The role of pattern goodness in the reproduction of backward masked patterns.

Abstract

Four experiments investigated the role of pattern goodness in backward masking using five- and four-dot patterns constructed by placing dots in the cells of a 3 x 3 matrix. In Experiment 1, subjects rated the goodness of these patterns and the results replicated previous work showing that good patterns had few alternatives. In Experiment 2, the dot patterns were the target stimuli in a backward masking task using a variety of masking stimuli. For all masking, good patterns were reproduced more accurately than poor patterns. In Experiment 3, the goodness of the masking stimulus was varied. The results indicated that good patterns were reproduced more accurately (replicating Experiment 2) and that good patterns were less effective as stimulus than were poor patterns. In Experiment 4, a long interstimulus interval which precluded masking was used to determine whether goodness affected encoding or memory. At these intervals, there were no differences among patterns, suggesting that the effect of pattern goodness was on rate of encoding. These results demonstrate the importance of configural properties in pattern perception.

Effects of good form and spatial frequency on global precedence

“…More current work has shown that good form benefits all stages of information processing. For example, good form facilitates encoding (Bell & Handel, 1976;Garner & Sutliff, 1974) as well as memory and comparison processes (Checkosky & Whitlock, 1973).Unlike the Gestalt view, recent work by Navon (1977, 1981, 1991) has suggested a different explanation for the priority of wholes over parts. Navon's primary hypothesis is that the pattern or global level of a form is always encoded faster than is more local information and hence is available sooner to recognition and response processes.…”

“…More current work has shown that good form benefits all stages of information processing. For example, good form facilitates encoding (Bell & Handel, 1976;Garner & Sutliff, 1974) as well as memory and comparison processes (Checkosky & Whitlock, 1973).…”

Is it really search or just matching? The influence of Goodness, number of stimuli and presentation sequence in same–different tasks

“…This result was later criticized for possibly probing into different encoding requirements, as the task was reproduction rather than comparison (Pomerantz, 1977). Sebrechts and Garner (1981) seemed convinced by the evidence by Bell and Handel (1976), but suggested that the influence of Goodness on encoding is relatively small compared to that on memory.…”

“…If stimuli with high Goodness are reproduced more accurately than stimuli with lower Goodness, this would suggest that good stimuli are encoded more quickly than poor stimuli. This approach was taken by Bell and Handel (1976), who indeed found better reproduction accuracy for good patterns, and this difference between good and poor stimuli disappeared when no mask was used. This result was later criticized for possibly probing into different encoding requirements, as the task was reproduction rather than comparison (Pomerantz, 1977).…”

“…Several previous studies using the Garner dot pattens (Garner & Clement, 1963) have attempted to distinguish between the influences of stimulus encoding and memory search (Bell & Handel, 1976;Checkosky & Whitlock, 1973;Clement & Varnadoe, 1967;Garner & Sutliff, 1974;Pomerantz, 1977;Sebrechts & Garner, 1981). One approach taken was to use the Sternberg (1969) methodology (the 'additive factor method') in which separate processes are assumed when two factors are shown to influence response times independently as demonstrated by a non-significant interaction in an analysis of variance.…”