The Missing-Phoneme Effect in Aural Prose Comprehension

Abstract

When participants search for a target letter while reading for comprehension, they miss more instances if the target letter is embedded in frequent function words than in less frequent content words. This phenomenon, called the missing-letter effect, has been considered a window on the cognitive mechanisms involved in the visual processing of written language. In the present study, one group of participants read two texts for comprehension while searching for a target letter, and another group listened to a narration of the same two texts while listening for the target letter’s corresponding phoneme. The ubiquitous missing-letter effect was replicated and extended to a missing-phoneme effect. Item-based correlations between the reading and listening tasks were high, which led us to conclude that both tasks involve cognitive processes that reading and listening have in common and that both processes are rooted in psycholinguistically driven allocation of attention.

Keywords

reading audio processing attention psycholinguistic processes letter search

Fifty years ago, Corcoran (1966) made a stunning discovery: When asked to read a text for comprehension while searching for a target letter, fluent readers of English miss a disproportionate number of es in the compared with other words. This finding has since been generalized to various words, and the same effect has been found in other languages (Arabic, Chinese, Dutch, French, German, Greek, and Hebrew; for reviews, see Healy, 1994; Klein & Saint-Aubin, in press). Such studies have repeatedly found that readers miss a disproportionately greater number of target letters in function words (which signify grammatical relationships in a sentence) than in content words (which carry meaning). A famous demonstration was provided by Read (1983), who asked participants to search for Fs in the following sentence: “FINISHED FILES ARE THE RESULT OF YEARS OF SCIENTIFIC STUDY COMBINED WITH THE EXPERIENCE OF YEARS.” Even though there are six occurrences of the letter F, between 85% and 90% of adult readers detect only three Fs when they read the sentence for the first time, missing those in the frequent function word of. Healy called this phenomenon the missing-letter effect, and it is formally defined as the higher miss rate for target letters embedded (a) in function than in content words and (b) in frequent than in rare words (for reviews of the earlier work, see Healy, 1994; Koriat & Greenberg, 1994). It is worth noting that in previous studies of the missing-letter effect, the contribution of word function has been isolated by contrasting function and content words of the same frequency (e.g., at vs. it, respectively), and the contribution of word frequency has also been isolated by contrasting content words of various frequencies (e.g., thought vs. thicket; e.g., Roy-Charland & Saint-Aubin, 2006). However, because function words are intrinsically more frequent than content words, in most studies in the missing-letter-effect literature, word frequency and word function covary.

For five decades, this missing-letter effect has been used to illuminate a wide range of cognitive factors involved in reading, including visual factors, such as eye-movement patterns (e.g., Greenberg, Inhoff, & Weger, 2006; Saint-Aubin & Klein, 2001) and letter position within a word (Assink & Knuijt, 2000; Guérard, Saint-Aubin, Poirier, & Demetriou, 2012); lexical and syntactic factors, such as word frequency (e.g., Minkoff & Raney, 2000; Moravcsik & Healy, 1995; Roy-Charland & Saint-Aubin, 2006) and word function (e.g., Koriat & Greenberg, 1991; Saint-Aubin & Poirier, 1997); and reading-specific factors, such as text familiarity (Greenberg & Tai, 2001; Saint-Aubin, Roy-Charland, & Klein, 2007), reading skills, and reading development (e.g., Cunningham, Healy, Kanengiser, Chizzick, & Willitts, 1988; Drewnowski, 1981; Saint-Aubin & Klein, 2008; Saint-Aubin, Klein, & Landry, 2005).

