twenty-three Holistic. Separate analysis of variance were performed on “same” and “different” reaction times of the two types of subjects and on their error rates, thus, four, two-way (mixed model) analysis of variances were performed altogether.
Reaction Times “same” response;
Mauchly’s Test of Sphericity gives a non-significant result (W = 0.988, p = 0.766). From the tests of within-subjects effects and test of between subjects effects tables of output the following can be inferred.
There was no main effect of complexity on performance [F (2,92) = .371; Mse = 3722, p = .69], indicating that performance on the block design memory task between both groups was not primarily affected by the levels of complexity.
There was a significant main effect of perceptual style [F (1,46) = 5.2; Mse = 99989, p = .028], clearly indicating that memory performance on the block design task was heavily influenced by the type of visual processing style adopted, holistic subjects responding much faster than analytic.
The main interaction effect of perceptual style, complexity and response times was significant [F=(2,92) = 7.58, Mse = 3722, p < .001]. Figure 1 presents the mean reaction time data for both, type of subject and complexity.
Figure 1. Mean “same” response times as a function of complexity and perceptual style.
Note that for holistic subjects, response times are generally much faster than analytic subjects, it is also clear to see that for holistic subjects, response times generally decrease monotonically in relation to the number of internal edges, whereas for Analytic subjects, response times generally increase monotonically in relation to number of internal edges.
Reaction Times “different” response;
Mauchly’s Test of Sphericity gives a non-significant result (W = 0.986, p = 0.731). From the tests of within-subjects effects and test of between subjects effects tables of output the following can be inferred.
There was no main effect of complexity on performance [F (2,92) = .602; Mse = 2670, p = .55], indicating that reaction times to “different” responses between both groups, was not significantly effected by levels of complexity on the block design memory task.
There was a significant main effect of perceptual style [F (1,46) = 9; Mse = 1057, p .004], clearly indicating that response times on the block design task were heavily influenced by the type of visual processing style adopted, holistic subjects responding much faster than analytic as before.
Most importantly the main interaction effect of perceptual style, complexity and response times was significant [F=(2,92) = 4.15, Mse = 2670, p = .019]. Figure 2 presents the mean reaction time data for both, type of subject and complexity.
Figure 2. Mean “different” response times as a function of complexity and perceptual style.
Note as before that holistic subjects display much faster response times overall than analytic subjects, the most noticeable difference was between the response times of both type of subject and complexity level of 2 edges, holistic subjects were able to process the image 170 msec faster than analytic. Despite this individual difference, overall it can be observed that generally the more complex the shape was the faster response times were for holistic subjects where as for analytic subjects the opposite performance was observed.
Error rates “same” response
Mauchly’s Test of Sphericity gives a non-significant result (W = 0.920, p = 0.152). From the tests of within-subjects effects and test of between subjects effects tables of output the following can be inferred.
The main effect of complexity on performance accuracy was significant [F (2,92) = 22.7; Mse = .0098, p < .001], indicating that performance accuracy of both groups on the block design memory task was affected by the complexity level of the shape.
There was a non-significant main effect of perceptual style [F (1,46) = .095; Mse = .038, p = .4 ], indicating that the type of visual processing style adopted did not influence performance accuracy on the block design task. I.e. there was no difference in errors made between the two types of subjects.
The main interaction effect of perceptual style, complexity and accuracy was non-significant [F=(2,92) = .095, Mse = .0098, p = .91]. Which was reasonable to expect since participants regardless of perceptual style consistently made similar errors in relation to complexity. Figure 3 presents the mean reaction time data for both, type of subject and complexity.
Figure 3. Mean error rates plotted as a function of complexity and perception styles made on “same” response condition.
By averaging error rates of both groups across the complexity levels of the shape in figure 3, it is possible to note the similarities in the performance accuracy of both groups as to why no interaction effect was found. More specifically the performance of both groups on the shape with two internal edges was significantly better than for 1 or 3 edges which reflects on the internal properties of the shapes. Overall it is clear to see that analytic subjects were slightly more accurate than holistic subjects although, these differences were too close to relate to subject type itself and that performance was significantly affected by complexity.
Error rates “different” response
Mauchly’s Test of Sphericity gives a non-significant result (W = 0.975, p = 0.570). From the tests of within-subjects effects and test of between subjects effects tables of output the following can be inferred.
The main effect of complexity on performance accuracy was significant [F (2,92) = 19.66; Mse = .023, p < .001], indicating that the complexity level of the shape affected performance accuracy of both groups on the block design memory task.
There was a non-significant main effect of perceptual style [F (1,46) = 1.85; Mse = 0.1, p = .18] indicating that performance accuracy on the block design task was not influenced by the type of visual processing style adopted. I.e. there was only a slight difference in amount errors made between the two types of subjects.
The main interaction effect of perceptual style, complexity and accuracy was non-significant [F=(2,92) = 1.6, Mse = .023, p = .198]. Which as in the previous analysis of “same” error rates was reasonable to expect since all participants regardless of perceptual style consistently made similar errors in relation to complexity. Figure 4 presents the mean reaction time data for both, type of subject and complexity.
Figure 4. Mean error rates plotted as a function of complexity and perception styles made on “different” response condition.
As with error rates of “same” response, it is possible once again to note that overall analytic subjects were slightly more accurate than holistic subjects, although these differences were too close to relate to subject type itself and thus explain why no interaction effect was found. At a closer inspection, it is also possible to note how, performance of both groups was significantly impaired by the least complex shape in comparison to the most complex shape thus it is clear to see that performance was significantly effected by complexity.
