Tracking Adults’ Eye Movements to Study Text Comprehension: A Review Article

Abstract

Adult readers’ cognitive effort during text processing is often associated with their reading comprehension, learning ability, and achievement scores. The aim of this review is to examine and analyze the current literature on the use of eye tracking technology as a tool for assessing text comprehension. A systematic review was conducted and, after the final screening, 13 articles were analyzed that fell into three main areas: (a) eye movements during reading in print and digital media, (b) eye tracking in text comprehension with perspective effects, and (c) eye tracking in text comprehension with instructional strategy effects. The findings of this review showed that during reading, the amount of cognitive effort invested in text processing, whether induced by the text, the task, or the readers themselves, is usually reflected in longer total fixation times, both as a result of higher fixation frequencies and longer fixation durations.

1. Introduction

Reading comprehension is the process of understanding the meaning of a text and the mental representations resulting from this process, and it involves perceptual, mental, and motoric functions that are essential for understanding individual words, phrases, sentences, and larger units of discourse (Kendeou et al. 2013; McNamara et al. 2014; Perfetti and Stafura 2014).

In recent years, computational models of reading have been developed to explain and simulate the aforementioned functions (Reichle 2021). Those models aim to explain the process of comprehension and the cognitive processes and skills that support and explain variance in accuracy, and they are supported by a range of research works (Gough and Tunmer 1986; Kim 2017, 2020; Kintsch 1988; Kintsch and van Dijk 1978). According to variance computational models of reading, successful decoding is a complex process and depends on cognitive skills such as word reading and oral language comprehension, including syntax and inference. These skills help to construct a coherent representation of individual clauses, sentences, and passages of text (Kim 2017, 2020). Furthermore, measuring reading comprehension accurately and reliably is challenging due to its complexity. Recent research has challenged the validity and reliability of standard reading comprehension measures, prompting researchers to explore alternative methods (Sabatini et al. 2019). Educational researchers have searched for research methods developed in various academic domains to present the process of learning from different perspectives (Anderson 2009). Eye tracking technology has recently gained interest among researchers in the field of education, as only researchers in the field of psychology have studied basic cognitive processes during reading as a perceptual process, as well as types of information processing (Rayner 1998, 2009). Eye movement tracking is a widely used method in reading research as it is a non-invasive, online, and ecologically valid approach to studying the reading process at the word, sentence, and discourse levels (Rayner et al. 2006). Tracking eye movements provides an online measure of the reading process, reflecting moment-to-moment control of the eyes during reading and aiming to comprehend a connected text (Just and Carpenter 1980). Research suggests a relationship between eye movements and reading comprehension processes. However, most studies manipulate linguistic variables to investigate their effect on eye movement behavior. The complexity and coherence of a text are key linguistic features used to assess reading comprehension and potential learning challenges. Over recent decades, there has been growing emphasis on the assessment of text complexity, as the selection of suitable texts for various educational levels has become an increasingly important aspect of curriculum design. This approach ensures that students can gradually enhance their reading and comprehension skills (Torres et al. 2021). Although there have been numerous studies on this topic (Chen and Chen 2020; Jian 2018, 2019, 2021; Kim et al. 2018; Liao et al. 2020; Mason et al. 2013; Schuster et al. 2020; Suckow and van Gompel 2018), the relationship between eye movement behavior during reading and specifically during text comprehension has not yet been well understood.

1.1. Reading Processing and Comprehension

Reading comprehension is a complex cognitive process that involves various linguistic, cognitive, and perceptual mechanisms. Understanding how adults process different types of materials during reading can provide valuable insights into the underlying mechanisms of comprehension. Eye tracking, a powerful methodology in cognitive science, allows researchers to investigate the temporal dynamics of reading by monitoring eye movements as individuals engage with text (Rayner 1998). Eye tracking studies have revealed intricate patterns of eye movements associated with different stages of comprehension, shedding light on the cognitive processes underlying reading (Reichle et al. 2003).

Furthermore, the type of material being read significantly influences reading processing strategies and comprehension outcomes in adults (Ackerman and Goldsmith 2011; Mangen et al. 2013). Print versus digital text, for example, has been shown to lead to differences in reading speed, fixation duration, and overall comprehension (Ackerman and Goldsmith 2011; Mangen et al. 2013). Additionally, the format of the material, such as paragraphs, lists, or tables, can affect how readers allocate attention and process information (Rayner 1998). Visual elements, such as images or diagrams, may also play a crucial role in comprehension by either facilitating or hindering the integration of textual information (Mayer 2020).