Surprisingly, even if it is well accepted that reading relies heavily on auditory processing, the auditory analog of the missing-letter effect has not been investigated. Our simple objective was to determine whether a similar missing-phoneme effect would be observed when participants are listening to a passage presented aurally while simultaneously listening for a target phoneme.¹ Although this empirical project was conceived before we developed our attentional-disengagement account of the missing-letter effect (Roy-Charland, Saint-Aubin, Klein, & Lawrence, 2007), we now recognize that it might be viewed as a test of that account. According to the attentional-disengagement model, miss rates are higher when attention disengages earlier from a lexical item containing the target letter. Attentional disengagement is assumed to depend critically on cognitive and psycholinguistic factors that would be operating in a relatively parallel manner whether one is reading or listening. According to this view, a missing-phoneme effect should be observed and ought to behave much like the missing-letter effect. Alternatively, it is difficult to see how purely visual accounts of the missing-letter effect would predict a missing-phoneme effect (e.g., Greenberg, Healy, Koriat, & Kreiner, 2004; Hadley & Healy, 1991). For instance, according to the parafoveal-processing hypothesis, letters are more frequently missed in function words than in content words, because the former are skipped more often than the latter. Because visual acuity is lower for skipped words, it would be harder to detect a target letter embedded in them. In this view, there is no a priori reason for observing an auditory equivalent to the missing-letter effect.

A standard phoneme-monitoring task, pioneered by Foss (1969), has been used to explore attention and a wide variety of psycholinguistic issues (Connine & Titone, 1996). In the present study, two groups of Francophone participants were recruited. One group searched for target letters in the standard paper-and-pencil version of the task. They read two carefully constructed French passages for meaning while crossing off all of the target letters. The other group listened to narrated versions of the same passages for meaning and pressed a key whenever they heard the target phoneme. The key dependent variable was how often each critical occurrence of the target (letter or phoneme) was missed.

Method

Participants

Ninety-six unpaid undergraduate students (66 women, 30 men) from the Université de Moncton took part. This sample size is typical of studies in the missing-letter-effect literature. All participants were native French speakers.

Materials

Two prose passages were presented to participants: the des text and the pour-cour text (see Saint-Aubin, Klein, & Roy-Charland, 2003, Experiments 1 and 5, respectively). The des passage contained multiple presentations of a single function word (des, a plural indefinite article) and fewer presentations of different content words. This passage is typical of normal speech and normal text, which usually contains a high rate of repetition of a few function words and a low rate of repetition of multiple content words. However, in the des text, frequency of occurrences within the text is necessarily confounded with word function. In contrast, the pour-cour text contained the same number of repetitions of a single function word (pour, a preposition meaning “for”) and a single content word (cour, a noun meaning “yard”). The pour-cour text is less natural than the des text, but this handicap is overcome by having the function and the content words repeat equally often. Using two different types of passage such as these helps increase the generalizability of the results.

The des text was composed of 593 words, and the target letter was d. In the critical words of this text (des and nine control words), the target letter was always associated with the phoneme /d/. The des text contained 24 instances of the French plural indefinite article des, with a frequency of 10,629 occurrences per million, and 24 instances of three-letter control content words beginning with the letter d, with an average frequency of 454 occurrences per million (New, Pallier, Brysbaert, & Ferrand, 2004). The control content words were drawn from a pool of nine different words—dis (“say”), dit (“says”), dix (“ten”), don (“gift” or “donation”), dos (“back”), duc (“duke”), duo (“duet”), due (“due”), dur (“hard”). These control words appeared between one and four times each, for a total of 24 occurrences across all nine words. The target letter d was also embedded in 14 noncritical words in the des text.

The pour-cour text was composed of 809 words, and the target letter was r. In the critical words of this text (pour and cour), the target letter was always associated with the phoneme /ᴚ/. The pour-cour text contained 16 occurrences of the preposition pour (“for”), with a frequency of 6,200 occurrences per million, and 16 occurrences of the noun cour (“yard”), with a frequency of 105 occurrences per million (New et al., 2004). In addition, the target letter r was embedded in 49 noncritical words in the pour-cour text.

For each text, each word containing the target letter or phoneme, whether critical or not, was separated from the previous word containing a target and the following word containing a target by at least four filler words without the target letter or target phoneme. The critical words were not included in the first and the last sentence of the text, they were never at the beginning or the end of a sentence, and they were never adjacent to a punctuation mark. Furthermore, in the pour-cour text, each sentence with the function word pour was paired with a sentence including the content word cour, in the following fashion. In each sentence of a pair, the word preceding the target word held the same number of letters and belonged to the same syntactic category (function or content word). The same pairing procedure was applied to the word following the target word.

