Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636  
1 
 
EYE MOVEMENTS IN ORAL READING BY PRIMARY-SCHOOL 
STUDENTS: THE RELEVANCE OF PUNCTUATION MARKS 
Movimientos oculares en la lectura oral de estudiantes de primaria: 
la relevancia de los signos de puntuación 
 
 
Nair Daiane de Souza Sauaia Vansiler *   
Angela Ines Klein **   
Romeu Miqueias Szmoski ***   
  
ABSTRACT 
This study aims to explore the relevance of punctuation marks during oral reading. To achieve this objective, we compared 
eye movement metrics of two  groups of 5th-year students, classified based on the Reading  Fluency variable. The fluent 
reading  group  (FL)  consisted  of  27  participants,  while  31  participants  were  in  the  poor  reading  fluent  group  (PF).  All 
participants  read three  types of  text:  a  simple  text  with punctuation, a  complex  text with punctuation  and a  simple text 
without punctuation. By analyzing the fixations and saccades, the data suggest that the complexity of the text influences the 
eye-tracking metrics of the PF group, but not the FL group. On the other hand, the presence of punctuation influence the 
reading behavior of the FL group, but not the PF group. These findings highlight the influence of punctuation and textual 
complexity  on  the  dynamics  of  oral  reading  and  offer  valuable  insights  into  how  reading  fluency  is  related  to  text 
comprehention. 
Keywords: oral reading, eye movements, punctuation marks. 
 RESUMEN 
Este estudio tiene como objetivo explorar la relevancia de los signos de puntuación durante la lectura oral. Para lograr este 
objetivo, comparamos métricas de movimiento ocular de dos grupos de estudiantes de 5º año, clasificados en función de la 
variable Fluidez Lectora. El grupo de lectura fluida (FL) estuvo formado por 27 participantes, mientras que 31 participantes 
estaban en el grupo de lectura fluida deficiente (PF). Todos los participantes leyeron tres tipos de texto: un texto simple con 
puntuación, un texto complejo con puntuación y un texto simple sin puntuación. Al analizar las fijaciones y movimientos 
sacádicos, los  datos sugieren que la complejidad del texto influye en las métricas de seguimiento ocular del grupo PF, pero 
no en el grupo FL. Por otro lado, la presencia de puntuación influye en la conducta lectora del grupo FL, pero no del grupo 
PF. Estos hallazgos resaltan la influencia de la puntuación y la complejidad textual en la dinámica de la lectura oral y ofrecen 
información valiosa sobre cómo la fluidez lectora se relaciona con la comprensión del texto. 
Palabras clave: lectura oral, movimientos oculares, signos de puntuación. 
 
 
 
 
 
 
*  Federal  University  of  Pará.  Belém,  Brazil.  PhD  in  Linguistic  Studies.  Email:  nairsauaia@gmail.com.  
ORCID: http://orcid.org/0000-0003-4947-738X  
**  Federal  University  of  Pelotas.  Pelotas,  Brazil.  PhD  in  Linguistics.  Email:  angela.klein@ufpel.edu.br.  
ORCID: http://orcid.org/0000-0001-6230-7938 
***  Federal  Technological  University  of  Paraná.  Ponta  Grossa,  Brazil.  PhD  in  Sciences  –  Physics.  
Email: rmszmoski@utfpr.edu.br. ORCID: https://orcid.org/0000-0002-0968-7158  
DOI: https://doi.org/10.15517/rk.v48i3.62716  
Recepción: 1/12/2023         Aceptación: 18/4/2024 
 

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

1. Introduction

During reading, our eyes are always moving and, due to the narrowness of the fovea area,

these movements are ceaseless, resulting in small eye movements to bring the words of a text to this

area, for example. Although the accuracy of visual information is limited to the area of the fovea, it is

known that the eye begins the process of identifying a word even before the eyes fixate it (Rayner,

1998), meaning that information from the parafoveal area is very important during reading. Figure 1

illustrates the fixation under the word “reads” during the reading of a text, in other words, the clearest

word represents the foveal area, while the other words, the imprecise ones, represent the parafoveal

area.

Figure 1.

Text reading - foveal and parafoveal region

Source: Dehaene (2012, p. 28).

Taking into consideration these characteristics of the visual field, during the processing of eye

movements in oral reading and their interaction with punctuation marks in writing, readers pause in

syntactically strategic places for the integration of information; this processing intensifies when there

is punctuation, causing an increase in reading times at syntactic or discursive boundaries (Hirotani et

al., 2006). Such an event has been called a “wrap-up” and is indicative of mental processes of updating

an interpretation that is still in progress but needs to be completed (Rayner, 1998). Hirotani et al.