Eye tracking studies have shed light on the temporal dynamics of reading, revealing distinct patterns of eye movements associated with different stages of comprehension. For instance, during the initial stages of reading, readers tend to fixate on content words while rapidly scanning the text for meaningful information (Rayner 1998). As comprehension deepens, readers may engage in regressive saccades, rereading specific passages or words to clarify meaning or resolve ambiguity (Reichle et al. 2003). The frequency and duration of these eye movements can vary depending on factors such as reading proficiency, text complexity, and individual reading strategies (Angele and Duñabeitia 2024; Just and Carpenter 2018).

In addition to material type, several other factors influence reading comprehension in adults as observed in eye tracking studies. These include individual differences in working memory capacity (Daneman and Carpenter 1980), prior knowledge of the topic (McNamara and Magliano 2009), and language proficiency (Perfetti 2007). Moreover, situational factors such as the reading environment, task demands, and cognitive load can also impact comprehension processes (Sweller et al. 2011). Understanding how these factors interact with material type can provide valuable insights into designing effective reading interventions and educational materials for adult learners.

Eye tracking studies offer a valuable tool for investigating the intricate processes underlying reading comprehension in adults. By examining how individuals process different types of materials, researchers can gain a deeper understanding of the cognitive mechanisms involved in reading and comprehension.

1.2. Eye Tracking Technology in Reading Comprehension Research

The measurement of reading comprehension is challenging due to its complexity, leading researchers to question the validity and reliability of established methods and to seek alternatives (Sabatini et al. 2013, 2020). Eye tracking is an experimental technique that records eye movement and gaze location during tasks and is commonly used to observe the allocation of visual attention (Carter and Luke 2020). Although eye tracking technology appears to be a modern method, its roots can be traced back to the 19th century. Charles Bell was the first to attribute the control of eye movements to the brain, to classify eye movements, and to describe the effect of eye movements on visual orientation (Bell 1823). By describing the physiological connection between the eyes and the nervous system, he linked their movement to neurological and cognitive processes. This connection provides a potential insight into the inner workings of the mind (Carter and Luke 2020).

Eye tracking technology is a non-invasive method that has been commonly used in reading research to investigate the reading process (Hyönä et al. 2003; Penttinen et al. 2013; Rayner et al. 2006). Eye tracking enables readers to freely read the text without being interrupted by a secondary task. Furthermore, it is possible to collect several indices of text processing simultaneously. By tracking eye movements, a large amount of data on the allocation of visual attention when performing complex tasks such as reading scientific texts can be provided (Ariasi and Mason 2011).

The eye tracking method has been used in studies of research on multimedia learning (Delgado and Salmerón 2022; Jeong and Gweon 2021; Jian 2022; Latini et al. 2020) and expository text reading (Hyönä and Lorch 2004; Kaakinen and Hyönä 2005, 2007; Penttinen et al. 2013). The eye tracking methodology is based on the eye–mind hypothesis, which suggests that the direction of human gaze is linked to the focus of attention as individuals process visual information. This link between locus of attention and eye position is particularly close in complex information processing tasks, such as scene perception, visual search, and reading (Just and Carpenter 1980; Rayner 1998; Rayner et al. 2006). Therefore, eye movements can be used to detect information processing and measure comprehension ability during reading (Hyönä et al. 2003).

Eye tracking technology provides two main measurements of eye movement: saccades and fixations. Saccades are rapid eye movements from one viewing location to the next, and fixations are the pauses between saccades where the eyes are relatively stationary. The text encodes information only during fixations, as vision is suppressed during saccades. Fixations typically last around 200–250 milliseconds, but there is considerable variability in the measurements. Regressions, where the eyes turn back to see part of the text again, constitute 10–15% of all saccadic movements. Therefore, the complexity of a text directly correlates with the number of fixations, shorter saccades, and regressions that occur during reading (Rayner et al. 2006).

Previous research has used eye tracking technology to study the relationship between reading comprehension and eye movement behavior. However, few studies have investigated the relationship between eye movement behavior and reading comprehension accuracy. Research has shown that better reading comprehension is associated with shorter fixations and fewer regressions in eye movement patterns (Kim et al. 2019; Parshina et al. 2021). On the other hand, some studies that have investigated the relationship between regressions and comprehension accuracy have concluded that making more regressions is associated with better reading comprehension (Schotter et al. 2014; Wonnacott et al. 2016).

