Eye movements as direct tests of the GO model for the missing-letter effect

Abstract

When asked to detect target letters while reading a text, participants miss more letters in frequently occurring function words than in less frequent content words. To account for this pattern of results, known as the missingletter effect, Greenberg, Healy, Koriat, and Kreiner (2004) proposed the guidance-organization (GO) model, which integrates the two leading models of the missing-letter effect while incorporating innovative assumptions based on the literature on eye movements during reading. The GO model was evaluated by monitoring the eye movements of participants while they searched for a target letter in a continuous text display. Results revealed the usual missing-letter effect, and many empirical benchmark effects in eye movement literature were observed. However, contrary to the predictions of the GO model, response latencies were longer for function words than for content words. Alternative models are discussed that can accommodate both error and response latency data for the missing-letter effect.

Solving the chicken-and-egg problem of letter detection and fixation duration in reading

“…First, we analyzed the subset of 36 d words from the d-search task. The aim of this analysis was to establish the baseline effects with regard to fixation, omission, and response latency among the subset of words used for the main analyses, as well as reproducing the effects observed by Roy-Charland et al (2007) with the complete data set of 64 participants. Second, the p words from the baseline condition were analyzed.…”

“…To promote reading for comprehension, the participants were informed that they would have to answer, on a sheet of paper, three multiple-choice comprehension questions after reading the entire text. These instructions were identical to those used by Roy-Charland et al (2007).…”

“…The data set from participants reading and searching for the target letter d is composed of the 42 participants in Roy-Charland et al (2007), as well as 22 participants who were added to the previous sample in order to create equal samples for the baseline condition in which the participants are searching for the target letter p and the d-search task. The overall sample of 64 participants in the d-search task consisted of unpaid undergraduate students with a mean age of 20 years.…”

“…Furthermore, they were also exactly the same as the ones used in the baseline condition, except that the participants were searching for d instead of p. In sum, 128 participants took part Fixation durations. All fixation duration measures used by Roy-Charland et al (2007) were used here: (1) first-fixation duration (i.e., the duration of the first forward fixation on a target word, independent of the number of fixations made on the word), (2) single-fixation duration (i.e., the duration of the fixation when only a single forward fixation was made on the target word), (3) gaze duration (i.e., the sum of all forward fixations on a target word prior to a saccade to another word), and (4) total fixation time (i.e., the sum of all fixations, including regressions and forward fixations if there was a regression on the target word). Measures 1-3 were conditional on the target word's being fixated on the initial pass through the text, whereas Measure 4 included all fixations made on a word, regardless of whether it was made on the initial pass through or not.…”

“…In a recent study, Roy-Charland et al (2007), following Saint-Aubin and Klein (2001), monitored eye movements while participants were reading for comprehension and searching for a target letter. The results revealed the usual missing-letter effect, with more omissions for frequent function words than for less frequent content words.…”