(2006) state that “Readers clearly exploit punctuation and use it to guide their eye movements in ways

parallel to those in which listeners use intonation” (p. 439). Thus, evidence is compiled that

punctuation interferes with eye behavior during reading.

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

Based on this evidence, this paper was designed, which focuses on the processing of

punctuation marks during oral reading by 5th-grade students, taking into consideration textual

complexity and reading fluency.

2. Eye Movements in Reading

In this section, we will present some eye aspects during reading and their implications for

processing, especially in young readers. The eye movements highlighted here are fixations and

saccades.

2.1 Fixations

Fixations represent the moment in which the eyes pause slightly under a certain area for

processing. The number of fixations is related to the difficulty in processing information (Forster,

2017), meaning that the more fixations made on a given word, the greater the difficulty in processing

said word (Rayner, 2014). The variables analyzed in reading studies that consider fixations are:

location, quantity and duration of fixations.

The fixations made during the reading are counted in milliseconds (ms): the larger the circle,

the longer the fixation duration. Figure 2 shows the fixations graphically in blue circles using the

BeGaze software.

Figure 2.

Fixations and saccades displayed by the BeGaze software

Source: the authors.

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

Studies (Just & Carpenter, 1987; Rayner & Duffy, 1986) have concluded that content words

are fixated 85% of the time, while functional words are fixated only 35% of the time. Furthermore,

according to these studies, this is due to the length of functional words, which tend to be short, that is,

the longer the word, the more likely it is to be fixated. In practical terms, a word with 2-3 letters can be

fixated around 25% of the time, while a word with 8 letters is more likely to be fixated, and even re-

fixated. Also, keywords for comprehension receive more fixations from fluent readers, which allows

them to process more quickly. In some cases, some words do not receive fixations during processing,

which is known as the “word skipping” phenomenon, because the word is “skipped” by means of a

longer saccade.

It is known that in terms of perceptual area, beginner readers display smaller letter spaces to

the right of fixation, around 12 spaces, when compared to adult readers, who display 14-15 spaces.

The size of the perceptual area also varies according to the difficulty of the text in both age groups

(Rayner & Duffy, 1986).

When it comes to identifying words, the time needed to identify a word is shorter when the

eyes initially fixate on the middle of the word, an event known as Optimal Viewing Position (OVP).

For Rayner (1978), the first landing position is classified as Preferred Viewing Location (PVL). It is

therefore considered that the OVP is the optimal position in terms of word perception, while the PVL

represents the actual position of the fixation when reading a word in the sentence (LIU & LI, 2013).

2.2 Saccades

Rayner (2014) defines saccades, or saccadic movements, as the rapid eye movements from

one fixation to the next while scanning and processing information. Examples of saccades are shown

as blue lines in Figure 2. Their length represents the amplitude of the saccadic movements, that is, the

distance from saccade start to end point.

In reading, saccades generally last from 20 to 40 milliseconds and span, on average, 7 to 9

letters, though typically within the range from 1 to 18 letters (Rayner, 2014). Aspects commonly

analyzed in these movements are duration, location and amplitude of the movements.

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

The saccades can be progressive or regressive. When reading Portuguese, where you move to

the right, the progressive will be the one made most of the time. Regressive saccades are movements

made in the opposite direction to reading; in the case of Portuguese, from right to left. According to

Rayner (2014) regressions are linked to comprehension difficulties.

Regressions occur 10 to 15% of the time during reading (Rayner, 2014; Macedo, 2007). For

Macedo (2007), it serves as a way of checking whether a word has been skipped or not understood.

Many regressions are only a few letters long and can mean that the reader has made a very long

saccade, making them need to go back to that information in order to process it; on the other hand, a

short regressive saccade can occur to solve a processing problem in the word that is being fixated in

the act of regression, for example, in the resumption of one of the syllables of an unknown or very

long word.

When compared to fluent adult readers, children usually make shorter saccades, as well as

longer fixations, and around 25% of their eye movements consist of regressions, which Blythe and

Joseph (2011) believe explains why reading is slower in this group.

3. Eye Movements in Oral Reading

Eye movements have distinct characteristics that depend on the type of reading. When reading

aloud, for example, even if you are only following a text that another person is reading aloud with

your eyes, the average duration of fixations is longer than when reading silently. This is due to the

tendency for the eyes to be ahead of the voice. In a study carried out by Lévy-Schoen (1981),

occurrences of refixations on a word were observed, almost as if the eyes were still. The intention of

the readers, who were following oral reading, was not to go too far ahead of the heard reading.