In their study, Inhoff et al. (2018) showed that eye movement measures can predict comprehension outcomes. Other studies have also concluded that the relationship between eye movements and comprehension is consistent across datasets. Specifically, more and shorter fixations are associated with better comprehension scores (D’Mello et al. 2020; Southwell et al. 2020). Studies investigating the relationship between eye movement behavior and reading comprehension accuracy have produced varied and sometimes conflicting results. As a result, the indicators of successful reading comprehension through eye movement have not been clearly defined to date.

The purpose of this review is to present research on the use of eye tracking technology as a method of defining adult reading comprehension and text processing. This review considers studies conducted between 2002 and 2022. The research question that this review aims to answer is as follows:

Q: How can eye tracking technology contribute to the understanding of the processes involved in adult text comprehension?

2. Method

This review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) (Moher et al. 2009) guidelines.

2.1. Eligibility Criteria

The inclusion criteria used in this review were as follows: (1) articles examining reading comprehension in a text; (2) the use of the eye tracking technology is clearly outlined and the eye movement measures are clearly stated; and (3) the participants are adults from a normative population (i.e., typical development) who are characterized in the literature as undergraduate or postgraduate students at universities with a mean age of 27 years, not exceeded by any research.

This study was restricted to scientific original research articles published in peer-reviewed journals in English between 2002 and 2022. Review articles, commentary articles, doctoral dissertations, and descriptive studies were excluded. Additionally, articles that focused on reading comprehension in sentences and words, and articles that did not use eye tracking technology, were also excluded. This review excluded studies that involved children or populations with special education needs or disorders, such as reading difficulties, cognitive impairments, or disorders, in relation to the participants.

The synthesis of the selected studies covered three main areas: (1) eye movements during reading in print and digital media, (2) tracking eye movements in text comprehension with perspective effects, and (3) tracking eye movements in text comprehension with the effects of instructional strategies.

2.2. Information Sources and Search Strategy

Comprehensive research was conducted electronically through two databases: (a) PsycInfo and (b) Education Resources Information Center (ERIC). These databases were chosen as they are commonly used by researchers to access related studies (Fraenkel et al. 2012). The search was limited to studies published between 2012 and 2022. To review studies of potentially more consistent quality, the researchers extended their search to include studies from 2002 until 2022 that focused on adults’ reading comprehension and eye tracking. The search was limited to journal articles. The terms ‘reading comprehension’ and ‘text processing’ were searched in combination with ‘eye-tracking’ and ‘eye movements’ across all databases. The section on the topic was searched for key terms, including title, abstract, and author keywords. Additionally, full-text English articles were searched using the same key terms. A review of the abstracts was conducted to determine which studies would be included in this review. The initial search yielded 60 publications in PsycInfo and 86 in Education Resources Information Center (ERIC).

2.3. Selection Process

The authors of this review selected original research articles that used eye tracking technology in reading comprehension research. Many of the articles were excluded from the selection process because they were not related to the field of educational research, were not original research articles, and did not use eye tracking technology in reading comprehension research. To clarify the focus of this article, eye tracking technology was defined as the use of eye movements to meet the needs of adults in reading comprehension. In addition, some articles were excluded from the initial reading of titles, abstracts, and keywords because they did not meet the inclusion criteria. The studies that finally emerged from the search process were read and critically appraised to ensure the validity of the inclusion criteria.

Reviews, commentaries, and doctoral theses were not included in this review. Descriptive studies that reported gains in reading comprehension as an additional benefit of implementing eye tracking technology, but did not have reading comprehension as a goal of text processing, nor did they report measurable outcomes on reading comprehension tasks, were also excluded. After eliminating duplicate studies, the authors independently applied the inclusion and exclusion criteria. Finally, 13 articles were included in this review.

2.4. Data Collection Process and Items

All 13 articles were independently reviewed by the two authors of this review. In each study, they identified the number, age, and typical development of participants, as well as native speakers and competent readers (L1) of the language, as stated in the articles included in the review, in which they were asked to read the survey texts; the research hypotheses of each study; the eye movement and performance measures; the learning topic; the properties of the text; and the results of each study. Following the independent reviews, cases where there were some discrepancies in data collection were discussed and resolved by consensus.

