Individual reading styles: a narrative approach to understanding reading behaviour


This article reports on the findings from a study that intended to consider the validity of a model of reading comprehension. In the process of coding the resulting verbal protocols, the ‘feel’ or experience of the actual readers was lost, as was the individuality of the participants. Narratives were then developed to describe the experience of the reading and to recapture that individual nature of each reader’s approach to making sense of the texts they read. After those narratives were developed, some patterns emerged that are discussed in this article as reading styles. Some of these styles resulted in the reader developing a good understanding of what they were reading and others did not. The successful styles were classified in two ways, by the main strategy used and by the deliberateness of the readers’ construction of understanding. There are two pairs of successful reading styles discussed here, making four successful styles in total. Readers using a visualisation strategy are described as either “movie makers” for those who consciously constructed their mental movies, and as “Movie watchers” for those who experienced the process as being more receptive. The other strategy used in successful reading was questioning, either as an active “Hunter-seeker” approach or in the less-active “Wondering” style. Three styles that were not associated with successful reading are also described. Some implications for future research, and for teaching and learning are also discussed.


Historically, research into reading comprehension has focussed on identifying specific behaviours and strategies of good readers, or on the related metacognition and self-regulation. This has resulted in some confusion in places. For example, authors that have focussed on the behaviour of good readers have tended to include monitoring as a strategy for developing meaning (e.g. National Reading Panel 2000) alongside other strategies such as making predictions. At the same time, researchers with a metacognitive focus have viewed monitoring as a part of the regulatory process or executive function (e.g. Jacobs and Paris 1987). In the last 15 years or so this has begun to change with a greater number of authors (e.g. Pressley 2002; Baker 2002; Gaskins et al. 2007; Kaplan 2008; Kennedy 2014) recognising that the relationships between metacognition, reading comprehension, and self-regulation need to be considered and could provide significant insights into the complex activity that is reading. This has been a positive development and has led to a fuller understanding of not only the strategies used and behaviours exhibited by good readers, but also how their use is regulated. One way of doing this is the use of verbal protocol analysis (Pressley and Afflerbach 1995; Israel 2015), which has enabled us to develop a much better understanding of what is going on in readers’ heads.

The use of verbal protocol analysis generates a large amount of rich data, which is generally rendered manageable through the use of a coding process (Evers 2011; Gu 2014). This process was originally used in the study that led to this article. The coding process was useful for identifying patterns of data and was aimed at supporting the development of a model of self-regulated reading comprehension. During coding however, the researcher identified that although useful for considering the patterns of reading behaviour across participants, much of the richness and understanding of individual readers’ behaviour was being compromised. Individual participants were being lost in the coding process.

The results of this study, as they are presented here, are intended to capture something of the process of reading as it is experienced by young readers as they read. In order to achieve this purpose, coding and other forms of analysis are not helpful (Gu 2014), as the richness and individuality of what was shared by participants would be lost.

Following the collection of the verbal protocol data and the researcher’s realisation of what was being lost during the coding process, narrative descriptions of individual reader’s reading process were developed. Doing so resulted in the identification of a number of reading styles, sometimes used by more than one participant in the study, and these reading styles are the focus of this article.

Literature review

We can define a good reader by saying that a good reader is one that has the word level and higher-level skills, knowledge, and strategies necessary for getting words off the page and for developing meaning from them. However, that is a gross over-simplification of what is a very complex activity. Vacca (2002) has described the majority of students in the middle grades (from about age 12) as being skilled in the mechanics of reading but insufficiently strategic in exploring text and developing meaning. There is more to the good reader than simply having more skills, more knowledge, or more well-taught cognitive reading strategies. Good readers are very involved with their reading from before they start until well after they have finished reading the text, and they are strategic. Their reading behaviour is controlled in a way that facilitates the achievement of a reading goal.

In the late 1990s and early 2000s several studies identified a number of features of good readers (e.g. Block and Pressley 2001; Pressley and Afflerbach 1995; Duke and Pearson 2002; National Reading Panel 2000; Pressley 2002) that have remained essentially unchanged since. These features are summarised as follows:

  • Good readers are active readers.

  • They have clear goals in mind, and they constantly evaluate their progress towards these.

  • Good readers look over the text before reading, they look for structural features and sections of the text that may be relevant to their goals.

  • They make frequent predictions or hypotheses, not just prior to reading.

  • They read selectively. They are constantly making decisions about whether to read carefully, quickly, what not to read, whether to re-read sections, or to jump forward or back in the text, although they generally read texts from beginning to end.

  • Good readers construct, revise, and question the meanings and understandings they develop as they read.

  • They try to determine the meaning of unknown words and concepts they encounter in the text, and find ways of dealing with inconsistencies or gaps in their knowledge.

  • They use, compare, and integrate their prior knowledge with what they find in the text.

  • Good readers think about the author(s) of the text and their intentions as they are reading.

  • They monitor their understandings, and alter their reading activity (such as the speed of reading or the strategies being used) as needed.

  • They evaluate the quality of the text and its value, and react to the text both intellectually and emotionally.

  • They change their approach to reading depending on whether they are reading narrative or expository text

As good readers move through the process of developing their understanding of a text they are active and goal directed (Pressley and Afflerbach 1995). They have the ability to use the strategies they have at their disposal constantly and in a fluid fashion. A large part of what makes a good reader is that they are able to orchestrate these strategies independently.

There are a number of ways in which the various strategies identified can be used by readers. Firstly, they are all things that the accomplished reader may be monitoring in order to ensure they are maintaining comprehension of what they read and may therefore be involved in the identification of what Klingner and Vaughn (1999) might have called the ‘clunk-point’. Secondly, they are all possible ‘fix-it’ strategies that may be utilised to correct an error. Thirdly, and possibly most significantly, they may also be a part of the reading process and not just a precursor or an outcome of it.

