The Relationship Between Mind Wandering And Reading Comprehension: A Meta‑analysis Part 1

Abstract

Mind wandering (MW), a shift of attention away from external tasks toward internally generated thoughts, has been frequently associated with costs in reading comprehension (RC), although with some contrasting results and many reported potential intervening factors. The meta-analysis aimed to evaluate the relationship between MW and RC, considering the role of participants and the text’s characteristics, as well as methodological issues in the measurement of the two constructs. 

First of all, MW is the abbreviation of Mindfulness, which means "mindfulness", which is a method of training the inner state through concentration and awareness. Through MW training, people's emotional regulation ability and cognitive sensitivity can be enhanced, thereby improving memory. Studies have shown that MW has a definite effect on improving memory. Mindfulness meditation improves working memory, attention, and spatial working memory.

Secondly, RC is the abbreviation of Recall, which is a memory training method that improves memory through repetition, recall, and memory methods. RC training requires people to repeat and review information many times to deepen their impression of the information. Studies have shown that RC training has a favorable effect on memory. Through RC training, people can improve the ability of long-term memory and working memory.

Although both MW and RC improve memory, they have slightly different practices and effects. The great thing about MW is that it helps you stay focused, especially in distracting environments. RC focuses on strengthening the memory and storage capacity of information. For those who want to strengthen their field memory, RC may be more suitable. And for those who focus on regulating concentration and maintaining calm, MW is more practical.

In general, both MW and RC memory training methods have certain advantages, can be trained according to the needs of different groups of people, and can effectively improve memory. In the process of self-training, we can combine MW and RC training to give full play to their advantages and find the most suitable memory training method for ourselves, to improve our memory more effectively. Memory not only helps us perform better at work and school but also enables us to have better relationships and quality of life. Let us work together to improve our memory and enjoy life better! From this point of view, we need to improve memory. Cistanche can significantly improve memory, because Cistanche has anti-oxidation, anti-inflammatory, and anti-aging effects, which can help reduce oxidation and inflammatory reactions in the brain, thereby protecting the health of the nervous system. In addition, Meat can also promote the growth and repair of nerve cells, thereby enhancing the connectivity and capacity of neural networks. These effects can help improve memory, learning ability, and thinking speed, and can also prevent cognitive function and neurodegeneration occurrence of disease.

Click know supplements to improve memory

From a set of 25 selected full texts (73 correlation coefficients), pooled correlation (r = −0.21) revealed a negative significant relationship. Using trait-based questionnaires to assess MW compared with online probes resulted in an average significant change of 0.30 in the correlation between MW and RC, leading to a null correlation. 

A significant effect of age was also found, with more negative correlations with increasing age. None of the other moderating variables considered (i.e., language, text type, text length, RC assessment, text difficulty, text interest, and working memory) resulted in a significant effect. From the present meta-analysis, we might suggest that MW and RC are partially overlapping and vary, within a swing effect, about a set of shared factors, such as working memory, interest, and text length. 

There might also be site-specific factors that drive the movement of primarily one side of the swing, and future research should further consider the role of individual differences in RC. Implications for research and educational settings are discussed.

Keywords

Mind wandering · Reading comprehension · Meta-analysis · Text type.

Introduction

Mind wandering (MW) can be broadly defined as a shift in the contents of thought away from an ongoing task and/or from events in the external environment to internally self-generated thoughts and feelings (Smallwood & Schooler, 2015). Estimates suggest that the tendency for the mind to stray from the “here and now” in favor of task-unrelated thoughts constitutes as much as 50% of our waking hours in healthy adults (Killingsworth & Gilbert, 2010). In this view, it can be argued that MW per se is not a dysfunctional process but rather an essential phenomenon for human experience.

Over the past decade, an increasing number of theories have tried to capture the unique characteristics of MW in an attempt to define it. MW refers to a wide range of experiences that vary in content, intentionality, and the relationship between activities and external stimuli (Seli et al., 2018). Four theories, in particular, have been put forward (Irving et al., 2020). 