3. Results

3.1. Study Selection

As a result of the search process, 146 studies were identified for possible inclusion in this review. Each of these studies were then assessed by the authors to determine whether they met the inclusion criteria in terms of participants, target outcomes, and use of eye tracking technology. A total of 13 articles met the inclusion criteria and were further categorized according to the topic defined by the study researchers and the reported study aims to investigate the intentional use of eye tracking as a tool to examine eye movements in relation to participants’ reading comprehension. As more than one experiment was reported in two articles, each of the experiments were counted as independent studies. As a result, a total of 15 studies were included in this review. The search and selection process are summarized in Figure 1.

3.2. Study Characteristics

3.2.1. Eye Movements During Reading in Print and Digital Media

The reading of print and digital texts, with a focus on potential differences in processing and comprehension, has become increasingly popular in research on teaching practice and educational policy over the last decade (Delgado and Salmerón 2022; Latini et al. 2020). In terms of comprehension, four of the studies included in this review investigate the reading medium used by readers and the extent to which reading comprehension is influenced by the medium.

Latini et al. (2020) used an illustrated text to investigate the effects of reading medium (print vs. digital) on integrative processing and integrated comprehension, and whether the reading medium indirectly influences integrated comprehension via integrative processing. In a research design with undergraduate and graduate students (n = 100), the researchers found that the participants who read the illustrated text on paper showed more integrative processing during reading than participants who read the exact same text on a computer. The researchers also examined the number of the participants’ gazes, and they found that those who made more cross-representational gaze transitions were also more likely to show integrated comprehension in their written responses. Integrative processing positively influenced comprehension performance, resulting in a mediated effect of reading medium on comprehension via integrative processing. However, there was no main effect of reading medium on integrated comprehension, and asking participants to pay attention to both text and illustration, as well as their correlation, had no main effect on integrative processing or integrated comprehension. There were also no interaction effects of reading medium with instructional prompts on integrative processing or integrated comprehension.

Furthermore, Jeong and Gweon (2021) examined the effects of reading medium (print vs. digital) on readers’ visual patterns, more specifically, fixation duration and fixation count, reading performance, and reading attitudes of readers (n = 74) who read articles using three reading media: print, computer, and tablet. The results of the research suggest that the performance gap between print and screen reading is narrowing, with print still being the preferred mode of reading. According to the results of the visual patterns, readers had a shorter fixation duration and a higher number of fixations when reading in print than when reading on the screen, thus strengthening the hypothesis that screen reading is related to higher cognitive loading and skimming (Shojaeizadeh et al. 2016). The results for reading performance, measured in terms of reading comprehension and reading time, showed that reading performance was equivalent for all three media. Finally, in terms of reading attitude, readers reported higher levels of perceived comprehension, perceived confidence, and perceived immersion, and lower levels of perceived fatigue when reading print than when reading from a device screen.

In addition, Delgado and Salmerón (2022) investigated the effect of reading medium and reading time on text processing, metacognitive monitoring of comprehension, and comprehension outcomes in a research design with undergraduate students (n = 116). Eye movements, and, more specifically, fixation time, number of fixations, and number of regressive fixations, were recorded while participants read texts in print and on a tablet under self-paced or timed conditions, and participants predicted their performance on a subsequent text comprehension test. The researchers found that the participants performed similarly across media in all the reading processes assessed, regardless of the reading time. However, participants fixated more on text titles when reading in print than on the tablet, and this was the only significant difference between print and digital reading media. Therefore, the results are in line with recent evidence suggesting that the reading superiority of print is negligible when reading on tablets.

Similarly, Jian (2022) investigated the differences in reading processes in print and digital media in an eye tracking experiment with undergraduate students (n = 50). Participants were randomly assigned to one of two groups, print or digital, and read a scientific article containing several diagrams either in print or on a tablet, and then they answered a reading comprehension test. While the results of the eye movement test, more specifically, the total fixation duration, the first-pass fixation duration, rereading (or second-pass) fixation duration, and the number of rereading instances across pages showed that the two groups spent about the same amount of time processing the article, text, diagrams, and diagram explanations, the time was not evenly divided between the first pass and the rereading phase. As a result, comprehension was better when reading in print than when reading on a tablet due to the different cognitive strategies used by the participants in the print group, who showed more selective and deliberate reading behavior than those in the digital group.