In general, what is known about oral reading and eye movements is that articulation in

oralization operates more slowly than cognitive processes. As a result of this time difference, readers

usually look at and process a word in the text that is two to three words to the right of what they are

pronouncing (Inhoff et al., 2011): “Given that the eyes are ahead of the voice in oral reading, readers

may only need to process the fixated word, and words to the right of that might be irrelevant” (p. 635).

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

Huey (1968) suggests a fairly simple relationship between silent and oral reading

modes; for him, oral reading processes can be considered essentially the same as silent reading

processes, with only the addition of articulatory demands. However, studies that have

examined eye movements in oral and silent reading in-depth indicate that eye movements in

both modes are strongly correlated (Anderson & Swanson, 1937; Søvik et al., 2000).

For Rayner et al. (2012), examining the demand on tasks in the parafoveal region and

analyzing the reading systems that capture information from this region during oral reading can make

important contributions to understanding oral and silent reading processes. The authors investigated

the role of parafoveal information in oral reading compared to silent reading. Intrigued by the

possibility of parafoveal information masking the delay in silent reading, the authors examined

whether the availability of parafoveal information affected oral reading. To this end, the authors

monitored the eye movements of participants reading sentences silently and orally under conditions in

which parafoveal information was sometimes available and sometimes not. The results indicate that

parafoveal information contributes directly to a reduction in fixation duration during both oral and

silent reading. On the other hand, the “number of fixations” in a sentence was reduced when

parafoveal information was available during silent reading, which was not the case during oral

reading. In other words, the data indicates that the availability of parafoveal information only helps

silent reading. The authors offer three explanations for this: firstly, following the results of another

study on the phonological representation of words in working memory

during sentence reading

(Inhoff, et al., 2004), the available resources for parafoveal processing are depleted as the

phonological working memory is required more, since words need to be stored in memory until they

are articulated. The second explanation is related to the phonological representations stored in working

memory for articulation, which could interfere with more recently activated phonological

representations and thus reduce the benefit of parafoveal information during reading. The third

explanation is that parafoveal information is treated similarly in the two reading modes, but readers do

not make use of this information to program fixation location during oral reading.

The working memory implies the temporary storage and manipulation of information that is considered

necessary for a wide range of complex cognitive activities (Baddeley, 2003).

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

4. Eye Movements and Punctuation Marks

The pause effect related to punctuation marks in reading has typically been attributed to

unfinished interpretive processing and the updating of discourse representation, according to Rayner et

al. (2012).

Aiming to further understand when and why closure effects occur, Hirotani et al. (2006)

investigated the effects of punctuation in clauses and sentences. For the authors, punctuation marks are

linked to the permanence hypothesis, through which the deceleration at the end of clauses is seen as

strategic. The authors explain that readers, in order to avoid returning to the beginning of the clause,

remain at the end of the clause until they have processed all the events contained therein, and then

move on to the next clause. This hypothesis is called “dwell-time”. This hypothesis characterizes

longer eye fixation durations at the end of sentences. According to the authors, when we compare the

number of regressions within a clause and at the end of it, especially when there is a punctuation mark

at the end of the clause, this hypothesis predicts that readers make fewer regressions at the end of

sentences than within the clause. Thus, characterizing reading eye movements in relation to

punctuation marks and their importance for sentence organization and, consequently, textual

comprehension.

Analyzing the impact on eye patterns when reading texts in Spanish, where the use of commas

is mandatory, Angele et al. (2024) organized an experiment with sentences with or without mandatory

commas. The data shows that even when commas are mandatory, their effect is primarily to shift when

processing takes place rather than to facilitate processing overall. This because:

There was a local increase in the go-past time for the pre-comma region when commas were

presented, which was balanced out by shorter first-pass and second-pass times on the

subsequent regions. In global sentence reading time, there was no evidence for an advantage

of presenting commas. (Angele et al., 2024, p.1)

So far, there has been a discussion about eye movements in reading, especially in oral reading,

and finally, there has been a relationship between some characteristics of eye movements and the

presence of textual punctuation marks. From this, the questions arise: what is the influence of

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

punctuation marks during oral reading? Does textual complexity play a role in differentiating these

characteristics? And what are the differences in reading processing between individuals with different

reading fluencies? In order to answer these questions, in the following sections we present an

investigation into the processing of eye movements during oral reading by elementary school students,

taking into account textual complexity (simple text, complex text, with or without punctuation marks)

and reading fluency (poor reading fluent group and fluent group).