The texts were recorded in a professional studio by a professional radio reporter from Radio-Canada, the Canadian French public radio corporation. The narrator was instructed to read the texts in a neutral manner, and no information was provided about the purpose of the experiment. The recording of the des text lasted 3 min 11 s, and the recording of the pour-cour text lasted 5 min 6 s. Overall, a minimum of 929 ms separated the onset of each target phoneme. While adult participants rarely need practice with the reading (letter-detection) task, pilot testing revealed the need for some practice with the less familiar listening (phoneme-monitoring) task. Consequently, in addition to the two experimental texts, three practice sentences and one practice text were recorded for participants in the listening group. Each practice sentence was composed of only one word with the target phoneme. A different target phoneme was used per practice sentence (/k/, /l/, and /t/). The recording of each practice sentence lasted between 3 and 5 s. For the practice text, participants searched for the target phoneme /a/, which was embedded in 31 words. The recording lasted 2 min 22 s.

Procedure

Participants were assigned equally to the two tasks (n = 48 each), which they completed individually in a private room, in one session lasting approximately 15 min. In the reading task, participants received a stapled package with an instruction sheet on the front. The experimenter read the instructions aloud while participants read them silently. The instructions encouraged participants to read the texts for comprehension at their normal reading speed. They were told that whenever they came across a target letter (either in upper- or lowercase), they were to circle it. They were also instructed not to slow down their reading speed to detect all target letters and not to backtrack to circle a letter they had missed. Participants were also told that they would read two texts with a different target letter for each text, and that after each text, they would have to answer five multiple-choice comprehension questions. This procedure was used to promote reading for comprehension. To familiarize participants with this task, we asked them to read the instruction sheet again and to circle all ts.

In the listening task, instructions were presented on the computer screen. The experimenter read the instructions aloud while participants read them silently. After reading the instructions, participants pressed the space bar, and a blank screen appeared. The experimenter then said the target phoneme and asked participants to repeat it to ensure proper perception of the phoneme. If participants produced the wrong phoneme, the experimenter repeated it and asked them to produce it again. After participants successfully produced the phoneme, they pressed the space bar, and the narration of the first sentence began. This procedure was repeated separately for the three practice sentences, the practice text, and the two critical texts. Participants were asked to click a mouse immediately when they heard a target phoneme and were encouraged to respond as quickly and accurately as possible, as both speed and accuracy were being measured.

Data analysis

As is usually the case with the reading task, only misses were included as errors, because with the paper-and-pencil procedure, participants never circle a wrong letter, and for the listening task, it is impossible to attribute a false alarm to a particular presented phoneme. For both texts in the listening task, target phonemes were considered detected when the response occurred during the first 929 ms after the onset of the word. This criterion was chosen because it was the shortest interval before the onset of another word in which the critical phoneme was embedded. Consequently, it was the longest possible interval during which a response could unambiguously be attributed to the critical target-containing word. A similar criterion has been applied in the past with the rapid-serial-visual-presentation procedure and with eye-movement monitoring (e.g., Healy, Oliver, & McNamara, 1987; Roy-Charland et al., 2007; Saint-Aubin, Kenny, & Roy-Charland, 2010; Saint-Aubin & Klein, 2001).

Results

Figure 1 shows miss rates for the two passages in the reading and listening tasks, separately for function words and content words, as well as the effect sizes with 95% confidence intervals (CIs) for both passages in the two tasks. (More detailed results can be found in Tables S1 and S2 in the Supplemental Material available online.) As can be seen in Figure 1, the missing-letter effect for both passages was strong and positive. For example, in the reading task, targets in function words were missed more often than targets in content words. The primary and novel finding is that this was also true in the listening task. Moreover, this missing-phoneme effect was substantially larger than the missing-letter effect in both passages.

Fig. 1.

Estimated effect sizes (top row) and mean miss rate (bottom row). Effect sizes are shown for each task, separately for the des and pour-cour texts. Error bars are 95% confidence intervals calculated with a stratified bootstrap procedure. The mean miss rate is shown as a function of task and word type (critical function words vs. control content words), separately for the des and pour-cour texts.