3.2.2. Tracking the Eyes in Text Comprehension with Perspectives Effects

A total of four studies reported gains in participants’ reading comprehension with perspective effects by tracking the eyes. Kaakinen et al. (2002) investigated the effect of reading perspective on online text processing by recording participants’ eye movements, measuring the participants’ total fixation time (first-pass fixation time and look-back fixation time) while reading. The researchers’ aim was to evaluate the encoding time and effortless encoding hypotheses as possible explanations for perspective effects in text comprehension. Students (n = 64) read an expository text about four countries with the goal of deciding which country, designated by the experimenter as the reading perspective, would be a good new place to live. The results of this study showed that reading perspective influences the online processing of the text by inducing readers to invest extra processing time on relevant information, resulting in better memory and comprehension for relevant than for irrelevant information. Also, the results showed longer total fixation time (the sum of first-pass and look-back fixations) on perspective-relevant information than irrelevant information.

In the same line, Kaakinen and Hyönä (2005) investigated the effects of perspective on reading comprehension using eye tracking. They recorded eye movements, measuring the first-pass progressive fixation time, the first-pass rereading time, and the look-back fixation time of the participants (n = 36) while they read an expository text from a particular perspective. The participants were asked to think aloud and identify relevant and irrelevant sentences in the text. The results of the study suggest that a reading perspective not only increases fixation time and recall for relevant information compared to irrelevant information, but also leads readers to use slightly different comprehension processes for relevant information.

In addition, Kaakinen and Hyönä (2007) investigated the detailed time course and nature of perspective effects in text comprehension. The aim of the study was to investigate the influence of perspective effects and perspective instructions on text processing during repeated reading. The text used in this study was the same as in their 2005 study, the results of which are not mentioned in this study. The participants, consisting of students (n = 62), were asked to read a text with high or low prior knowledge twice from a given perspective while their eye movements, more specifically gaze duration (first-pass processing measures, second-pass reading of words), probability of skipping words, and regression, were recorded. The results of the study suggest that prior knowledge played a crucial role in how readers allocated visual attention during reading, resulting in a more comprehensive memory for perspective-relevant information. Words in the relevant sentences attracted longer gaze durations than words in the irrelevant sentences, and repeated reading facilitated processing, as indicated by all eye movement measures.

Kaakinen and Hyönä (2008) investigated whether the framework of perspective-driven text comprehension applies to the comprehension of narrative texts. An eye tracking experiment was designed for undergraduate students (n = 64) measuring their progressive fixation time (duration of fixations), first-pass rereading time (total fixation time) and look-back fixation time (second-pass fixations). The participants read either a transparent version of the text, in which the relevance of text segments to perspective was made apparent, or an opaque version of the text, in which perspective was not directly mentioned. The study found that prior knowledge of perspective influences the perspective effects observed in online text processing. In addition, signaling relevance or irrelevance helps to encode important information in memory. A clear perspective effect was observed in text recall, despite the absence of a perspective effect (i.e., longer processing of relevant than irrelevant text information) in any of the fixation time measures in the high prior knowledge condition. The study concludes that the proposed framework is effective.

3.2.3. Tracking the Eyes in Text Comprehension with Effects of Instructional Strategies

Five studies in this review investigated eye movements as indicators of reading effects. Hyönä and Lorch (2004) studied the effects of topic headings on the processing of multiple-topic texts by tracking eye movements, measuring the readers’ first-pass fixation time, first-pass forward fixation time, first-pass rereading time, look-back fixation time, and look-from-fixation time. The study included participants (n = 66) who read two types of texts: one with topic shifts signaled by topic headings and one without. The authors concluded that the use of topic headings can aid in reading comprehension and improve memory retention of text topics. This facilitatory effect was observed in both the initial reading fixations and subsequent look-backs directed towards the topic sentences.

Rayner et al. (2006) further investigated the relationship between reading comprehension processes and eye movements, measuring the average fixation duration, number of fixations, and total fixation time of adult readers. The researchers conducted two studies to confirm that eye movements are sensitive to global text passage difficulty and inconsistencies in text during reading. In Study 1, the eye movements of 16 participants were monitored, and in Study 2, the eye movements of 18 were monitored, while reading texts. The results showed that processing times and fixations increased when the text was difficult to comprehend, and readers fixated more on the area where an inconsistency occurred. In both studies, the probability of making a regressive eye movement increased.