Duke and Pearson (2002) identified six “Effective Individual Comprehension Strategies”, prediction, think-aloud, story structure, visual representations of text, summarisation, and questions/questioning. The National Reading Panel (2000) identified these same six strategies but also included comprehension monitoring and co-operative learning. These strategies appear consistently throughout the literature (National Reading Panel 2000; Pressley 2002; Davis 2016) although sometimes with differing names. Each of these strategies has been shown to be an effective tool for developing good comprehension of text. As research in the last 15 years or so has seen little change to the generally accepted strategies, there appears to be a degree of agreement on the strategies involved in reading comprehension. This is evidenced by the number of references to studies from the 1990s and early 2000s that are seen in more recent literature (e.g. Kennedy 2014; Davis 2016; Conley 2016) when discussing the strategies used by readers. Despite the apparent agreement among researchers as to which strategies are used by readers, some issues remain.

A number of authors (e.g. Phillips 1988; Pressley and Afflerbach 1995; Pressley and Gaskins 2006) have identified that competent readers as a group employ a large range of cognitive strategies to achieve their goals and understand the text they are reading. The expectation that individual readers will have available, and use, a significant number of strategies is an important part of constructively responsive reading (Pressley and Afflerbach 1995) and much of the subsequent literature. What is not clear from these studies is whether individual readers regularly use a number of these strategies or whether individuals have preferences for a smaller number of strategies, and whether the strategy used in a particular situation is determined by circumstance. Those studies that have used quantitative analysis in connection with this kind of question leave this open to interpretation. For example, Phillips (1988) identified a large number of comprehension strategies used by young readers and gave the mean number of times each strategy was used. Even the most commonly reported strategies had standard deviations that were large in comparison to their average, these varied from a standard deviation that was equivalent to 21% of the mean to a standard deviation that was 160% of its related mean. Some of the readers in the study must have used particular strategies repeatedly while others did not use them at all. Studies of this sort mean that we do not have a clear picture of individual readers’ strategy use and we may be making assumptions based on generalisations.

Our understanding of how readers manage the reading process is based on what we know of a generalised understanding of what good readers do and on models of self-regulation. This study considers information-processing based models to be particularly instructive and the model developed by Winne (2001) informs the interpretation of the data and their discussion. We do not know all we need to know about individual’s strategy use. We need to better understand the strategy use of individuals to know how we should support students to use the strategies they develop for themselves or that we teach them. Should we be expecting and teaching students to use and choose between large numbers of strategies, or should we encourage them to be judicious in their choice of a single or small number of strategies? That is, perhaps good readers are not just better in their use of a range of strategies; maybe they manage the complexity better by limiting the range they apply in a given situation. This is the question considered in this article: How do individual readers orchestrate and manage their use of reading comprehension strategies?


The method used in this study involves the collection of verbal protocols or think-alouds. The method itself, and the detail with which it is reported here follow the recommendations of Pressley and Afflerbach (1995) and Ericsson and Simon (1993) in their seminal works.


In the current study, purposive sampling (Creswell 2007) was used to ensure appropriate participants were involved in the study, including the fact that all identified English as being their first language. Homogeneous sampling (McCrudden and Kendeou 2014) was used to identify two groups of participants, a group that has high reading comprehension skills (HC group) and a group that has low reading comprehension skills (LC group). A number of criteria were used to identify these groups. Firstly, a standardised test of reading ability, the Progressive Achievement Test (PAT) of reading comprehension was used (Darr et al. 2008). The PAT distributes scores using stanines. In general terms, a stanine score of one to three is considered low, stanines four to six are in the middle range, and stanines seven to nine are considered to be indicative of higher achievement. In the current study, a good reader is defined as someone who has a stanine of seven or higher, and a poor reader as someone with a stanine of three or lower. For the purposes of sampling these criteria were restricted further. For the LC group, only readers who had achieved a stanine of three (and not lower or higher) were included. For the HC group, only readers who had achieved a stanine of seven (and not higher or lower) were included. This restriction was to ensure that the LC readers had sufficient decoding ability to read a text and basic reading comprehension skills, but that these skills were not well developed and that the HC readers were very proficient but not exceptional readers. As individuals become more expert their behaviour becomes more automatic, and therefore less available for reporting. Including extremely poor readers and exceptional readers would also reduce the applicability and generalizability of the results.

A total of 12 prospective participants agreed to take part in the study, all were aged 11 at the time of initial contact. The HC group consisted of seven participants. All female, five of the HC group were of Pacific Island heritage and the other two were New Zealand Europeans. All seven identified English as their first language and had achieved stanine seven on a recent PAT. The LC group (stanine three on the PAT) consisted of a total of five participants, comprised of two male and three female students, all were from Pacific Island backgrounds and considered English to be their first language.


Participant materials

The texts used covered a range of topics with the intention that all participants should be able to find one that was of at least some interest. Each text was levelled using the noun-frequency method (Elley and Croft 1989) of assessing readability. The HC group were given texts with a reading age of 11–13 years and the LC group were given texts with a reading age of approximately nine. During reading all of those taking part made some verbal reports of what they were thinking.

All texts used were selected from the School Journal instructional text series used in New Zealand schools. Participants were initially given a page with three titles accompanied by a brief one-line precis of each text. From these, participants selected the text to read during the research phase. Once they had made their selection they were presented with the text. Both the text-selection page and the texts themselves had a small red dot placed at the end of each sentence. These dots were there as a reminder that participants should be reporting their thinking rather than as an indicator that there was a requirement to report at the point.

Researcher script

The researcher used a script to ensure consistency in the instructions and information given to all participants. Where participants raised questions then those were answered regardless of whether the information was already contained in the script. The script also contained a range of neutral prompts and probes that could be used at appropriate points during the research process in order to both maintain consistency and also to help ensure that researcher bias was not introduced at any point.

Research procedure

General principles and intentions

There were two general principles underlying the method used. The first of these was that the researcher should be involved as little as possible. The second was that the processes captured should mimic as closely as possible the personal reading styles of the participants when they are reading for their own purposes, as opposed to reading to achieve an external goal such as one set by a teacher. While it is accepted that no research procedure using methods currently available can be completely naturalistic, the procedure used was designed to create as little external influence as possible on the internal processes of participants’ reading while still enabling the collection of sufficiently rich data.