5. Sample

The sample for this study consisted of 58 fifth grade students from public schools in the city of

Ponta Grossa, Paraná, Brazil. Using the Multidimensional Scale for the Assessment of Expressive

Reading

(Vansiler, 2015; Picanço & Vansiler, 2014) the participants were divided into two groups:

the fluent reading group (FL) and the poor reading fluent group (PF), as shown in Table 1.

This scale assesses oral reading in the following dimensions: intonation and emphasis,

phrasing, and fluidity of rhythm. Depending on the reading performance, the reader receives scores

from 0 to 4, with 0 referring to the lowest score. Based on the average between the dimensions, the

reader's level of fluency in oral expressiveness is measured: non-fluent (PF) and fluent (FL).

Table 1.

Mean and SD of Reading Score per Group

Fluency-Group

boys

girls

Total

Source: Vansiler (2021, p. 102).

The project was submitted to the ethics committee of Plataforma Brasil and was approved

under #2.757.810, CAAE: 86296517.8.1001.0018 on 05/06/2018.

“Escala Multidimensional para avaliação da leitura expressiva”. Portuguese version available.

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

6. Space Characterization and Technical Specifications

The data was collected at the Visual Processing Laboratory (LabPV), located at UTFPR in

Ponta Grossa, Brazil. The RED 500 device from German manufacturer SMI detected eye movements

while the participants read the texts. The device is sensitive enough to register any binocular

movement every 2 milliseconds and has an accuracy of 0.4. It consists of an infrared set, a projection

screen and a notebook on which there is SMI software for recording eye movements, setting up the

experiment and processing the data. The infrared is located below the monitor and is positioned at a

distance of 60 to 80 cm from the participant. Before the participant began the collection, the height of

the table was adjusted so that the participant was in a comfortable position during the experiment.

Calibration was carried out with 5 points with a limit of 0.5º on the x and y axes. For each deviation

found above this limit, the calibrations were repeated and validated in order to continue with the

experiment.

7. Description and Application of the Instrument

The data was collected individually in a quiet room at the LabPV. Each participant sat in front

of the monitor with the tracker attached, resting the upper front of their head on a support to reduce

sudden movements during the experiment and read aloud, that is, orally. The collection instrument

consists of three texts which vary according to textual complexity and the presence or absence of

punctuation marks.

Text 1 (Figure 3) corresponded to a complex 49-word school text, the content of which

features sentence inversion, less frequent words in the Portuguese language and is highly informative,

but the vocabulary is accessible at a school level. The punctuation marks are visible in the text.

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

Figure 3.

Screen of the Experiment Center with the Text “Matrizes culturais do Brasil”

(Complex Condition with Punctuation).

Source: elaborated by the authors.

Text 2 is composed of a 37-word non-complex school text and also has semantic content

accessible to the level of the students, punctuated according to the norms of the written Portuguese

language. Figure 4 shows the screen with the text.

Figure 4.

Screen of the Experiment Center with the Text “Nunca, nuncaquinha” (Easy Condition with Punctuation)

Source: elaborated by the authors.

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

Text 3 is also a non-complex text with 34 words, but part of the text has no punctuation marks.

Figure 5 shows the image of stimulus-text 3.

Figure 5.

Screen of the Experiment Center with the Text “A herança” (Easy Condition without Punctuation)

Source: elaborated by the authors.

Questions were asked at the end of each text to ensure that the students were reading carefully.

However, these questions and answers were not taken into account when analyzing the data.

8. Analysis of the eye movement data

Considering the three stimuli, quantitative analyses were carried out to check: a) Number of

Fixations (NF) and the b) Number of Saccades (NS). To find out whether or not the two eye-tracking

metric mentioned followed a normal distribution, the Shapiro-Wilk test was used, according to which,

a distribution is considered normal if the value of the parameter p obtained from the test is greater than

0.05. Table 2 displays the test result for the three stimuli considered.

Text commonly used in the classroom to work on punctuation marks. For an example of how to

work in class, see: https://brainly.com.br/tarefa/8181984

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

Table 2.

Classification of Normality of Eye-Tracking Metric Distribution

Stimulus (S)

Number of fixations (NF)

Number of saccades (NS)

S1 - PF

Normal

S1 - FL

Non-normal

S2 - PF

Normal

Non-normal

S2 - FL

Normal

Non-normal

S3 - PF

Normal

S3 - FL

Normal

Non-normal

Source: elaborated by the authors.

To compare the distribution of eye-tracking metrics for each stimulus and for each group, we

used the Mann-Whiteney statistical test. The Mann-Whitney test is a non-parametric test used to

compare two independent samples and determine whether there is a significant difference between

them. It is considered the non-parametric equivalent to the t test; however, it does not require the data

to be normal.