Furthermore, Ariasi and Mason (2011) examined whether reading a refutational text or a non-refutational text would elicit different cognitive processing, as revealed by eye movement analyses, measuring the readers’ first-pass duration, total fixation duration, second-pass fixation, and total reading time. Participants (n = 40) were assigned to read one of the two types of text. The findings of the study indicate that the readers of a refutational text perceived more than those who read non-refutational text in the post-test (off-line measure). In addition, the results showed that readers of refutational texts fixated on scientific concepts for a longer time than readers of non-refutational texts, particularly during the second-pass reading. Additionally, refutational text readers spent less time fixating on the refutational segments compared to non-refutational text readers for the control segments.

Penttinen et al. (2013) investigated the interplay of conceptual change, text comprehension, and eye movements, more specifically fixations, during reading. Two studies were conducted with 15 and 23 participants, respectively. The participants first read a text, then explained their reading processes retrospectively, cued with their own gaze videos, and finally answered pretests and post-tests. The results of Study 1 revealed connections between rereadings and high-level cognitive processing, and an increase in eye movements between two or three or more sentences of texts during reading. Study 2 found that learners undergoing conceptual changes differed from other learner groups in terms of prolonged overall reading time, and that a relatively high amount of textbase construction was expressed at the beginning of the retrospective reporting. The study suggests that complementing eye tracking with cued retrospective reporting is beneficial when examining high-level cognitive processes during reading. It also highlights the importance of constructing the textbase carefully in conceptual change.

Ariasi et al. (2017) evaluated the effectiveness of eye tracking methodology in understanding the refutation text effect on online text comprehension, capitalizing on the association between reading comprehension and eye movements, measuring the readers’ first-pass progressive fixation time, first-pass reinspection time, and look-back fixation time (second-pass fixation). It was found that a refutation text acknowledges the reader’s alternative conceptions about a phenomenon, refutes them, and presents the correct conceptions. The study investigated two hypotheses: the coherence hypothesis, which suggests that a refutation text is more coherent than a standard text, making it easier to comprehend, and the elaboration hypothesis, which suggests that a refutation text involves deeper processing, also facilitating comprehension. The study involved participants (n = 40) who read one refutation text and one non-refutation text on two different topics. The results of the offline data analysis showed that readers of refutation texts recalled more scientific facts than those who read non-refutation text. Additionally, online eye tracking measures confirmed a decrease and an increase in reading time in response to the refutation statements. This was due to the fact that topic-medial text sentences with correct scientific facts were fixated on for a shorter time when first encountered in the refutation text. However, refutation statements increased integrative processing at the end of each text paragraph. This was indexed by longer look-back fixation times on topic-final sentences with the science concepts, as well as longer look-back fixation times directed to the refutation statements.

3.3. Results of Individual Studies

Table 1. Research articles of the review on text comprehension with eye tracking.