Initial contact and practice

The researcher held an initial meeting with all participants present to provide further information about the process involved, introduce them to the equipment to be used, and to answer any questions they had. Following a review of the process that would be followed, the participants had few questions about the research. Most of these questions revolved around practical issues such as whether they could choose the times when future contact would occur. Participants were also reminded that they had the opportunity to continue or withdraw from the research at any time.

At this initial contact, a brief introduction to the process of thinking aloud was carried out. This was completed using a puzzle activity rather than reading. A different activity provides the opportunity to discuss the process of thinking aloud while reducing the potential to confound the research data to be collected later.

The researcher initially modelled the process of thinking aloud, and then a number of the participants volunteered. These participants appeared to find it a relatively natural process and reported a number and range of thoughts during this activity. Encouragingly, they reported several metacognitive thoughts relating to the process of deciding what to do next. They also reported a number of thoughts not directly related to actual problem solving or actions. These reports were mostly indicative of being ‘stuck’ (e.g., “ummm…” or “I don’t know”) and in some cases even included mild profanity, the inclusion of which would suggest that participants had taken seriously the request not to censor their reports, or had become largely oblivious to the presence of the researcher.

Finally, participants were reminded of the aim of the research and of what they would be asked to do next time. In some previous research (Afflerbach 1990) participants have been asked to consider their thinking for a time prior to actual data gathering as a way of increasing participants’ mindfulness of their thinking and therefore increasing their ability to report during think-alouds. In the current study participants were given the instruction to “pay attention to (their) thinking while reading” over the period (approximately one week) between this initial meeting and when they would next meet individually with the researcher.

Collection of concurrent verbal protocols

At the next meeting, participants met with the researcher individually. Once participants were seated and indicated they were comfortable and ready to proceed, they were reminded of the task. Instructions were based on a script in order to ensure consistency, although the actual instructions given to each participant varied slightly as participants were encouraged to ask questions or seek clarification as they felt necessary. Participants were then given a short warm up passage to read. In order to increase reliability (Ericsson and Simon 1993), these warm up materials were of a similar length to the experimental materials and were presented in the same form. Participants knew this was a warm-up and that the researcher would ask them to stop reading once it appeared they were reporting their thoughts regularly. During the warm-up exercise the researcher provided verbal reminders and prompts to the participants. Such prompts were kept neutral and as short as possible, (e.g., “Keep talking” or “Remember to think out loud”) in order to reduce the risk of intruding into participants thinking and to avoid cueing particular responses (Taylor and Dionne 2000). The overall aim of the warm up activity was to reduce the need for researcher input during the experimental task.

In the data collection phase, participants firstly received a page that gave them a choice of texts to read. Participants were then asked to think aloud from the moment they received this selection sheet and to report everything they were thinking while selecting the text they wished to read and from that point on. The researcher needed to give some prompts to remind participants to verbalise their thinking at this point. The prompts were kept as neutral as possible; variations on “remember to keep talking” were used. Once participants had made their choice, the researcher gave them the text they had chosen and given a final reminder to “Please say everything that comes into your head”.

Participants read out loud while reading the texts. A few of the participants required further reminders to think out loud at some point during the reading of the text. The same prompts used in the warm up were used here. Again, these were kept neutral so as to avoid prompting particular responses or encouraging participants to report when there was nothing to report. Particular circumstances led to the use of the prompts. Participants were given a reminder if they had completed an entire paragraph and had not reported anything (other than reading out loud), where they had paused in their reading but were not reporting any thinking, and where their body language (facial expressions, eye movements etc.) suggested confusion or engagement with the text but nothing was being reported.

At the conclusion, participants were asked to give a brief summary of their understanding of the text, this was primarily to look for consistency between their understanding and the verbalisations given earlier. It also provided an opportunity for the researcher to judge whether the participant’s processing of the text had been successful in developing a full understanding of the text. The entire process was recorded using a camera connected to a laptop.

Collection of cued retrospective reports

Immediately following the collection of the concurrent verbal protocols, the participant and the researcher moved to be in a position where they could both see the video. Instructions were given that made it clear there were three available approaches; the researcher could ask the participant for clarification of something they had reported earlier, the researcher could ask about periods where there appeared to be a lack of reporting (particularly if the participant appeared to be thinking), or the participant could pause the video in order to add to what they had already reported without waiting for the researcher to prompt them. Participants were told that they should say anything that occurred to them as they no one else really knew what they had been doing while they were reading.

Following these instructions, the video of the just completed reading task was replayed. During this part of the process there was a mixture of participant- and researcher-led interactions, some participants quickly began to add significant amounts of information to their accounts, while others were more likely to respond to researcher prompts. During this process, participants were able to refer to the text they had read, and several made use of this to put their thoughts into context. This was particularly true where the researcher asked for clarification on a particular point. Participants would often refer to the text and the sections immediately preceding that they were referring to at the time. Because participants had read aloud it was an easy process to connect their verbal reports to the portion of text they were reading, and subsequently for them to put their reports into context when clarifying.

In accordance with the recommendations of Taylor and Dionne (2000), any prompts that were used were again kept as general as possible, particularly initially. There was a focus on asking very general “what” questions such as “Can you tell me what you were thinking about there?” and “What did you mean?” (Rather than “Can you tell me how you decided to do that?” or “Did you mean…?”). Prompts and questions sometimes became more specific following these general ones in the context of following a line of reporting initiated by the participant, either spontaneously or as a result of a general probe.