For the listening task, it was also possible to compute response latencies between the onset of the word presentation and the button press indicating a detection. These analyses were based on 46 of 48 participants for the des text, because 2 participants did not detect the target phoneme in any of the function words (des), and on 35 of the 48 participants for the pour-cour text, because 13 participants did not detect the target phoneme in any of the function words (pour). As we expected (e.g., Saint-Aubin et al., 2003), results showed that participants took longer to detect a phoneme in the function word des (M = 622 ms) than in control words (M = 564 ms), t(45) = 5.60, Cohen’s d for repeated measures = 0.84, p < .0001, 95% CI for the difference between means = [36.93, 78.38]; similarly, they took longer to detect a phoneme in the function word pour (M = 674 ms) than in the content word cour (M = 523 ms), t(34) = 6.02, Cohen’s d for repeated measures = 1.04, p < .0001, 95% CI for the difference between means = [99.63, 201.31].² Two further analyses were conducted to increase our confidence in the similarity of these two effects.

We first conducted correlational analyses of miss rates across all the items in both passages. The correlation between items in the reading and listening tasks was significant for both the des text, r(46) = .60, p < .0001, and the pour-cour text, r(30) = .74, p < .0001 (see Fig. 2). We conducted a second follow-up analysis because we recognized that in a spoken passage, there might be acoustic differences between function words and content words. Because misses might be affected by the acoustic properties of the target phonemes, we first determined whether the two classes of words differed on a variety of acoustic properties. One property on which they differed significantly in both passages was duration: There were longer durations for content words (M = 0.319 s) than for function words (M = 0.148 s), t(78) = 7.96, Cohen’s d = 2.06, p < .0001, 95% CI for the difference between means = [0.128, 0.214]. Moreover, the raw correlation between word duration and miss rate was significant, r(78) = −.59, p < .0001.

Fig. 2.

Scatterplot (with best-fitting regression line) showing the relationship between mean miss rate on the listening task and mean miss rate on the reading task for each target-containing word in both des and the pour-cour texts. Unfilled circles indicate identical data from two different items.

Given this finding, we created and compared two models to determine whether the role of the word (function vs. content) had any explanatory power over and above word duration. To create the models, we used multilevel logistic regressions with both participants and items as random effects (Baayen, 2008); the Akaike information criterion (AIC) was the metric of model fit. With the AIC, only relative scores matter, with lower values indicating better model fit. Between models, AIC differences greater than 10 can be considered large (Symonds & Moussalli, 2011). In our analysis, the comparison model contained both word role and word duration as predictors. When word duration was removed, leaving only word role, the AIC changed very little (it was reduced by 1). This suggests that word duration barely explains variability over and above the extra complexity it adds to the model when just word role is present. Removing word role, leaving only word duration, raised the AIC by 46, a very substantial increase that suggests that word role is a much more important explanatory variable in the model than word duration. In short, a model with word role explained the misses about equally well whether or not word duration was included, but one with only word duration was much poorer. Therefore, even though word duration had a significant effect on misses, it did not account for our finding that phonemes of function words are missed substantially more than phonemes of content words.

Discussion

The results of the present study are clear and straightforward: We obtained a large missing-phoneme effect in the listening task, and it correlated well, on an item-by-item basis, with the missing-letter effect in the reading task. In addition to missing more phonemes in function than in content words, listeners also took longer to detect the target phoneme in the former than in the latter. This pattern of response latencies nicely replicates what has been observed in the reading domain (e.g., Roy-Charland et al., 2007; Saint-Aubin et al., 2003).

