Lab Report Assignment: role of iconic memory PDF

Preview

Summary

Effect of Number of Required Stimuli and Post-stimulus Cues on Percentage of Stimuli Correctly Identified: Role of Iconic Memory...

Description

Running Head: ROLE OF ICONIC MEMORY

Laboratory Report: Effect of Number of Required Stimuli and Post-stimulus Cues on Percentage of Stimuli Correctly Identified: Role of Iconic Memory

1

ROLE OF ICONIC MEMORY

2

Abstract The experiment was conducted to investigate the partial report superiority effect with a view to building understanding of pre-categorical iconic memory. It included 347 students and involved three conditions: Whole Report, Partial Position Report and Partial Category Report. Alphanumeric matrices with eight items in two rows were presented briefly. In Whole Report, all items were required to be reported. In Partial Position Report, poststimulus cue of spatial information was given, requiring stimuli in one row. In Partial Category Report, cue of stimuli identity asked for reports of either letters or digits. The hypothesis that partial report superiority effect was present in Partial Position Report compared to Whole Report and not present in Partial Category Report was supported by the order of average percentage of stimuli correctly identified in three conditions. This provided support for the presence of iconic memory and for its characteristics as a pre-categorical sensory storage. More experiments could be conducted to solidify the presence of more than one memory storages, one of which including different cue delays and a mask was suggested.

ROLE OF ICONIC MEMORY

3

Effect of Number of Required Stimuli and Post-stimulus Cues on Percentage of Stimuli Correctly Identified: Role of Iconic Memory Questions are often raised on how brief visual representations are perceived and how much information can be retained and accessed. To investigate this, Sperling (1960) conducted multiple experiments based on two main conditions. They were often known as Whole Report (WR) and Partial Report (PR), in which all items versus a portion of presented items respectively were required to be reported. More stimuli overall could be stored in memory for PR, as represented by PR performance, than in WR, the difference between which was known as partial report superiority effect (Coltheart, 1983; Sperling, 1960). Experiments found that around four to five items could be correctly reported in WR yet around nine in PR (Sperling, 1960). One main explanation for this considered two memory storages. The initial stage, “iconic memory” (Neisser, 1967), was pre-categorical, or nonprocessing, rapidly decaying and high-capacity. The subsequent stage, “durable storage” (Coltheart, 1980, p. 186), was post-categorical, long lasting yet low-capacity. It was where stimuli were processed, retrieved and then reported. Relevant stimuli could be selected for selective transfer between these two storages. Alternatively, non-selective transfer could occur with no selection conducted (Coltheart, 1980). To build an understanding of this partial report superiority effect and iconic memory, an experiment was created using similar design to Sperling’s experiment. In addition to WR, two conditions of partial reports were conducted. Partial Position Report (PPR) was where spatial location indicated required stimuli while in Partial Category Report (PCR), identity of stimuli, letters or digits, was criterion. Based on Sperling’s experiment, it was hypothesised partial report superiority effect was present in PPR compared to WR, and not in PCR. Higher number of stimuli required, in WR, and complicated cue requiring semantic information – stimuli identity, in PCR, would decrease the percentage of stimuli correctly identified. The reason was that if about four to

ROLE OF ICONIC MEMORY

4

five items could access durable storage (Coltheart, 1980), only about half of required stimuli could be correctly identified in WR. However, the given cue of spatial information could facilitate the selective transfer during PPR to achieve higher accuracy in four items reported. As iconic memory did not process presented information, cue for selective transfer could only be based on episodic information, such as location, rather than semantic information. Therefore, selection of letters and numbers happened poorly, if at all, hence the absence of superiority effect in PCR. Critical analysis could propose iconic memory might not be the only reason for partial report superiority effect. Instead, afterimages, allowing stimuli to be visible to participants for several seconds to minutes after stimulus offset, (Berry & Imus, 1935; Colheart, 1980) could have been in place to produce higher accuracy in PPR compared to WR. However, if the absence of superiority effect in PCR was tested to be true, afterimages could not provide persuasive explanations. In fact, afterimages were rejected by Sperling (1960) based on participants’ reports of seeing just single brief flash rather than lasting afterimages. To further understand iconic memory and test the above hypothesis, the experiment was as follows. Method Participants The participant group included 347 undergraduates from the psychology unit, PYB102, at QUT. No specific criteria were used for selection. Design The independent variables were the number of stimuli required to be reported and the type of information required in post-stimulus cues. Three conditions were included. WR required participants to report as many stimuli as they could, in the correct position. In PPR,

ROLE OF ICONIC MEMORY

5

with cue using episodic information – location, four stimuli in one of the two rows, were required in correct position. In PCR, cue using semantic information was provided and four stimuli of either letters or numbers in their correct position were required. The dependent variable was the percentage of stimuli correctly identified. Apparatus and Procedure Participants were given a link to access the experiment when they had Internet connection on any devices. Activation of the link presented detailed instructions of one in three conditions for the subsequently displayed five trials. Instructions were also shown in the other two conditions after each five trials. After clicking ‘Next’, a fixation cross was shown, followed by an alphanumeric matrix with eight items in two rows, two digits and two letters in each row. Fifteen matrices for three conditions were displayed for 200ms each, with counterbalanced order of item presentation. After this, a blank page was presented, followed by a response page with eight empty squares to fill in required stimuli. For WR, as many stimuli as participants could recall were required to be typed into correct squares from keyboards. For the two partial reports, two auditory probes, high-pitched and low-pitched, were played with no delay after stimulus offset to give post-stimulus cues. Previously on ‘Instructions’ page, participants had been given a chance to listen to these probes; use of headphones was of participants’ choice. For PPR, high-pitched probe meant top row stimuli were required and low-pitched meant bottom row. For PCR, letters were required if highpitched sound was heard and numbers if low-pitched probe. Responding time was determined by participants. Participants clicked ‘Next’ after finishing responding for each trial. Results

ROLE OF ICONIC MEMORY

6

Average Percentage of Stimuli Correctly Identified in Three Experimental Conditions

Stimuli Correctly Identified (%)

60

50 49.45 40 39.16 30 30.61 20

10

0 Whole Report

Partial Position Report

Partial Category Report

Experimental Conditions Figure 1. Average percentage of stimuli correctly identified in three experimental conditions: Whole Report, Partial Position Report and Partial Category Report. The data presented the average percentage of stimuli correctly identified in three conditions, illustrated in Figure 1. The largest percentage was recorded in PPR, at mean percentage of 49.45% and standard deviation, SD = 18.05%. This was higher than the figure obtained in WR, Mean = 39.16%, SD = 9.94%, which was higher than mean percentage of 30.61% (SD = 15.77%) of stimuli were correctly identified in PCR. The variability of participants’ percentage of accurately reported stimuli varied considerably. The highest variability recorded in PPR nearly doubled the lowest figure in WR. The difference between the second highest figure of PCR and the highest one was approximately half as much as that with the lowest variability figure. Analysis of t tests and p values indicated statistical significant difference in the percentage of stimuli accurately identified between PPR and WR,

ROLE OF ICONIC MEMORY

7

t(346) = 10.78, p...