Data analysis

As discussed earlier, a coding scheme was initially used to code the verbal protocol data from this study. During the process of coding the data it became apparent to the researcher that while the information was useful in terms of specific aspects of reading, much was being lost in the process of coding. One reason for the use of both concurrent and cued retrospective reports was to increase the richness of the data gathered. Breaking the data down into specific codes led to much of the richness becoming hidden. Recording the codes and the process used by readers is useful and interesting, as well as making the quantity of data gathered manageable for analysis. However, each reader had their own way of approaching and managing the task of developing reading comprehension and the coding process hid this. The process of dissembling and reassembling the data using codes (Yin 2011) yields the ability to view patterns across multiple individuals within the analysis, but loses the individual voices of those individuals at the same time. There is a tension between maintaining the richness that is the aim of verbal protocol analysis and reducing the volume of data to manageable chunks (Gu 2014; Evers 2011). In this paper, the aim is to capture the individualistic nature of reading comprehension and to outline some of the reading styles employed by individual readers so coding is not only unnecessary (Gu 2014), but would be counterproductive. The use of narratives has been employed as a way of maintaining the richness of the data and to understand the reading process as it is for the individual readers who were participants in the study.

This is a descriptive study, and therefore the data is used to tell the story of the participants’ experience of reading (Taylor et al. 2015). Narrative descriptions of each participants’ text reading were developed by combining the information gathered from both the concurrent and retrospective verbal protocols into a longitudinal description of the reading event. As well as simply including the behaviours, other information from the participants such as feelings of confusion or frustration are also included to give a fuller description of the participants’ experience of reading.


All readers in the study set some form of goal when selecting the text to be read, in most cases (and in all cases for the HC group) this was related to interest levels and/or a meaning related goal, which may have been a specific question or a more general inquiry of the text. For the LC group, the majority of goals were related to accuracy of reading rather than comprehension. The form of these goals was reflected in some aspects of reading behaviour, particularly in connection to readers’ responses to decoding errors. Those focussed on accuracy attempted to correct every error they made, while those focussed on understanding ignored most errors that did not affect meaning, often without even realising they had made an error, but at other times consciously deciding not to worry about it.

Once they began reading, each participant had a default strategy connected to their definition of the reading task (McKoon and Ratcliff 1992) that they used without further consideration, with some commenting “I just (did) what I always do”. In most cases, this approach varied for different text forms. Although there was a predominant method for developing understanding, e.g. visualisation, other strategies such as inference and prediction were used as part of overall process of developing understanding. However, if a text did not fit their expectations then another approach may be used, “If (the text) looks strange I might (think about what to do).”

There were seven different styles identified, four successful and three that were not associated with good comprehension. Each style is defined primarily by two things; the first is the automatic comprehension strategy that is used, and the second is the deliberateness of the actions. The four successful styles came in pairs, each pair was connected by the strategy being used and the two styles within each pair were distinguished by the nature of the actions involved.

The first pair of successful styles used questioning as the primary comprehension strategy. The first is termed the “Wondering” approach. Readers using this style utilised a questioning strategy and asked themselves a number of questions as they read. The questions asked were not used to prompt a search for related information, rather they reflected the reader wondering what will happen next, or developing or testing of a hypothesis (“I wonder if…”). The questions provide focus for the reading and are used by the reader to monitor their understanding. A second style also makes use of questioning as a strategy. The “Hunter Seeker” asks more specific questions, but the major difference is in the intentionality of the approach to answering them. In contrast to the “Wonderer”, the “Hunter Seeker” deliberately sets out to answer their question(s). They read quickly through the text, searching for key words or another signal that indicates when they may be able to answer the question. They are less interested in the text as a whole and more in seeking an answer. This strategy was primarily seen when reading non-fiction texts but was also observed in one case of fiction reading. It is perhaps not surprising that questioning was a relatively common strategy, given its prevalence in schools, but participants in this study were reading as they would for their own pleasure. Simply asking a question and seeking an answer does not reflect what many would think of as reading for pleasure. Wilhelm and Smith (2016) draw on the work of Dewey and identify the intellectual pleasure or satisfaction gained from this sort of reading as providing value to the reader.

The second pair of styles focuses on visualisation strategies. One is termed the “Movie maker” and the other “Movie watcher”. While both described watching a movie, different participants arrived at this position in different ways. The movie maker could be considered as analogous to a film director. Individuals using this style actively used the information in the text to build their own movie, looking for the information they needed and including predictions as to what the movie was going to look like. At times the movie they were constructing appeared to be predictive and ahead of the text they were reading, this enabled them to monitor their success through checking whether or not the movie in their head matched what they were reading. The Movie watcher style was less proactive. While still involving a movie, the involvement of the reader in this style is more like that of the audience. The reader is watching a movie and waiting to see what will happen next. In this style the movie was built from the text being read and was not predictive of upcoming text. Both of these styles making use of visualisation strategies can be viewed as creating mental simulations of the information contained in the text (Glenberg 2011; Kiefer and Pulvermüller 2011) as a means to developing understanding. The result of reading in this way is what might be referred to in Deweyan terms as immersive play pleasure and the pleasure gained from social interactions with others or the characters themselves (Wilhelm and Smith 2016).

The three styles that resulted in poor comprehension all showed lower levels of involvement from the reader. The first of these styles (“Passive”) was observed in a high comprehender as well as low comprehenders. This style can be characterised as bored reading, particularly in the case of the good reader who exhibited this style on one occasion. At first glance it may appear to be good reading as the reading is largely fluent. There is little or no attempt to make sense of the text however, the general impression is of a reader reading without purpose and waiting for the text to end. The other two styles associated with poor reading are different, in that the reader is relatively active, but is no more focussed on understanding. In the first of these, the “Fluent decoder” the reading is again fluent, but the reader is focussed on reading accuracy. They stop to correct mistakes that may be ignored by a reader focussed on comprehension, and miss errors that look correct but would affect meaning. The fluent decoder is focussed on a goal (accuracy) and is sufficiently skilled that they can read fluently, and understand that this is important. The “extreme decoder” is focussed on accuracy to the exclusion of all else. They read word by word, attempting to ensure each is pronounced correctly with no real attempt to understand what they are reading, neither is there a concern with the fluency and phrasing of their reading.

