Methods
Twenty subjects were used in this experiment, all were students at Alma College and were personally recruited by the experimenters.
Two versions of the experiment were created using Superlab software on Macintosh computers. One version was a three-by-three grid of numbers the other was made of various shapes(figures 1 and 2). Both versions of the experiment included partial and whole report sections which were presented separately. The statistics were kept on paper until the experiment ended at which time they were examined using Statview software.
Subjects were placed in groups depending on what order they arrived to participate in the experiment(they were not assigned specific times). Four groups were used: 1)Letters with ten trials of whole report followed by ten trials of partial report, 2)shapes with ten trials of whole report followed by ten trials of partial report, 3)letters with the whole and partial report in reverse order and 4)shapes with whole and partial report in reverse order. The presentation of whole and partial report was varied in order to control for one effecting the other but no noticeable difference between the sub-groups was detected, so they were combined into two basic groups(letters and shapes) ten subjects in each. A warm-up trial of five sets of partial report number grids was given before the beginning of each trial in order to familiarize the subjects with partial report.
Grids were presented to subjects in random order. Each was flashed on the screen for 60 milliseconds, at which time the subjects were to write down what they had seen on a score sheet. The grid was then re-presented for an indefinite amount of time and the subjects were instructed to score themselves based on the number of letters or shapes they had recalled, they were then allowed to move on to the next grid. This cycle was controlled by the subject using the spacebar on the computer's keyboard. Partial report was assigned to either the top, middle or bottom row by a voice that immediately followed the grid presentation. The pitch of the voice coincided with the commands "high," "medium" or "low".
Results
In both cases(partial and whole report) the subjects, on average, were able to correctly identify more letters than shapes. A repeated measures ANOVA showed that subjects performed signficantly better with letters than shapes [F(1,198)=184.63, (p= .0001)].(figure 3). And within both groups the adjusted partial report was higher than the whole report. There was a significant interactin between type of report and stimulus type indicating the that partial report effect is stronger for letters than shapes [F(1,198)=85.51, (p=.0001)]. Finally, the interaction between shapes with and without sound against letters with and without sound proved significant using the same repeated measures ANOVA [F(1,198)=9.02, (p=.003)].
Discussion
Judging by the results, subjects had an easier time recalling the letters than the shapes. It is likely due to the fact that letters are more common than the shapes that were used. The importance of these findings is that because shapes are less common they may prove to be a better measure of iconic memory in future experiments. Since real life is not entirely about what is common, there is quite a bit of uncertainty as to what types of stimuli one will encounter in a day-to-day routine. It is also uncertain what the individual will do with a stimulus once it has been experienced. An expansion of this experiment may be to mix the shapes in with the letters in the same grid and measure how each effects the outcome of the other.
References
Di Lollo, V. & Peter Dixon (1992). Is the icon's worth apples and oranges? Some fruitful thoughts on Loftus, Duncan and Gehrig. Journal of Experimental Psychology: Human Perception & Performance, 18(2), 550-555.
Gegenfurther, K. R. (1993). Information transfer in iconic memory experiments. Journal of Experimental Psychology: Human Perception & Performance, 19(4), 845-866.
Sperling, G. A. (1960) The information available in brief presentation. Psychological Monographs, 74, Whole No. 498. Cited in Anderson, J. R. (1995) Cognitive Psychology and its Implications. W.H. Freeman Company, New York, 85-87.