When the Mann-Whitney U test is applied to two independent samples, one of the results is

the U parameter, which indicates the magnitude of the difference between the two samples. A high U

value suggests that one of the samples tends to have higher values than the other, while a low U value

suggests the opposite. In addition to the U parameter, the test provides the p-value, which indicates the

probability of observing a difference between the samples. If the p-value is less than a certain pre-

defined significance level (e.g. 0.05), the null hypothesis that there is no difference between the

samples is rejected.

With the Mann-Whitney test it is also possible to calculate the z-score and the effect size r.

The z-score corresponds to the normalized value of U. It indicates how many standard deviations the

U value is away from the mean. A high Z-score (in absolute value) indicates that the difference

between samples is greater than would be expected by chance. In turn, the effect size r indicates the

magnitude of the difference between the samples. It is obtained by dividing the z-score by the square

root of the total number of observations. Knowing these parameters provides a comprehensive view of

the difference between the two samples and whether this difference is statistically significant.

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

The results of applying the Mann-Whitney test as well as the box-plot of the eye-tracking

metrics are described below and are presented by stimulus, comparing the two fluency groups and the

three stimuli.

8.1 Analysis of Stimulus 1: Complex Condition with Punctuation

In Figure 6, we present the boxplot for the two eye-tracking metrics we've been discussing, the

number of fixations and the number of saccades. As we can observe, the PF group demonstrated a

higher number of fixations and saccades compared to the FL group. Considering the mean and median

of fixations, the PF group had 306.2 and 332 fixations, respectively, while the FL group had 197.5 and

159 fixations. Regarding saccades, the mean and median of the PF group were, respectively, 517.2 and

488, while that of the FL group was 285.6 and 277.

Figure 6.

Boxplot of NF and NS by Group FL and PF for Stimulus 1

Source: elaborated by the authors.

In order to verify whether this difference between the behavior of the ocular variables of the

two groups is significant, the Mann-Whitney Test was employed. The Table 3 displays the test result

obtained for the two metrics considered. Since the U value is high and the p value < 0.05, it can be

stated that there is a significant difference between the groups.

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

Table 3.

Result of the Mann-Whitney test in relation to eye-tracking metrics for stimulus 1

Eye-tracking metrics

U (PF)

U(FL)

Number of fixations

417

9.09E-05

3.742

0.557

Number of saccades

44.5

435.5

2.15E-06

4.596

0.693

Source: elaborated by the authors.

Looking at Figure 6 and Table 3, the PF and FL groups differ significantly, with the FL group

having lower averages in all ocular variables in stimulus 1. As this stimulus corresponds to a complex

text, the result indicates that textual complexity directly interferes with eye behavior during oral

reading, causing participants with low fluency to need to make more saccades and more fixations.

Therefore, ocular variables could be used as parameters to classify reading fluency when the stimulus

involves textual complexity.

8.2 Analysis of Stimulus 2: Easy Condition with Punctuation

Figure 7 shows the boxplot of the fixations and saccades of stimulus 02 for each fluency

group. Also in this stimulus, the PF group exhibited a higher number of fixations and saccades than

the FL group. Regarding the number of fixations, the first had a mean and median equal to 202.9 and

195, respectively, while the second, 157.4 and 150. Also, in relation to the saccades, the mean and

median of the PF group (332.1 and 297) performed a slightly higher than the FL group (240.8 and

212).

Figure 7.

Boxplot of NF, NS for FL and PF in Stimulus 2

Source: elaborated by the authors.

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

As with stimulus 1, the Mann-Whitney Test was employed and the results (high U and p value

< 0.05) suggest that there is a significant difference between the groups. Table 4 displays the test result

obtained for each of the test parameters.

Table 4.

Results Of the Mann-Whitney Test in Relation to Eye-Tracking Metrics for Stimulus 2

Eye-tracking metrics

U (PF)

U(FL)

Number of fixations

120,5

455.5

2.86E-04

3.44

0.497

Number of saccades

148

428

2.01E-03

2.877

0.415

Source: elaborated by the authors.

The results of the eye movements were also statistically different between the fluency groups

in stimulus 2, which involved a non-complex text. Participants in the fluent group made significantly

fewer fixations and saccades, meaning that the performance of low-fluency readers was once again

lower, even though this stimulus was made up of a non-complex text.

8.3 Analysis of Stimulus 3: Easy Condition without Punctuation

The behavior observed in the previous texts for the number of fixations of the PF groups was

also observed in text 3. The figure 8 shows the boxplot containing the mean and median measurements

of the ocular variables observed in stimulus 3 for each fluency group.

Figure 8.