Excerpts from descriptive narratives

This section provides some excerpts to illustrate the narrative type descriptions of the reading styles above. The full narratives are an attempt to capture the flow of the reading and to provide an illustration of how individuals approached their reading. The excerpts here are to provide some examples of what that might look like as the space restrictions of an article do not allow for the inclusion of the full narratives which were up to 1500 words in length. The examples presented here are intended to be representative and cannot cover all aspects of every style. They should not be read as being a full description of each reading style. Each participant has been given a pseudonym, and illustrative excerpts have been chosen, therefore the description of a single approach may include excerpts from more than one participant.

Questioning-based styles: the “wondering” approach

Pre reading

…Michelle began by reading the titles and précis of the choices and then attempted to “weigh one against the other” in order to decide which was more appealing. Once she had selected a text based on her level of interest she set an idea goal for her reading that related to “getting to know” the characters in the story… she set a goal for reading that was an ‘idea goal’ in that it was related to the content or ideas in the text rather than a ‘level goal’ related to how well she wanted to understand the story. This did not take the form of a specific question to be answered but was more general in nature “I wonder what’s going to happen?”.

…Sarah considered the possible choices of text separately. She read the title and précis of the story and quickly considered what each one might be about. She indicated in the review that she wasn’t sure what the stories were about, but that she chose the one she did because it attracted her interest more than the others; she wanted to “know more about it”.


…As she began to read the story, Michelle almost immediately began to ask herself questions about the text she was reading. Some of these questions were essentially repeats of her initial wondering about the story and what was coming, others were more specific and related to particular incidents in the text “Why would you need a screwdriver?” while still others, also related to specific events, appeared to be used as a strategy to focus Michelle’s attention on a perceived lack of understanding. Michelle indicated in the review that she doesn’t generally attempt to answer these questions at the time but waits to see, “I look for answers as I’m reading”. These questions came frequently throughout her and appear to the observer to be quite intrusive. For Michelle however, they were anything but, helping her to focus and to move through the text in a purposeful fashion.

…As the story drew to its conclusion, Michelle began to find the answers to her wonderings “Oh, it’s a lizard, that’s what he should have told her.” She appeared more relaxed and was smiling as she read the last few paragraphs. As she finished, she disengaged from the text using her default strategy (“I wonder what it is?”) considering what the story’s protagonist might be planning.


…Michelle then spent some time disengaging from the story. She indicated in review that she was considering earlier parts of the text and “putting some of it together better” and that she was now able to understand much of the text better once she knew the ending.

Questioning-based styles: the “hunter seeker”

Pre reading

…Fiona quickly read the titles and précis for the three texts… Having selected the most interesting text, she was aware that she wanted to understand the text well as it had aroused her interest. She also set an idea goal relating to getting to know the characters and their relationships.


…Once she started to read Andie read fluently for 18 seconds before asking a question about two characters (“Are they friends?”) before deciding that they were and continuing.

…Andie indicated that she only really paid attention to parts of the text that related to her goal “otherwise it doesn’t really matter.” After 61s a new character entered the story and this prompted Andie to ask, “How does she fit in?”. This question was followed by a period of less fluent reading lasting six seconds while she attempted to answer her question. The next use of her questioning strategy immediately followed this period of less fluent reading and was a more specific question “So are they friends?” Andie said in review that she had “pretty much decided they were” but was “sort of checking by asking myself to look again.”

… Andie indicated in the review that she felt she had answered it successfully but that she was also aware that the end of the story was getting closer and that she needed to make sure she had everything “sorted out” before it finished.


…Andie completed the text. She indicated in the review that she felt she had read the text successfully in that she had a good feel for the characters and their relationships. She also spent a few seconds considering the possibilities stemming from a cliff-hanger ending before completely disengaging from the text.

Visualisation-based styles: the “movie-maker”

Pre reading

…Stevie began by selecting the text to read based on her interest levels. She then defined the task at hand. She did this in two ways, firstly she defined the task as interesting and secondly she defined it by text type, in this case a narrative. Immediately prior to beginning reading she set an idea goal. This was not a specific goal but was general “I want to find out what it’s about”.

…Rose began by selecting the text to be read based on interest, commenting “I want to know more about that one, it sounds funny ‘cause it says ‘locked in the loo’.” In review Rose indicates that she is already thinking about what that might look like and beginning to form a picture in her head.


…During reading she laughs in response to the text before continuing reading. In the review Rose said, “When the toilet paper drops it’s like you’re there watching the thing.” Later in the text there are other indications that when her reading is successful Rose places herself in the position of the main character. Rose’s next comment is “That’s like my Nana and Granddad’s one”, and indicated in the review that this was again as if she was there looking at it.

… carrying this out was a very deliberate activity; each piece of new information was carefully assimilated into the movie and assessed for whether it fitted with the information already contained in the movie. There was also an almost constant stream of comments or responses to the movie (e.g. “Sounds scary”, “Oh my god!”). During her reading there were a number of decoding errors, some were ignored or not even noticed if they are not relevant to the movie, while others were corrected if they impacted on her ability to make the movie clearer.

…After about one and a half minutes of this, Stevie identified a discrepancy in her level of understanding, … she was unable to assimilate a new piece of information into her movie. As a result of this discrepancy, Stevie evaluated the success of her default strategy negatively and disengaged from the text … In this instance she decided to use a different strategy (questioning) to enable her to resolve the issue with her movie. After its use she decided that this strategy had been sufficiently successful for her to continue reading with her original default strategy. Interestingly, she did not find an answer to her question, Stevie indicated that what the question did do was enable her to define the part of the movie that was incomplete so that she could carry on with “a blurry bit” to be filled in later if needed or possible.


…As she finished reading with a clear movie in her head, the goal was achieved. Stevie then reviewed her movie; she indicated that during this time the abbreviated movie included those sections during which she was using an alternative strategy while reading the text. Stevie completed her disengagement from the story by relating this story to one she had previously read “That was a bit like…”.