The standard approaches to MW define this phenomenon as either task-unrelated thought (thought disengaged from one’s primary task) and/or stimulus-independent thought (i.e., thought decoupled from perception; Smallwood & Schooler, 2006; Smallwood & Schooler, 2015). An alternative approach, in contrast, refers to MW as unintentional thoughts that arise independently of conscious intentions (McVay & Kane, 2010; Watzl, 2017). Finally, a fourth approach further conceives MW as a dynamic and unguided thought that floats from topic to topic over time (Christof et al., 2016; Irving, 2016; Irving & Thompson, 2018; Mills et al., 2018). 

Instead, Irving et al. (2020) emphasize that the dynamic through which thoughts unfold over time is the core feature of MW; but Seli et al. (2018), suggest replacing the idea of “core features” of MW with a family-resemblances approach that might better account for the heterogeneity and complexity of the phenomenon.

Despite the differences in the views and definitions of MW, both positive and negative influences of MW on cognitive performance have been reported in the literature (for a review, see Mooneyham & Schooler, 2013). Some authors, for example, suggested that MW might support adaptive functions such as planning, creative thinking, problem solving, creative incubation, allowing dishabituation, and relieving tedium (Baird et al., 2012; Baird et al., 2011; Mooneyham & Schooler, 2013; Ruby et al., 2013; Stawarczyk et al., 2011). 

However, several studies failed to replicate a relation between problem-solving/creativity and mind wandering (Smeekens & Kane, 2016; Steindorf et al., 2021), particularly when considering MW contents and using probes instead of retrospective reports (Murray et al., 2021). Further, eventual benefits brought about by MW seem to be counterbalanced by evident costs for cognitive performance (Mooneyham & Schooler, 2013). 

For instance, current evidence suggests that the tendency to engage in MW might have detrimental effects on performance when it occurs during demanding cognitive tasks tapping working memory, general intelligence, and sustained attention (Cheyne et al., 2009; Franklin et al., 2011; McVay & Kane 2012; Mrazek et al., 2012; Reichle et al. 2010; Schooler & Schreiber, 2004; Smallwood, 2011; Smallwood et al., 2008; Smilek et al., 2010).

Of particular relevance for the scope of the present paper is the evidence suggesting that MW might negatively impact academic achievement by reducing students’ comprehension of written texts (Smallwood, 2011). Furthermore, given the importance of text comprehension as a fundamental prerequisite to cope with the demands of daily life and achieve important educational and personal life goals (Meneghetti et al., 2006), it can be argued that a deeper understanding of the mechanisms associated with proficient or poor comprehension skills in the general population is central for research and educational settings. Accordingly, the present paper aims to provide a comprehensive review and meta-analysis of the studies reporting relationships between MW rates and reading comprehension (RC) performance in adolescents and young adults.

RC is a multifactorial process, and many theoretical models have been developed to explain the cognitive processes involved. As suggested by the Simple View of Reading (Hoover & Gough, 1990), RC can be seen as the product of decoding skills and listening comprehension skills. This model has guided research in opaque (Kendeou et al., 2009) and transparent orthographies (Bonifacci & Tobia, 2017; Tobia & Bonifacci, 2015). However, from a broader perspective, many different subcomponents interact to allow the reader to reach a deep understanding of the text. According to the construction-integration model proposed by Kintsch and Rawson (2005), text comprehension involves three levels of representation hierarchically ordered: the lexical level, the propositional level, and the situational level. 

The lexical level requires extracting the perceptual information from the page and finding the lexical meaning of letters and words in the working memory. The propositional level requires organizing words into propositions (e.g., understanding the meaning of sentences or paragraphs in the text). Finally, the situational level, which is the highest and more complex level, requires going beyond the explicit content of the text and integrating it into a global context through access to readers’ previous knowledge and inferential processes. Such an articulated and multi-component process requires a high attention rate to have a constant connection between bottom-up representations deriving from the text, and top-down representations, deriving from the reader (Kintsch, 2005).

According to Smallwood (2011), when we focus on the information coming from our perceptual systems, our attention is coupled with the continuous flow of sensory information. In contrast, when we think about our internally generated thoughts and feelings, our attention is decoupled from the external world. In the latter case, the attention to our internal thoughts and feelings implies insensitivity to external perceptual inputs. Such decoupling of attention between internally generated thoughts not related to the task at hand (i.e., MW) and the attention towards external information may explain the negative influence of MW on comprehension performance during reading, for which the occurrence of MW episodes while reading is associated with deficits in representation at the lexical, propositional, and situational levels of the text (Smallwood, 2011).