A strictly bottom-up visual hypothesis, the parafoveal-processing hypothesis (e.g., Greenberg et al., 2004; Hadley & Healy, 1991), has also been proposed to explain the missing-letter effect. This hypothesis posits that word-form information is picked up parafoveally, whereas letter identification is largely dependent on the detailed information provided by foveation of the word. In this view, the high rate at which function words are skipped can be due to their orthographic properties alone (Angele, Laishley, Rayner, & Liversedge, 2014; Angele & Rayner, 2013) or in conjunction with psycholinguistic information about the syntax and semantics of the passage one is reading (Koriat & Greenberg, 1991). Therefore, detecting a target letter would be difficult if the word were skipped. Hence, according to this hypothesis, a missing-letter effect is generated by the higher probability of skipping rather than fixating function words in reading (Carpenter & Just, 1983; Gautier, O’Regan, & Le Gargasson, 2000; O’Regan, 1979). The similar results we generated here when participants listened to a passage while searching for a phoneme, in conjunction with previous results we obtained in the reading domain (Roy-Charland et al., 2007; Saint-Aubin et al., 2003), encourage us to reject the parafoveal-processing hypotheses.

The missing-letter and missing-phoneme effects appear to have the same cause: Less attention is devoted to function than to content words when a person is reading a passage for comprehension. Because function words are more predictable (Drieghe, Rayner, & Pollatsek, 2005; Roy-Charland et al., 2007), their meanings are primed before the word is encountered. This implies that when attention is allocated to function words, the reader has a head start in processing them semantically. Attention, whether operating within the visual or auditory modality (Klein & Lawrence, 2011), disengages from function words sooner than from content words (Roy-Charland et al., 2007). Although not tested here, our model also assumes that attention is disengaged faster from frequent words than from rare words. It further assumes that information about the physical (visual or acoustic) properties of words accumulates when attention is focused on them. Consequently, faster attention disengagement means that physical information needed for the detection tasks is less accurately represented. Detections are assumed to be based on quality or strength of representation of the physical properties of each target, and these representations are assumed to decay once observers disengage from the target-containing word. In this attentional-disengagement model, more strongly activated target representations will generate a higher percentage of hits and faster detection times (see Roy-Charland et al., 2007, for a more complete presentation of this model and how it predicts the reaction time disadvantage for targets embedded in function words).

The remarkable similarity of findings across the two domains should not come as a surprise for researchers, considering that reading was built on listening processes. As a matter of fact, the results observed here are a logical extension of the ideas put forward by Koriat and Greenberg (1994) when they developed their structural account aimed at explaining the missing-letter effect. They described their hypothesis as follows:

It is our thesis that the processes underlying the comprehension of both spoken and written messages recapitulate the general architecture of speech production: The extraction of structure precedes and paves the way for the extraction of meaning. Essentially, both the listener and the reader strive to quickly establish a rudimentary frame-and-slot organization of the phrase or the sentence into which incoming units can be placed. (p. 346)

Having demonstrated the presence of a missing-phoneme effect in listening and its similarity to the missing-letter effect in reading, the current study opens the door for researchers to further explore the relationship between reading and listening.

Footnotes

Action Editor

Charles Hulme served as action editor for this article.

Declaration of Conflicting Interests

The authors declared that they had no conflicts of interest with respect to their authorship or the publication of this article.

Funding

This research was supported by discovery grants from the Natural Sciences and Engineering Research Council of Canada to J. Saint-Aubin and to R. M. Klein. While working on this research, M. Babineau was supported by an undergraduate student research award from the Natural Sciences and Engineering Research Council of Canada.

Supplemental Material

Additional supporting information can be found at

Notes

References

Angele

Laishley

A. E.

Rayner

Liversedge

S. P.

(2014). The effect of high- and low-frequency previews and sentential fit on word skipping. Journal of Experimental Psychology: Learning, Memory, and Cognition, 40, 1181–1203. doi:10.1037/a0036396

Angele

Rayner

(2013). Processing the in the parafovea: Are articles skipped automatically? Journal of Experimental Psychology: Learning, Memory, and Cognition, 39, 649–662. doi:10.1037/a0029294

Assink

E. H.

Knuijt

P. A.

(2000). Reading development and attention to letters in words. Contemporary Educational Psychology, 25, 347–362. doi:10.1006/ceps.1999.1010

Baayen

R. H.

(2008). Analyzing linguistic data: A practical introduction to statistics. New York, NY: Cambridge University Press.