…On completion of the text, Rose reflects on the fact that she “quite liked that story” and thought back through the story and remembering some of the movie scenes in some detail and referring to those reminded her of events and places in her own experience.

Visualisation-based styles: the “movie watcher”

Pre reading

…Sally identified the one that she found most interesting and selected that for reading. Sally set herself an idea goal that took a reasonably general form “I wonder what sort of trouble the magpie will get in to?” As part of her preparation for reading she makes a prediction as to what the answer to this question might be, and plans to use questions to check and see if she’s right. She also has a level goal in mind, although this is not connected to an explicit idea but is qualitative, “I want to understand it well because it seems interesting.”

…As she begins reading Sally begins to visualise the story. She does this automatically and without a deliberate decision to do so. Sally indicates in review that the movie she develops is not consistently clear, there are parts that are more in focus than others. As she reads Sally asks herself questions. These questions are not used to look for answers but are the mechanism used to focus her attention on relevant parts of her movie. In this way the questions she asks are more like the comments that other participants who used visualisation strategies made. They are not questions that require an answer but merely serve to draw her attention to aspects of the story. Sally also makes comments (e.g. “Eew!”) and indicates in the review that she does not see them as different; rather the questions are more like comments.

…after 31s of reading she pulls a face at something in the text without pausing in her reading. After a further four seconds there is a brief pause, in the review Sally indicated she “just needed some time to build up” her picture at that point. This was followed by another 31s period of fluent reading that included a number of brief indications of monitoring such as laughs or face-pulling. Sally then encountered another difficulty when she couldn’t reconcile a particular phrase from the text with her visualisation. Here again she chose to pause her reading and decided to make a prediction of what was coming in the text and match her picture to that so she could check whether she was right. After two seconds of reading she decided that she appeared to be correct and made another prediction before continuing reading.

…At the conclusion of the story however, Julia had to pause in her reading owing to an unexpected ending, saying “That’s funny, because I thought…” She indicated in the review that she needed to “rearrange” her movie scene to cope with the unexpected nature of the conclusion. This process took about five seconds. Once she had the difficulty resolved Sally read for a further 15s to finish the text.


…Once completed Sally looked away from the text for a second and appeared to be finished with it. She then looked back to the text however, smiled and considered her final picture before predicting what the scene would look like and picturing that. Following this final visualisation, Sally disengaged from the text.

Accuracy-based styles: “passive”

Pre reading

…Victoria began by reading the précis for all texts. She indicated in the review that having done this and found that none of the available texts caught her interest she had already defined the task as being of little interest and therefore one to be completed as quickly as possible. She chose apparently based on interest saying, “That one seems more interesting.” However, she indicated in the review later that it was actually a case of choosing the least uninteresting, and that a more accurate interpretation was that the one she had selected appeared less boring. Victoria stated that under other circumstances she would have been unlikely to read any of the texts available.

…Having defined the task and selected a text Victoria set a goal for her reading based on her intended level of understanding. This goal might be referred to as an ‘anti-goal’ in that she was clear that she didn’t care whether she understood the text or not. This meant that there was no motivation for her to do any real work towards developing her understanding of the text. Aware of her lack of interest and motivation, Victoria began to read.


…The first 26s of her reading were not particularly fluent, she appeared obviously bored and uninterested in the text. Victoria then continued reading fluently without interruption for a further one minute and 20s. During this time there were two periods of less fluent reading (1 and 2s respectively) when Victoria encountered an unfamiliar word. After a total of one minute and 46s Victoria ceased reading and looked around. In her retrospective report, she indicated that she had not been thinking about the text at all during this four second break and was effectively looking for something more interesting. She also said that she would normally have discarded the text at this point (if she had started it at all) and considered doing so here.


…When finished, she sat up, put the text down and said “Done!” This was the full extent of her disengagement from the text. Victoria indicated she felt she had completed her task successfully as she had done what she wanted. She knew that she didn’t really know what the text was about.

Accuracy-based styles: the “fluent decoder”

Pre reading

…James began by selecting a text based on interest (“Because I like drawing”). Although he did not report this at the time, James said in his retrospective report he wanted to see what sort of drawing the character might do. Once James had the text he made note of the length of the text and pulled a face. In review James indicated that he thought it looked “a bit long” (the text was approximately one and a half pages) and therefore might be hard.


…a sustained period of fluent reading (two minutes and 21s) with only a few pauses to decode difficult words. James’ phrasing and reading of punctuation is good. His fluent reading sounds like that of a good reader, except it is not broken by comprehension-related processing. In fact, because of the lack of processing, James’ reading sounds better than most of the cases in the HC group. After this period of fluent reading, James encountered two names in the text. He spent 22s attempting to decode these two words, before carrying on once he was satisfied.


…James does not actively disengage from the text, but simply puts it down and waits. When asked, James has forgotten that he was interested in finding out about the drawing…

Accuracy-based styles: the “extreme decoder”

Pre reading

…Sophie selected a text, saying she hoped that it would be easier to read.


…She began reading the text in a staccato fashion that paid little or no attention to punctuation and phrasing within the text. There were a number of brief interruptions where she said “Oh, that’s not right, go back” or similar. These comments were related to the accuracy of her decoding and all occurred immediately following an error. After each of these interruptions, Sophie re-read the word in question and attempted to correct her error. She only reread the individual word, never a longer phrase, and was sometimes satisfied with a ‘correction’ that appeared visually accurate but was still erroneous.


…although she felt she had read well, Sophie was unable to give even a brief idea of what the text had been about. She indicated that this was not unusual and that when she is reading she is focussed on individual words and getting them correct. If she needs to (i.e., if asked by a teacher or required to by a test) she will go back and look for information after she has finished, but she does not consider meaning while reading.