Boxplot of NF and NS for FL and PF in Stimulus 3

Source: elaborated by the authors.

Also in stimulus 3, the number of fixations and saccades in the PF group were higher than in

the FL group. While the mean and median number of fixations for the first were, respectively, 193.2

and 192, for the second they were 134.8 and 136. Regarding saccades, the mean and median were

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

respectively 276.5 and 264 for the PF group, and 200.5 and 183.5 for the FL group. Table 5 shows the

result of the statistical test for the ocular variables in stimulus 3.

Table 5.

Result of the Mann-Whitney Test in Relation to Eye-Tracking Metrics for Stimulus 3.

Eye-tracking metrics

U (PF)

U(FL)

Number of fixations

133

673

7.86E-06

4.318

0.571

Number of saccades

179.5

626.5

1.76E-04

3.573

0.473

Source: elaborated by the authors.

As in the previous stimuli, also here the mean of the eye magnitudes investigated was lower

for the FL group compared to the PF group. However, for this stimulus the difference between the

means was smaller than in the previous stimuli, even so the difference between the groups was

statistically significant.

8.4 Analysis of Stimulus 1 X Stimulus 2

After comparing the eye-tracking metrics in relation to the groups, they were also compared in

relation to the stimulus. Table 6 displays the result of applying the Mann-Whitney Test comparing the

number of fixation and number of saccades in the complex text with punctuation (S1) with those in the

easy text with punctuation (S2).

Table 6.

Results of the Mann-Whitney Test for Each Group Comparing Stimuli 1 and 2

Eye-tracking metrics

U (S1)

U(S2)

Number of fixations (PF)

60.5

443.5

6.92E-06

4.346

0.647

Number of fixations (FL)

232.5

343.5

1.28E-01

1.135

0.163

Number of saccades (PF)

81.5

422.5

5.47E-05

3.869

0.577

Number of saccades (FL)

241.5

334.5

1.71E-01

0.949

0.137

Source: elaborated by the authors.

As the p values obtained for the PF group were lower than 0.05 for both eye-tracking metrics,

it can be concluded that there is a significant difference between the group's behavior in relation to the

complex text with punctuation and the simple text with punctuation. On the other hand, p values

greater than 0.05 obtained for the FL group reveal that there is no significant difference for fixations

and saccades in this group. These results suggest that the complexity of the text influences the eye-

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

tracking metrics of the PF group, but not the FL group, which may have implications for reading

comprehension and cognitive processing. However, more research is needed to explore these

relationships in different contexts and populations.

8.5 Analysis of Stimulus 2 X Stimulus 3

The PF and FL groups were also analyzed separately in relation to stimuli 2 and 3, which refer

to simple texts with and without punctuation, respectively. Table 7 displays the result of applying the

Mann-Whitney Test, comparing the number of fixation as well as the number of saccades in relation to

two simple texts, one with punctuation and the other without.

Table 7.

Results of the Mann-Whitney test for each group comparing stimuli 2 and 3

Eye-tracking metrics

U (S1)

U(S2)

Number of fixations (PF)

342.5

401.5

3.11E-01

0.492

0.066

Number of fixations (FL)

199.5

424.5

1.48E-02

2.175

0.308

Number of saccades (PF)

276.5

467.5

5.34E-02

1.613

0.217

Number of saccades (FL)

212.5

183.5

2.92E-02

1.893

0.268

Source: Elaborated by the authors.

Unlike what was observed in relation to complexity, analysis of the data in table 07 reveals

some interesting observations about the influence of punctuation on reading, especially when we

consider semantics. Semantics is heavily influenced by punctuation. Punctuation helps define the

structure and meaning of sentences, indicating where pauses should occur and how different parts of

the sentence relate to each other.

In the case of the FL group, we observed significant differences in the number of fixations and

the number of saccades between simple texts with and without punctuation. This suggests that the

presence of punctuation, which contributes to the semantics of the text, can influence the reading

behavior of this group. Punctuation can help guide the reader through the text, making it easier to

understand the sentence structure and therefore the overall comprehension of the text.

On the other hand, for the PF group, we did not observe significant differences between

fixations or saccades for simple texts with and without punctuation. This may indicate that, for this

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

group, the punctuation has a lesser impact on reading. Perhaps these readers are more focused on the

meaning of each word in the text and less on the meaning of the sentence or paragraph.

These findings highlight the importance of punctuation and semantics in reading and

understanding text. However, they also show that the influence of the punctuation can vary depending

on the reader, possibly due to differences in reading skills, experience, or reading strategies.