Carpenter

P. A.

Just

M. A.

(1983). What your eyes do while your mind is reading. In Rayner

(Ed.), Eye movements in reading: Perceptual and language processes (pp. 275–307). New York, NY: Academic Press.

Connine

C. M.

Titone

(1996). Phoneme monitoring. Language and Cognitive Processes, 11, 635–645. doi:10.1080/016909696387042

Corcoran

D. W. J.

(1966). An acoustic factor in letter cancellation. Nature, 210, 658.

Cunningham

T. F.

Healy

A. F.

Kanengiser

Chizzick

Willitts

R. L.

(1988). Investigating the boundaries of reading units across ages and reading levels. Journal of Experimental Child Psychology, 45, 175–208. doi:10.1016/0022-0965(88)90029-X

Drewnowski

(1981). Missing -ing in reading: Developmental changes in reading units. Journal of Experimental Child Psychology, 31, 154–168. doi:10.1016/0022-0965(81)90009-6

10.

Drieghe

Rayner

Pollatsek

(2005). Eye movements and word skipping during reading revisited. Journal of Experimental Psychology: Human Perception and Performance, 31, 954–969. doi:10.1037/0096-1523.31.5.954

11.

Foss

D. J.

(1969). Decision processes during sentence comprehension: Effects of lexical item difficulty and position upon decision times. Journal of Verbal Learning and Verbal Behavior, 8, 457–462. doi:10.1016/S0022-5371(69)80089-7

12.

Gautier

O’Regan

J. K.

Le Gargasson

J. F.

(2000). “The-skipping” revisited in French: Programming saccades to skip the article “les.” Vision Research, 40, 2517–2531. doi:10.1016/S0042-6989(00)00089-4

13.

Greenberg

S. N.

Healy

A. F.

Koriat

Kreiner

(2004). The GO model: A reconsideration of the role of structural units in guiding and organizing text on line. Psychonomic Bulletin & Review, 11, 428–433. doi:10.3758/BF03196590

14.

Greenberg

S. N.

Inhoff

A. W.

Weger

U. W.

(2006). The impact of letter detection on eye movement patterns during reading: Reconsidering lexical analysis in connected text as a function of task. Quarterly Journal of Experimental Psychology, 59, 987–995. doi:10.1080/17470210600654776

15.

Greenberg

S. N.

Tai

(2001). Letter detection in very familiar texts. Memory & Cognition, 29, 1088–1095. doi:10.3758/BF03206377

16.

Guérard

Saint-Aubin

Poirier

Demetriou

(2012). Assessing the influence of letter position in reading normal and transposed texts using a letter detection task. Canadian Journal of Experimental Psychology, 66, 227–238. doi:10.1037/a0028494

17.

Hadley

J. A.

Healy

A. F.

(1991). When are reading units larger than the letter? Refinement of the unitization reading model. Journal of Experimental Psychology: Learning, Memory, and Cognition, 17, 1062–1073. doi:10.1037/0278-7393.17.6.1062

18.

Healy

A. F.

(1994). Letter detection: A window to unitization and other cognitive processes in reading text. Psychonomic Bulletin & Review, 1, 333–344. doi:10.3758/BF03213975

19.

Healy

A. F.

Oliver

W. L.

McNamara

T. P.

(1987). Detecting letters in continuous text: Effects of display size. Journal of Experimental Psychology: Human Perception and Performance, 13, 279–290. doi:10.1037/0096-1523.13.2.279

20.

Klein

R. M.

Lawrence

M. A.

(2011). The modes and domains of attention. In Posner

M. I.

(Ed.), Cognitive neuroscience of attention (2nd ed., pp. 11–28). New York, NY: Guilford Press.

21.

Klein

R. M.

Saint-Aubin

(in press). What a simple letter detection task can tell us about cognitive processes in reading. Current Directions in Psychological Science.

22.

Koriat

Greenberg

S. N.

(1991). Syntactic control of letter detection: Evidence from English and Hebrew nonwords. Journal of Experimental Psychology: Learning, Memory, and Cognition, 17, 1035–1050. doi:10.1037/0278-7393.17.6.1035

23.

