Academic journal article Memory & Cognition

The Role of Verbal Memory in Regressions during Reading

Academic journal article Memory & Cognition

The Role of Verbal Memory in Regressions during Reading

Article excerpt

Published online: 15 August 2012

© Psychonomic Society, Inc. 2012

Abstract During reading, participants generally move their eyes rightward on the line. A number of eye movements, called regressions, are made leftward, to words that have already been fixated. In the present study, we investigated the role of verbal memory during regressions. In Experiment 1, participants were asked to read sentences for comprehension. After reading, they were asked to make a regression to a target word presented auditorily. The results revealed that their regressions were guided by memory, as they differed from those of a control group who did not read the sentences. The role of verbal memory during regressions was then investigated by combining the reading task with articulatory suppression (Exps. 2 and 3). The results showed that articulatory suppression affected the size and the accuracy of the initial regression but had a minimal effect on corrective saccades. This suggests that verbal memory plays an important role in determining the location of the initial saccade during regressions.

Keywords Reading * Memory

Because of the structure of the eye, the acquisition of information during reading is limited to a small area around the fixation point, which prompts readers to constantly move their eyes on the line. Research shows that during reading, the eyes generally move rightward (Rayner & Pollatsek, 1989). However, 10 % to 15 % of these eye movements, called regressions, are made in the opposite direction, to words that have already been fixated. It has been suggested that regressions are made in order to reanalyze words that have been incorrectly interpreted (see, e.g., Frazier & Rayner, 1982; Kennedy & Murray, 1984). In these cases, the eyes are not directed randomly, but aim at a precise word location (Kennedy & Murray, 1987), suggesting that regressions are guided by memory. Some studies have suggested that regressions are mostly guided by a spatial code attached to each word (see, e.g., Kennedy, 1992; Kennedy, Brooks, Flynn, & Prophet, 2003). Although some studies point to a contribution of verbal memory during regressions (see, e.g., Rawson & Miyake, 2002; Weger & Inhoff, 2007), little convincing evidence exists. The objective of the present study was therefore to examine the contribution of verbal memory1 during regressions through the use of verbal interference.

In order to examine the role of memory during regressions, researchers usually measure the pattern of eye movements. For instance, a precise regression is taken as evidence that the reader remembers the word's location. In Kennedy and Murray's study (1987; see also Kennedy et al., 2003), participants were asked to read a sentence that was followed by a test word presented visually at the end of the line. Their task was to identify whether or not the test word had been presented in the sentence. Kennedy and Murray (1987) found that when participants made an eye movement toward the target word in the sentence, the regression was very accurate and landed on the target word. The researchers argued that such a high level of precision would only be possible if the spatial coordinates of the words were retained in memory during reading. Additional evidence for this theory was provided by Kennedy et al. (2003) using a similar procedure. They showed that if the sentence was shifted slightly to the right before the regression was initiated, participants overshot the target word more often than during a condition with no shift, suggesting that regressions are guided by memory for the position of the word relative to the page layout, rather than by the visual configuration of the sentence (see also Inhoff, Weger, & Radach, 2005).

Other studies have investigated the contribution of verbal memory during regressions. For instance, Therriault and Raney (2002) showed that after reading a text, participants were no better than chance at relocating where information was presented relative to the page layout. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.