9. Discussion

The results of the ocular data analysis presented in the previous section show that the low

fluency group had higher values for all eye-tracking metrics than the fluent group for all stimuli. This

corroborates the findings of Forster (2017) and Rayner (2014) that the number of fixations is related to

difficulty in processing information. These findings suggest that readers with lower fluency may spend

more time fixating on words and make more saccades, or quick eye movements, as they read. This

could be because they are spending more time processing the information in the text, which could be

due to a variety of factors such as lower reading speed, less familiarity with the language, or more

difficulty understanding the content.

It’s also interesting to note that these patterns were observed across all stimuli, suggesting that

these differences in eye-tracking metrics are not just due to the complexity or difficulty of specific

texts, but are likely reflective of broader differences in reading ability or strategy between the two

groups. However, when comparing the values that differentiate the groups from each other, it can be

seen that textual complexity directly interfered with the scores: in the complex text, the PF participants

were much more distant from the ocular behavior of the FLs. And this need is more evident in the FP

participants.

The textual and typographical characteristics also influenced the eye movements of the two

fluency groups: although both stimuli 1 and 2 were composed of informative text, the first text

contained formal language and the second colloquial. As colloquial language is simpler, this made it

easier to understand the text in stimulus 2, reducing the difference in means between the PF and FL

groups.

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

In the text without punctuation marks, which is stimulus 3, the difference between the values

of the two groups in the ocular aspects is much smaller. It can be understood from these results that the

text without punctuation was difficult to understand for both groups, even though it was a simple text.

The participants had to choose where to pause, giving their own interpretation to the text. Depending

on where the pause was (often marked by a punctuation mark), the nephew and the tailor would

receive an inheritance or not. Deciding on one interpretation or another required more fixations in both

groups, corroborating the findings of Rayner et al. (2006) that “eye movements are sensitive to global

text difficulty” (p. 251). In native Spanish fluent readers, when they reading texts without commas,

they had fewer fixations, and according to the authors, skilled readers could easily analyze the

syntactic structure and semantic relationships in a sentence, even in the absence of commas, at least

for syntactically simple sentences (Angele et al., 2024).

Considering the discussion so far, the data shows that the presence of punctuation marks in

texts during oral reading is of utmost importance, even surpassing the level of reading fluency and

textual complexity. In the two previous stimuli, where the difference was textual complexity, the

fluency groups also differed significantly.

10. Final considerations

In order to find out the relevance of punctuation marks during oral reading, this research

analyzed ocular aspects (number of fixations, number of saccades) during oral reading of 58 fifth

grade students from a school in Brazil, taking into account the reading fluency variable (fluent group

and low fluent group) and textual complexity (simple text with punctuation, complex text with

punctuation and simple text without punctuation).

In response to the questions that motivated this work −such as: a) what role do punctuation

marks play during reading?−, we can see that the punctuation marks during oral reading in the context

studied here played a role in textual organization for both fluency groups, since in the text without

punctuation, the two fluency groups showed closer eye movements, even so the difference between the

groups was statistically significant. Regarding the second question −b) Does textual complexity play a

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

role in differentiating these characteristics?−, it was possible to observe that the fluency groups

differed significantly in their eye behavior from a complex text to a simple one, with the fluent group

showing lower averages; finally, in relation to the third question −c) What about reading fluency, what

are the differences in reading processing?−, it is noteworthy that the low fluency group had the highest

averages in all the stimuli and in all the ocular variables.

From continuous writing to the addition of punctuation marks in a text, our brain has adapted

synaptic areas for the evolution of its species that guarantee it will pass on to future generations one of

its greatest inventions: writing. There is no doubt that there is a place in the brain for the processing of

writing. Just as there is no doubt that a reader's brain has different physical and chemical

characteristics to a non-reader's brain. Skills with printed material (we're even broadening the concept

of reading here, because everything that is printed is processed: reading the world, reading

impressions, reading expressions, etc.) are acquired as the reader enters and is introduced to the world

of writing.

It is necessary to stimulate the skills of memory, oralization, expressiveness, knowledge of the

world and many others related to the ability to read, in order to guarantee the best performance from

students. Developing this skill means ensuring that these citizens are actively included in society.

References

Anderson, I. H., & Swanson, D. E. (1937). Common factors in eye-movements in silent and oral

reading. Psychological Monographs, 48, 61–77. https://doi.org/10.1037/h0093393

Angele, B., Gutiérrez-Cordero, I., & Marcete, A. (2024). Reading (,) with and without commas.

Quarterly Journal of Experimental Psychology, 77(6), 1190-1200.

https://doi.org/10.1177/17470218231200338

Baddeley, A. (2003). Working memory and language: an overview. J Commun Disord, May-

Jun;36(3):189-208. https://doi.org/10.1016/s0021-9924(03)00019-4

Blythe, H. I., & Joseph, H. S. S. L. (2011). Children's eye movements during reading. In S. P.