Studies of reading strategy use that have emphasised individual strategies have greatly increased our understanding of good readers. However, they may have led to a perception that the use of these strategies is somewhat separated from the reading itself and that readers make use of comprehension strategies in order to “fix” a problem with their understanding. Many of these studies have also (at least) implied that individual readers make use of a range of strategies. This article addresses that implication. The results of this study suggest that individuals do not make use of a range of strategies in a single text, but rather use a limited number of strategies (usually 1 or 2) that they know to be effective in a particular context.

The narratives displaying good reading and the comments of some of the participants discussed earlier make it clear that good readers have particular, and individual, ways of making sense of what they are reading. For some participants, particularly those making use of visualisation strategies, this is an immersive experience. Some reported not even really being aware of reading. The experience for these readers was one of seeing and hearing the text rather than of reading words and constructing meaning. In addition to this different view of strategy use, it is clear that good readers do not generally select from a range of strategies as needed during reading. They identify the strategy they will use even before they begin reading and the selection process itself may not have been conscious. As reflected in some of the comments reported above (e.g. “I don’t usually think about it much, I just do what I always do.”). This is particularly true where the task is familiar to the reader and is therefore accompanied by a default strategy (e.g. McKoon and Ratcliff 1992). Only where there was a reading difficulty to overcome did strategy selection become conscious during reading itself. During the reading of a text, readers only use a small number of strategies appropriate to the text and their goals. Individual readers have their own preferences and strategies they rely on. The participants in this study used a range of comprehension strategies in the course of their reading, but individuals used a single strategy and sometimes a second if they had to.

There is discussion in the literature on reading for pleasure that covers similar ground to the idea of reading being an immersive experience. Parry and Taylor (2018) use the example of the character Meggie in Inkspell (Funke 2005) to illustrate the fact that meaning resides not with the author alone, but in the reader’s experience of a text. The immersive nature of reading illustrated in the narratives presented here connects directly to the engagement and motivation of readers (Parry and Taylor 2018). Many countries are experiencing decreases in rates of reading for pleasure (Garces-Bacsal et al. 2018; McNaughton 2020) and connections have been made to the increasingly utilitarian expectations of reading in schools as a result of standards-based assessment measures. If good reading is the immersive experience it seems to be then teachers need to know how to assist the development of that kind of reading, and not simply encourage strategy use for the purpose of achieving a set purpose.

The development of the narratives of individual reading styles provides a useful tool for understanding the experience of readers with different styles as well as the process they use for developing their understanding and the way they control the skills and strategies at their disposal. The real benefit of the narratives as a tool is likely connected to teaching and learning. There are two major issues relating to the effective teaching and learning of reading comprehension. We know that although students receive tuition in the use of reading comprehension strategies many do not develop the ability to use them independently (McNaughton 2007). We also know that good modelling by teachers is one of the most effective ways to develop metacognition in students (Louden et al. 2005) and that effective modelling is difficult to develop in teachers (Kennedy 2014). The major potential impact of the development of the narratives of individual reading styles is in the ability for them to be used as a practical tool to assist students to understand the reading process, and for teachers to be able to both understand the process as it is actually experienced by readers and to develop their ability to model it for students.

It is unlikely that students will receive good modelling from their teachers if those teachers are either not used to thinking about the underlying processes involved in controlling the development of reading comprehension, or are unfamiliar with the processes used by good readers in a range of different styles of effective reading. There is a risk that teachers (who are presumably good readers themselves) will not see their style is as one way to read successfully, but as the way. A misunderstanding of this sort would lead teachers to model only one style that may or may not be appropriate for the students they are working with. The identification of the individual reading styles in the current study and their illustrations using the narratives provide an opportunity to show teachers how good young readers read, and to encourage them to practise reading in a variety of different styles themselves in order that they can then model those styles to their students. It is noted that we have known a lot about what good readers do for some time, and yet this has not transferred into the classroom in a meaningful way. The various routines (Duke and Pearson 2002) for teaching reading comprehension have their place as methods for teaching students how to use a variety of strategies, but the critical issue has been shown to be the teaching of self-regulation of those strategies. Professional development for teachers that enables them to effectively model different reading styles to students (using examples such as those from the narratives) may be the key to maximising the benefits from existing methods of instruction and to developing truly self-regulating readers. This study gives us some indication of the ways in which individuals make use of the reading comprehension strategies they have at their disposal, and that should be included in instructional design.

For good readers, strategy use is not something that good readers do to help them with their reading, it is their reading. This needs to be encouraged in the course of their journey to becoming good readers.

Data availability

Datasharing is not applicable to this paper.


  1. Afflerbach P (1990) The influence of prior knowledge and text genre on readers’ prediction strategies. Journal of Reading Behavior 22:131–148

    Article  Google Scholar 

  2. Baker L (2002) Metacognition in comprehension instruction. In: Block CC, Pressley M (eds) Comprehension instruction: research-based best practices. Guilford Press, New York, pp 77–95

    Google Scholar 

  3. Block CC, Pressley M (2001) Comprehension instruction: research-based best practices. Guilford Press, New York

    Google Scholar 

  4. Conley M (2016) Improving adolescent comprehension: developing comprehension strategies in the content areas. In: Israel SE, Duffy GG (eds) Handbook of research on reading comprehension, 2nd edn. Guilford Publications, New York, pp 406–427

    Google Scholar 

  5. Creswell JW (2007) Qualitative inquiry and research design, 2nd edn. Sage, Thousand Oaks, CA

    Google Scholar 

  6. Darr C, McDowall S, Ferral H, Twist J, Watson V (2008) Progressive achievement test: reading. NZCER Press, Wellington, NZ

    Google Scholar 

  7. Davis AJ (2016) Teaching reading comprehension , Second edition. Eleanor Curtain Publishing, Hong Kong

    Google Scholar 

  8. Duke NK, Pearson PD (2002) Effective practices for developing reading comprehension. In: Farstrup AE, Samuels SJ (eds) What research has to say about reading instruction, 3rd edn. International Reading Association, Newark, Del., pp 205–242

    Google Scholar 

  9. Elley W, Croft C (1989) Assessing the difficulty of reading materials: the noun frequency method, Revised. New Zealand Council for Educational Research, Wellington

    Google Scholar 

  10. Ericsson KA, Simon HA (1993) Protocol analysis: verbal reports as data, Revised. MIT Press, Cambridge, MA

    Google Scholar 

  11. Evers, J. C. (2011). From the past into the future. How technological developments change our ways of data collection, transcription and analysis. Forum: Qualitative Social Research.