	ID Author	Research Purposes	Participants	Eye Movements Measures	Performance Measures	Focus of the Study
1	Ariasi and Mason (2011)	This study examined whether reading a refutational or non- refutational text would induce different cognitive processing, as revealed by eye movement analyses.	40	First-pass duration, total fixation duration, second-pass fixation, total reading time	Pretest and post-test, knowledge test	Text comprehension with effects of instructional strategies
2	Ariasi et al. (2017)	This study used an eye tracking methodology for a deeper understanding of the refutation text effect on online text comprehension.	40	First-pass progressive fixation time, first-pass reinspection time, look-back fixation time (second-pass fixation)	Pretest and post-test knowledge questions	Text comprehension with effects of instructional strategies
3	Delgado and Salmerón (2022)	This study examined the effect of the reading medium and the reading time-frame on text processing, metacognitive monitoring of comprehension, and comprehension outcomes.	116	Fixation time, number of fixations, number of regressive fixations	Prior knowledge and topic interest questionnaire, comprehension questions	Text comprehension in print and digital media
4	Hyönä and Lorch (2004)	This study investigated the effects of topic headings on the processing of multiple-topic expository texts with the help of readers’ eye fixation patterns.	66	First-pass fixation time, first-pass forward fixation time, first-pass rereading time, look-back fixation time, look-from-fixation time	Text summary	Text comprehension with effects of instructional strategies
5	Jeong and Gweon (2021)	This study examined the effects of the reading medium (print vs. digital) on readers’ visual patterns, reading performance, and reading attitudes.	74	Fixation duration, fixation count	Reading comprehension level test, pre-survey questionnaire, reading time, reading comprehension test, reading attitudes questionnaire	Text comprehension in print and digital media
6	Jian (2022)	This study involved an eye tracking experiment to investigate the differences in reading processes in print and digital media.	50	Total fixation duration, first-pass fixation duration, rereading (or second-pass) fixation duration, number of rereading instances across pages	Demographic survey, test of prior knowledge, reading comprehension test	Text comprehension in print and digital media
7	Kaakinen et al. (2002)	This study examined the effect of a reading perspective on online text processing, studied by recording readers’ eye movements during reading.	64	Total fixation time, first-pass fixation time, look-back fixation time	Reading span test, recall test	Text comprehension with perspective effects
8	Kaakinen and Hyönä (2005)	This study investigated the recording of participants’ eye movements while reading an expository text from a given perspective, and think-alouds were probed after 10 relevant and 10 irrelevant sentences.	36	First-pass progressive fixation time, first-pass rereading time, look-back fixation time	Recognition of target items, think aloud, recall performance	Text comprehension with perspective effects
9	Kaakinen and Hyönä (2007)	This study examined the influence of perspective instructions on the online processing of expository text during repeated reading.	62	Gaze duration (first-pass processing measures, second-pass reading of words), probability of skipping word, regression	Recall performance	Text comprehension with perspective effects
10	Kaakinen and Hyönä (2008)	This study examined whether the perspective-driven text comprehension framework applies to comprehension of narrative text.	64	Progressive fixation time (duration of fixations), first-pass rereading time (total fixation time), look-back fixation time (second-pass fixations)	Recall performance	Text comprehension with perspective effects
11	Latini et al. (2020)	This study investigated effects of reading medium (print vs. digital) on integrative processing and integrated understanding of an illustrated text, as well as whether reading medium indirectly affected integrated understanding via integrative processing.	100	Number of gazes	Demographic survey, writing report, inventory, reading comprehension measure, diagram comprehension score	Text comprehension in print and digital media
12	Penttinen et al. (2013)	This study examined the interplay of conceptual change, text comprehension, and eye movements during reading, and developed and tested methods suitable for such explorations.	15 + 23	Fixations	Pretest (open-ended questions) and post-test (open-ended questions and two interviews) questionnaires, average reading time	Text comprehension with effects of instructional strategies
13	Rayner et al. (2006)	This study examined the results of two experiments about the eye movement data to assess moment-to-moment comprehension processes: (a) global text passage difficulty and (b) inconsistencies in text.	16 + 18	Average fixation duration, number of fixations, total fixation time	Comprehension test, total reading time	Text comprehension with effects of instructional strategies

presents a summary of the studies included in this review.

3.4. Results of Syntheses

The first section of this review included research articles that investigated eye movements during reading in print and digital media. The results of these studies showed that there were no significant differences between reading in print and reading in digital media in terms of reading comprehension, and the slight superiority in reading comprehension tests of print over digital reading that was found was negligible. However, the results of the visual patterns showed that readers had a higher number of fixations when reading in print than when reading on screens. Therefore, the researchers claim that this slight superiority when reading in print over reading on screens can be attributed to different cognitive strategies and to more integrative processing that occurs when reading in print compared to reading with a digital medium.

The studies included in the second section of the review examine eye movements during text comprehension with perspective effects. Summarizing the findings, it is concluded that reading comprehension and text processing are affected by perspective effects. The results of the studies suggest that a reading perspective increases fixation time, and readers tend to use slightly different comprehension processes and to store more perspective-relevant text information during reading. Furthermore, the results of the studies reviewed in this section suggest that assigning a comprehension perspective and activating readers’ prior knowledge can enhance the learning of perspective-relevant information, not only from expository material but also from narrative texts.

Finally, the third section of this review includes research articles that examine eye movements during text comprehension with the effects of instructional strategies. All the studies reviewed in this section found that processing times and fixations increased when the texts exhibited instructional strategy effects, as opposed to when the texts did not. Thus, the researchers conclude that eye movements in reading comprehension are affected by instructional strategy effects.