Liversedge, I. D. Gilchrist, & S. Everling (Eds.), The Oxford handbook of eye movements (pp.

643–662). Oxford University Press.

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

Dehaene, S. (2012). Osneurônios da leitura: Como a ciência explica nossa capacidade de ler. Penso.

Forster, R. (2017). Aspectos da utilização do rastreamento ocular na pesquisa psicolinguística.

DELTA: Documentação e Estudos em Linguística Teórica e Aplicada, 33(2): 609-644.

https://doi.org/10.1590/0102-445095461720767529

Hirotani, M., Frazier, L., & Rayner, K. (2006). Punctuation and intonation effects on clause and

sentence wrap-up: Evidence from eye movements. Journal of Memory and Language, 54(3),

425-443. https://doi.org/10.1016/j.jml.2005.12.001

Huey, E. B. (1968 [1908].). Psychology and Pedagogy of Reading. Cambridge, MA: MIT Press.

Just, M.A. & Carpenter, P.A. (1987). The Psychology of Reading and Language Comprehension.

Boston, MA: Allyn and Bacon.

Lévy-Schoen, A. (1981). Flexible and/or rigid control of oculomotor scanning behavior. In D.F.

Fisher, R.A. Monty, & J.W. Senders (Eds.). Eye movements: Cognition and visual perception

(pp. 299-316). Hillsdale, NJ: Erlbaum.

Liu, P.& Li, X. (2013). Optimal viewing position effects in the processing of isolated Chinese words.

Vision Research, 81, 45–57. https://doi.org/10.1016/j.visres.2013.02.004

Macedo, E., Lukasova, K., Ariente, L., Koakutu, J., & Schwartzman, J. S. (2007). Processos

perceptuais e cognitivos na leitura de palavras: propriedades dos movimentos oculares.

Revista Semestral da Associação Brasileira de Psicologia Escolar e Educaciona, 11(2), 275-

283.

Picanço, G. L., & Vansiler, N. D. S. S. (2014). A prosódia e a leitura fluente. Gragoatá, 19(36), 157-

174.

Rayner, K. (2014). The gaze-contingent moving window in reading: development and review. Visual

Cognition, 22, 242–258.

Rayner, K. (1998). Eye movements in reading and information processing: 20 years of research.

Psychological Bulletin, 12, 372-422.

Rayner, K. (1978). Eye movements in reading and information processing. Psychological Bulletin 85,

618-660.

Káñina, Rev. Artes y Letras, Univ. de Costa Rica XLVIII (3) (Setiembre-Diciembre) 2024: 1-22/ISSNe: 2215-2636

Rayner, K., Pollatsek, A., Ashby, J. & Clifton, C. Jr. (2012). Psychology of reading (2th ed.).

Psychology Press.

Rayner, K., Chace, K. H., Slattery, T. J., &Ashby, J. (2006). Eye Movements as Reflections of

Comprehension Processes in Reading. Scientific Studies of Reading, 10(3), 241–255.

Rayner, K. & Duffy, S. A. (1986). Lexical complexity and fixation times in reading: effects of word

frequency, verb complexity and lexical ambiguity. Memory & Cognition, 14, 191- 201.

Inhoff, A. W., Solomon, M., Radach, R., & Bradley, A S. (2011). Temporal dynamics of the eye–

voice span and eye movement control during oral reading. Journal of Cognitive Psychology,

23, 543-558.

Inhoff, A. W., Connine, C., Eiter, B., Radach, R., & Heller, D. (2004). Phonological representation of

words in working memory during sentence reading. Psychonomic Bulletin & Review, 11(2),

320–325.

Søvik, N., Arntzen, O. & Samuelstuen, M.S.V. (2000). Eye-movement parameters and reading speed.

Reading and Writing, 13, 237–255.

Vansiler, N. D. S. S. (2015). Expressividade oral e fluência em leitura: Monitoramento e diagnóstico

de cinco escolas estaduais de Belém do Pará. 110p Dissertação (master's degree in

Linguistics). Universidade Federal do Pará, Belém.

Vansiler, N. D. S. S. (2021). Movimentos oculares e prosódia de leitura oral: análise dos marcadores

prosódicos gráficos na leitura de alunos do 5º ano do ensino fundamental. 231p. Tese (PhD in

Linguistics) Universidade Federal do Pará, Belém.

Esta obra está disponible bajo una licencia https://creativecommons.org/licenses/by-nc-nd/4.0/