  12. Funke C (2005) Inkspell. Chicken House, Frome Somerset

    Google Scholar 

  13. Garces-Bacsal RM, Tupas R, Kaur S, Paculdar AM, Baja ES (2018) Reading for pleasure: whose job is it to build lifelong readers in the classroom? Literacy 52(2):95–102.

    Article  Google Scholar 

  14. Gaskins IW, Satlow E, Pressley M (2007) Executive control of reading comprehension in the elementary school. In: Meltzer L (ed) Executive function in education. Guilford Press, New York, pp 194–215

    Google Scholar 

  15. Glenberg AM (2011) How reading comprehension is embodied and why that matters. International Electronic Journal of Elementary Education 4(1):5–18

    Google Scholar 

  16. Gu Y (2014) To code or not to code: Dilemmas in analysing think-aloud protocols in learning strategies research. System 43:74–81

    Article  Google Scholar 

  17. Israel SE (2015) Verbal protocols in literacy research: Nature of global reading development. (Routledge research in educational psychology.). Routledge, New York

    Google Scholar 

  18. Jacobs JE, Paris SG (1987) Children’s metacognition about reading: issues in definition, measurement, and instruction. Educational Psychologist 22(3 & 4):255–278

    Article  Google Scholar 

  19. Kaplan A (2008) Clarifying metacognition, self-regulation, and self-regulated learning: what’s the purpose? Educational Psychology Review 20:477–484

    Article  Google Scholar 

  20. Kennedy E (2014) Raising literacy achievement in high-poverty schools. Routledge, New York

    Google Scholar 

  21. Kiefer M, Pulvermüller F (2011) Conceptual representations in mind and brain: theoretical developments, current evidence and future directions. Cortex 48(7):805–825

    Article  Google Scholar 

  22. Klingner JK, Vaughn S (1999) Promoting reading comprehension, content learning, and English acquisition through collaborative strategic reading (CSR). Reading Teacher 52:738–747

    Google Scholar 

  23. Louden W, Rohl M, Barrat-Pugh C, Brown C, Cairney T, Elderfield J et al (2005) In teachers’ hands: effective literacy teaching practices in the early years of schooling. Australian Government Department of Education, Science and Training, Canberra, ACT

    Google Scholar 

  24. McCrudden MT, Kendeou P (2014) Exploring the link between cognitive processes and learning from refutational text. J Res Reading 37(1):116–140.

    Article  Google Scholar 

  25. McKoon G, Ratcliff R (1992) Inference during reading. Psychol Rev 99:440–466

    Article  Google Scholar 

  26. McNaughton, S. (2007). Cultural contexts for schooling improvement. In: Redesigning pedagogy: culture, knowledge and understanding. National Institute of Education, Singapore

  27. McNaughton S (2020) The literacy lanscape in Aotearoa New Zealand. Office of the Prime Minister of New Zealand, Wellington

    Google Scholar 

  28. National Reading Panel (2000) Teaching children to read: an evidence-based assessment of the scientific literature on reading and its implications for reading instruction. National Institute of Child Health and Human Development, Bethesda, MD

    Google Scholar 

  29. Parry B, Taylor L (2018) Readers in the round: children’s holistic engagements with texts. Literacy 52(2):103–110.

    Article  Google Scholar 

  30. Phillips LM (1988) Young readers’ inference strategies in reading comprehension. Cognition and Instruction 5(3):193–222

    Article  Google Scholar 

  31. Pressley M (2002) Metacognition and self-regulated comprehension. In: Farstrup AE, Samuels SJ (eds) What research has to say about reading comprehension, 3rd edn. International Reading Association, Newark, Del, pp 291–309

    Google Scholar 

  32. Pressley M, Afflerbach P (1995) Verbal protocols of reading: the nature of constructively responsive reading. Erlbaum, Hillsdale, NJ

    Google Scholar 

  33. Pressley M, Gaskins IW (2006) Metacognitively competent reading comprehension is constructively responsive reading: how can such reading be developed in students? Metacognition and Learning 1:99–113

    Article  Google Scholar 

  34. Taylor KL, Dionne J-P (2000) Accessing problem-solving strategy knowledge: the complementary use of concurrent verbal protocols and retrospective debriefing. J Educ Psychol 92(3):413–425

    Article  Google Scholar 

  35. Taylor SJ, Bogdan R, DeVault M (2015) Introduction to qualitative research methods: a guidebook and resource. Wiley, Hoboken

    Google Scholar 

  36. Vacca RT (2002) From efficient decoders to strategic readers. Educational Leadership 60(2):6–11

    Google Scholar 

  37. Wilhelm J, Smith M (2016) The power of pleasure reading: what we can learn from the secret reading lives of teens. English J 105(6):25

    Google Scholar 

  38. Winne PH (2001) Self-regulated learning viewed from models of information processing. In: Zimmerman BJ, Schunk DH (eds) Theories of self-regulated learning and academic achievement: an overview and analysis, 2nd edn. Lawrence Erlbaum Associates, Mahwah, NJ, pp 153–189

    Google Scholar 

  39. Yin RK (2011) Qualitative research from start to finish. Guilford Press, New York

    Google Scholar 

Download references

Author information



Corresponding author

Correspondence to John Milne.

Ethics declarations

Conflicts of interest

None declared.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Milne, J. Individual reading styles: a narrative approach to understanding reading behaviour. SN Soc Sci 1, 48 (2021).

Download citation


  • Reading comprehension
  • Narrative
  • Reading style
  • Reading strategy
  • Metacognition