Koriat

Greenberg

S. N.

(1994). The extraction of phrase structure during reading: Evidence from letter detection errors. Psychonomic Bulletin & Review, 1, 345–356. doi:10.3758/BF03213976

24.

Minkoff

S. R. B.

Raney

G. E.

(2000). Letter-detection errors in the word the: Word frequency versus syntactic structure. Scientific Studies of Reading, 4, 55–76. doi:10.1207/S1532799XSSR0401_5

25.

Moravcsik

J. E.

Healy

A. F.

(1995). Effect of meaning on letter detection. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 82–95. doi:10.1037/0278-7393.21.1.82

26.

New

Pallier

Brysbaert

Ferrand

(2004). Lexique 2: A new French lexical database. Behavior Research Methods, Instruments, & Computers, 36, 516–524. doi:10.3758/BF03195598

27.

O’Regan

(1979). Saccade size control in reading: Evidence for the linguistic control hypothesis. Perception & Psychophysics, 25, 501–509. doi:10.3758/BF03213829

28.

Read

J. D.

(1983). Detection of Fs in a single statement: The role of phonetic recoding. Memory & Cognition, 11, 390–399. doi:10.3758/BF03202454

29.

Roy-Charland

Saint-Aubin

(2006). The interaction of word frequency and word class: A test of the GO model’s account of the missing-letter effect. Quarterly Journal of Experimental Psychology, 59, 38–45. doi:10.1080/17470210500269428

30.

Roy-Charland

Saint-Aubin

Klein

R. M.

Lawrence

(2007). Eye movements as direct tests of the GO model for the missing-letter effect. Perception & Psychophysics, 69, 324–337. doi:10.3758/BF03193753

31.

Saint-Aubin

Kenny

Roy-Charland

(2010). The role of eye movements in the missing-letter effect revisited with the rapid serial visual presentation procedure. Canadian Journal of Experimental Psychology, 64, 47–52. doi:10.1037/a0016850

32.

Saint-Aubin

Klein

R. M.

(2001). Influence of parafoveal processing on the missing-letter effect. Journal of Experimental Psychology: Human Perception and Performance, 27, 318–334. doi:10.1037//0096-1523.27.2.318

33.

Saint-Aubin

Klein

R. M.

(2008). The influence of reading skills on the missing-letter effect among elementary school students. Reading Research Quarterly, 43, 132–146. doi:10.1598/RRQ.43.2.2

34.

Saint-Aubin

Klein

R. M.

Landry

(2005). Age chances in the missing-letter effect revisited. Journal of Experimental Child Psychology, 91, 158–182. doi:10.1016/j.jecp.2005.01.007

35.

Saint-Aubin

Klein

R. M.

Roy-Charland

(2003). Direct assessments of the processing time hypothesis for the missing-letter effect. Journal of Experimental Psychology: Human Perception and Performance, 29, 1191–1210. doi:10.1037/0096-1523.29.6.1191

36.

Saint-Aubin

Poirier

(1997). The influence of word function in the missing-letter effect: Further evidence from French. Memory & Cognition, 25, 666–676. doi:10.3758/BF03211308

37.

Saint-Aubin

Roy-Charland

Klein

R. M.

(2007). The influence of multiple readings on the missing-letter effect revisited. Memory & Cognition, 35, 1578–1587. doi:10.3758/BF03193492

38.

Schneider

V. I.

Healy

A. F.

Ericsson

K. A.

Bourne

L. E.

(1989). Letter detection errors in reading, auditory, and memory tasks. Journal of Memory and Language, 28, 400–411. doi:10.1016/0749-596X(89)90019-3

39.

Schneider

V. I.

Healy

A. F.

Gesi

(1991). The role of phonetic processes in letter detection: A reevaluation. Journal of Memory and Language, 30, 294–318. doi:10.1016/0749-596X(91)90038-L

40.

Symonds

M. R. R.

Moussalli

(2011). A brief guide to model selection, multimodel inference and model averaging in behavioural ecology using Akaike’s information criterion. Behavioral Ecology and Sociobiology, 65, 13–21. doi:10.1007/s00265-010-1037-6

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.09 MB