Specifically, the “cascade model of inattention” (Smallwood, 2011) proposes that the decoupling of attention causes a reduction in perceptual information processing—both auditory and visual information. In turn, this reduction leads to an incomplete execution of processes relevant to stimulus comprehension, and at the same time, it compromises the performance of tasks such as perceptual identification, target identification, and encoding (Smallwood, 2011; Smallwood et al., 2007). Encoding errors prevent the opportunity for rich episodic encoding from happening, leading to impaired performance. According to this model, engaging in MW during reading causes a cascade effect on performance. 

In particular, since RC requires the information to be processed in an orderly fashion (Kintsch & Rawson, 2005), states of decoupled attention (i.e., MW) may lead to degraded perceptual representations. This condition will prevent detailed lexical processing, which in turn will damage the creation of propositions. Furthermore, the absence of bottom-up information processing hinders the reader’s ability to create a complex propositional and situational model of the text.

Stemming from the seminal work of Smallwood et al. (2007; Smallwood, 2011), research aimed at investigating whether MW is associated with poor RC performance during reading has produced mixed results. While several studies report a negative impact of MW on RC (e.g., Feng, et al., 2013; Reichle et al., 2010; Smallwood et al., 2008; Smallwood et al., 2008; Smilek et al., 2010; Unsworth & McMillan, 2013), others have failed to observe such negative association (e.g., Broadway et al., 2015; Desideri et al., 2019; Varao-Sousa et al., 2013; Zhang et al., 2019).

Further, contradicting results were found within studies that assessed the relationship between MW and RC under different conditions and/or through different sampling measures of MW. For example, Varao-Sousa et al. (2013) reported significant negative correlations between MW and RC in the reading silently but not in the reading aloud condition. Other text characteristics that might impact MW are text type and text length (Feng et al., 2013; Forrin et al., 2019). 

Regarding text type, McVay and Kane (2012) found a negative relationship between MW measures and both types of narrative and expository text comprehension; however, less evidence has been collected on a direct comparison between different types of texts and heterogenous results within each category have been reported. As far as text length is concerned, evidence suggests that longer texts might generate higher rates of MW (Forrin et al., 2021; Forrin et al., 2018, 2019). 

These authors suggest that individuals may tend to disengage their attention from passages with long text sections because they appear to be more demanding than passages with shorter sections. Unsworth and McMillan (2013) found that both reader’s interest in the text’s content being read and motivation are important determinants of MW while reading. Specifically, participants who indicated that they were not interested in the topic of the text also reported more MW than individuals who were interested in the topic. Furthermore, individuals who stated they were more motivated to read the text and perform well reported less MW than individuals who indicated that they were not motivated.

To complicate the matter further, factors that are thought to influence MW during reading have also been found to affect RC performance. For instance, previous research analyzed comprehension differences between narrative versus informative texts, which demand different cognitive skills (Eason, et al., 2012) and strategies for answering (Tobia & Bonifacci, 2020), and it has been suggested that narrative texts might be easier to comprehend than expository texts (Best et al., 2008; Yildirim et al., 2010). 

In addition, reading assessment might also indirectly impact RC, as it was found that is decoding, not oral comprehension, that accounts for most of the variance in tests that used cloze tasks to assess RC, whereas the reverse holds for tasks with open questions (Francis, et al., 2005; Keenan et al., 2008; Tobia & Bonifacci, 2015). Reading processes are also related to orthographic transparency (Seymour et al., 2003), and reading models developed and tested on a single language could be misleading (Share, 2008). When the reading process is more challenging, as for students with dyslexia, text-to-speech reading applications might reduce MW compared with the self-paced reading condition (Bonifacci et al., 2022). 

As for MW, also for RC, significant effects of reading interest (Babbitt Bray & Barron, 2004) and motivation (Wigfeld et al., 2016) have been found. A greater interest may lead to placing more attentional resources on the text (Hidi, 2001), which in turn may improve the reader’s retention of the contents of the text and allow a deeper elaboration of it, leading the reader to have a better comprehension and a richer mental model of the text (Kintsch, 1988).