Discussion
Individual differences in perceptual processing (Cooper, 1976) were found in relation to visual memory on the block design task, pattern complexity also affected the efficiency to which perceptual processing occurred within the two visual styles. Furthermore the current findings suggest that in line with Cooper (1976, 1982) efficient performers follow a holistic strategy. The findings of which, displayed in figures 1 to 4 will be outlined in brief.
For the holistic type subjects : (a) correct “same” response times were faster than correct “different” response times; (b) correct different response times were effected by complexity; (c) overall error rates consistently similar for both “same” and “different” responses; (d) reaction time correct “same” responses decreased as complexity increased: (e) overall response times fast. For the analytic subjects : (a) correct “same” response times faster than correct “different responses; (b) correct “same response times increased monotonically in relation to complexity; (c) errors were more or less equal for both “same” and “different” responses; (d) response times relatively slow.
Explanations for these marked performance difference are as follows. According to Schorr et al., (1982, p.480) efficient performers in block design tasks are able to mentally segment pattern components and attend equally to each one. Royer (1977, cited in Schorr, 1982) believed efficient subjects are able to perform greater amounts of “mental slicing”. Schorr et al., (1982, p.480) argued that segmentation is dependant on the “number of perceptual cues indicating edges”.
Such views map accordingly with current findings. Note, that differences observed between the two perceptual styles can be explained by Schorr et al., (1982) in relation to “a person’s sensitivity to edge cues”. Also understood in terms of a person’s preference for either global or local processing (Navon, 1977, cited in Eysenck & Keane, 2002). Schorr et al., (1982, p.480) believe efficient performers “ follow interior edges in aligning a local, one block, mental grid with block edges” thus holistic processing style attend to local and global properties of a pattern whereas analytic style attend only to global properties i.e. the gestalt appearance.
It follows that complexity has advantage effects on reaction time for holistic subjects but not analytic. Navon (1977, cited in Eysenck, 2002, p.85) suggested that “perceptual processing proceeds from global structuring to more fine grained analysis” presumably at a more local level. Holistic subjects therefore may be more efficient in their overall processing from global to local information than analytic, thus by processing information faster they are by virtue able to segment the pattern into components easier than analytic.
Such explanations are consistent with previous research (Cooper, 1976, Schorr, 1982) as they are able to account for both the qualitative and quantitative differences between the two perceptual strategies. However the task domains with which such research has been compared present significant methodological differences (Kimichi, 1992), for instance, overall pattern complexity has been characterized in many different forms throughout the proposed literature (Cooper, 1976, 1982, Kimichi, 1992., Schorr et al.,1982., Royer, 1977, cited in Schorr, 1982., Navon, 1983) further studies reveal that eccentricity and relative size of stimulus also influence a global processing advantage (Navon, 1983, Pomerantz & Pristach, 1989 cited in Kimichi, 1992).
According to Kimichi (1992) assumptions regarding global advantage cannot be rightly inferred, until more is known about perceptual units of a stimulus structure, such discrepancies may advocate improper characterization of perceptual processing. In line with Kimichi (1992), it is thus considered that the true nature of perceptual processing must be treated carefully in respect to the mapping of one processing style onto another, as no specific processing strategy is necessarily inherent of the same characteristics as another.
In summary, perceptual processing styles were found in relation to visual memory on the block design task. These strategies can be understood in a variety of forms (Schorr, et al, 1982, Navon 1977, cited in Eysenck, 2002) however it is consistently noted that these are qualitative as well as quantitative differences (Cooper, 1976, 1982). Interpretation of the current results suggests that the perceptual processing styles identified by Cooper (1976) can be used to assess visual functioning in a reliable format. Future research into the holistic style of processing is likely to yield more fruitful observations, as the cognitive capabilities of such processing is far from understood. Further research points to the perceptual structure or complexity of visual patterns. Finally, it is suggested that in order to make reliable conclusions about the nature of perceptual processing, research within the field need concern the use of “converging operations” and task demands that allow such inferences to be made (Kimichi, 1992, p.36).
References
Cooper, L. A. (1976). Individual differences in visual comparison processes. Perception & Psychophysics, 19(5), 433-444.
Cooper, L. A. (1982). Strategies of Visual Comparison and Representation: Individual Differences. In R.J.Sternberg. (Ed.). Advances in the Psychology of Human Intelligence. Vol.1. CH 2, 77-124.
Eysenck, M., & Keane, M. (2002). Cognitive Psychology: A Students Handbook. Philadelphia: Taylor and Francis Inc.
Hoffman, J. E. (1980). Interaction between global and local levels of a form. Journal of Experimental Psychology: Human Perception & Performance, 6, 222-234.
Kimichi, R. (1992). Primacy of wholistic processing and global/local paradigm. A critical review. Psychological Bulletin, 112, 24-38.
Navon, D., & Norman, J. (1983). Does global precedence really depend on visual angle? Journal of Experimental Psychology: Human Perception & Performance, 9(6), 955-965.
Palmer, S.E. (1999). Vision Science: Photons to phenomenology. Cambridge, Massachusetts: Bradford Books.
Schorr, D., Bower, G.H., & Kiernan, R. (1982). Stimulus variables in the block design task. Journal of Consulting and Clinical psychology, 50(4), 479-487.