4. Discussion

In the present study, we conducted an updated review of empirical studies that have examined the use of eye movement data to assess comprehension processes in adult readers. The results showed a relationship between the cognitive processes activated during adult reading comprehension and the associated eye movements. Thus, they indicate that eye tracking technology can be seen as a complementary medium that supports the study of cognitive processes during reading and provides accurate indicators of comprehension processes that can be collected during reading tasks.

In the studies reviewed in this article, the reading process is influenced by the reader’s word recognition, comprehension, and motivation (McLaughlin 2012). In addition, physical and behavioral factors, such as eye strain and emotions, and performance factors, such as reading comprehension, are aspects of reading that should be considered when comparing print and screen reading (Keller 2012). According to the results of the studies, there were no significant differences between reading in print and reading in digital media, and therefore it is concluded that reading comprehension is not affected by medium. However, the results of the visual patterns indicated that printed text is still the best medium in a learning environment because, as has been previously claimed (Jeong and Gweon 2021), readers have more positive reading attitudes towards print than digital devices. The results of eye movement analyses show that reading on digital devices is associated with a higher cognitive load, which may prevent readers from reading in depth (Destefano and Lefevre 2007). As the transition from print to electronic sources has increased over the years, studies examining the impact of the medium on reading comprehension and the general reading experience are timely and necessary.

Reading comprehension is also related to the memory representation that is constructed around the text characteristics, the reading perspectives, and instructional strategies. A number of studies in this review examined the relationship between reading comprehension and reading perspective effects or instructional strategies, while tracking eye movements, in order to investigate whether a reading perspective effect or instructional strategy influences the processing of a text. Studies in which the participants either read an expository text from a given perspective, or were provided perspective instructions for text processing during reading, focused on the relationship between reading comprehension and the activation of readers’ relevant prior knowledge. The way in which visual attention, and more specifically eye fixations, is allocated to the text is modulated by the availability of prior knowledge. Therefore, when the reader encounters perspective-relevant textual information or instructional strategies, the available prior knowledge is activated with the textual input (Kuperberg and Jaeger 2016). The activation of relevant prior knowledge, either with perspective instructions or with linguistic signals embedded in the text, facilitates the encoding of relevant information in memory regardless of the type of text (Kaakinen and Hyönä 2008).

The findings from this review are consistent with a growing body of research on the link between learning and eye movements (Lai et al. 2013). Although there is not enough research on the interplay between reading comprehension, text processing, and eye movements in adult reading, there seems to be a clear link between reading comprehension and eye movements, and between reading comprehension and cognitive processes. The corpus of research on eye movements during reading shows that the amount of cognitive effort invested in text processing, whether induced by the text (e.g., text difficulty), the task (e.g., reading goals), or the readers themselves (e.g., use of reading strategies), is usually reflected in longer total fixation times, both as a result of higher fixation frequencies and longer fixation durations (Delgado and Salmerón 2022). In summary, the use of eye movement monitoring provides a research tool to further investigate moment-to-moment comprehension processes.

However, the conclusions drawn from the studies on the use of eye tracking technology in reading comprehension and text processing analyzed in this review should be treated with caution, as they may be influenced by publication bias. The 13 articles included were published, whereas unpublished studies were not included. Eye tracking studies that do not achieve statistically significant results tend to remain unpublished, and the omission of such studies could bias conclusions about the use of eye tracking technology in reading comprehension and text processing.

In addition, reading comprehension is a complex task that involves a variety of cognitive and metacognitive processes. Some of the studies included in this review did not describe in detail the processes involved in reading a text, and most of the studies that did specify the processes involved different types of activities. Future studies should investigate the use of eye tracking technology and consider specific aspects of eye movements that may contribute to improved reading comprehension.

Eye tracking and eye movement as indicators of effects on reading comprehension can provide a context for investigating the comprehension skills and cognitive processes of adults. Future research should focus on the types of eye tracking measures (e.g., fixation duration, time to first fixation, etc.) and the appropriate measures for each research situation. Researchers should also focus on individual differences between participants, which will contribute to the development of appropriate adaptive learning systems that take into account the characteristics of the reader. Analyses of the frequency and systematic patterns of eye fixations can therefore complement other online methods used to study processing during reading (Hyönä 2010). In conclusion, future research that will take into consideration all the parameters above could help unravel the process of text comprehension among adult readers and be a step forward towards the application of eye tracking procedures in the learning process.