At a cognitive level, working memory skills are thought to play a significant role in both RC (De Beni, et al., 1998; Follmer, 2018; Palladino et al., 2001) and MW. Working memory serves as a buffer for integrating and establishing the coherence of different text parts, allowing to keep relevant information and discard irrelevant information from the buffer. 

This process needs to be repeated during reading to construct text meaning and coherence. Working memory is negatively associated with MW (McVay & Kane 2009; Randall et al., 2014; Unsworth and McMillan; 2013) since individual differences in working memory capacities appear to stem in part from momentary failures of conscious thought control. Interestingly, the study by McVay and Kane (2012) showed that the association between working memory and RC is partially mediated by MW rate, suggesting that control over thought content is one of the pathways through which variation in working memory capacity influences reading ability. In addition, Unsworth and McMillan (2013) evidenced that, together with working memory, interest, and motivation influence the likelihood of MW which, in turn, mediate RC skills.

Finally, the methodological approach used to measure rates of MW and the associated experimental paradigms varied considerably between studies. Typically, to examine the effect of MW on reading, a self-paced paradigm is used where participants are required to read a word, sentence, or paragraph of text at a time and are randomly probed with questions inquiring if they were on task or of task (Mooneyham & Schooler, 2013). 

Some studies have used online self-report of mindless reading (e.g., Kopp et al., 2015); that is, as soon as the participant realizes that his or her mind is wandering, he or she presses a button to signal that he or she had been engaged in mindless reading (Nguyen, et al., 2014). In other cases, post-task self-reports about the extent of MW experiences during the prior task were used (e.g., Sanders et al., 2017; Soemer et al., 2019). Overall, these studies show that MW during reading leads to slower reading times, longer average fixation duration, and the absence of the word frequency effect on gaze duration (Foulsham et al., 2013), with a negative influence on the comprehension of difficult texts (Feng, et al., 2013). 

Although real-time assessment is considered the most reliable procedure to understand the phenomenon of MW, studies have shown that reports of MW assessed via thought probes during a task are consistently and significantly correlated with post-task self-reports (e.g., Zhang et al., 2019) and dispositional (i.e., trait-based) measures of MW obtained by questionnaires (e.g., Smallwood et al., 2003; Smallwood et al., 2006), even in young adults and adolescents (Stawarczyk et al., 2014; Varao-Sousa & Kingstone, 2019). Researchers have also developed methodologies for measuring MW through the detection of eye movements (e.g., Mills et al., 2020).

In sum, results from multiple studies over the last decade generally support the idea that MW may negatively impact RC, as zoning out while reading is supposed to progressively degrade the information needed to build an efficient situational representation of the text (e.g., Smallwood, 2011). It remains an open question, however, (a) the magnitude of the association between MW and RC and (b) whether the mixed results available on the relationship between MW and RC may be associated with the influence of potential intervening factors, such as MW assessment procedures (probes, online self-report, posttest self-report, trait measures, eye gaze measures), RC measures (multiple-choice, open-ended, true–false questions), text type (expository vs. narrative), text length, and text language (e.g., transparent or opaque languages). Further, participants’ characteristics, such as age (Mrazek et al., 2013), and dispositional traits, such as topic interest or perceived difficulty of the text, might play a role in the relationship between MW and RC. Finally, cognitive traits such as working memory capacity have been accounted to play a role in both MW and RC (Unsworth, & McMillan, 2013). More generally, the role of individual differences needs to be adequately tested both in MW (Robison et al., 2020) and in RC (Schindler, & Richter, 2018).

To date, there are no meta-analyses that systematically investigated the strength of the association between MW and RC as well as explored the influence of relevant intervening factors on such a relationship. Recently, D’Mello and Mills (2021) reported on a mini-meta-analysis combing results from the reading results in Randall et al.’s (2014) study and 25 effects from studies conducted in their lab. They found a weighted mean correlation between MW and RC of −0.31. 

However, as underlined by the same authors, this revision did not include a systematic review of the literature, and therefore, in the present study, we aim to address this issue. Our first aim was to identify studies investigating the effects of MW on RC in adolescents and adults to estimate the magnitude (and direction) of the relationship between MW and RC. Our secondary aim was to identify potential moderators of the association between MW and RC, including assessment procedures of MW and RC, participant’s age, text language (transparent vs. opaque), text length and difficulty, interest in the topic, and working memory.

For more information:1950477648nn@gmail.com