Abstract
Doodling and fidgeting—traditionally viewed in educational contexts as markers of inattention and poor classroom behaviour—have more recently been considered as possible routes to improve performance by reducing boredom and its negative impact on memory. However, there is a surprising lack of well-controlled studies examining this possibility, despite the widespread adoption of fidget toys and doodling exercises within classroom settings. Here we report two experiments (total N = 222) that assess the impact of doodling on boredom, attention, mind-wandering, and subsequent recall of auditory information. In Experiment 1, participants first listened to a 15-min section of a lecture known to induce boredom. Immediately thereafter they were asked to jot down important information from a short voicemail that they listened to while either doodling (adding shading to shapes) or doing nothing in between note-taking. In Experiment 2, participants listened to a 45-min section of the same lecture under one of four conditions: structured doodling (i.e., shade in shapes), unstructured doodling, note-taking, or listen-only. Thought probes assessed self-perceived levels of state boredom, mind-wandering, and attention throughout the lecture. Across studies, doodling neither reduced boredom or mind-wandering nor increased attention or retention of information compared with other conditions. In contrast, attention and test performance were highest (and boredom and mind-wandering lowest) for those focused solely on note-taking.
Most people have felt the aversive experience of boredom, which results from various factors that affect an individual’s ability to engage in a satisfying activity (Eastwood et al., 2012). Evidence suggests that boredom is related to decreased attention (Carriere et al., 2008; Damrad-Frye & Laird, 1989; Fisher, 1998; Gerritsen et al., 2014; Hamilton et al., 1984; Hunter & Eastwood, 2018; Malkovsky et al., 2012; Seib & Vodanovich, 1998), increased instances of mind-wandering (Critcher & Gilovich, 2010; Pachai et al., 2016; Seli et al., 2015; Steinberger et al., 2016), and can hinder performance on a variety of tasks (Fritea & Fritea, 2013; Pekrun et al., 2014; Tze et al., 2016). Students are at a higher rate of suffering the negative impacts of boredom due to its overwhelming presence in the academic environment (Sinclair et al., 2018). During such periods of boredom, there may be observable increases in fidgeting behaviour—an onset and an increase in unnecessary body movements (Mehrabian & Friedman, 1986). A form of fidgeting—task-irrelevant doodling—may be a functional method of reducing the negative impact of boredom on memory (Andrade, 2010). However, there is a surprising lack of well-controlled studies examining the possibility that doodling or fidgeting is beneficial, despite the adoption of fidget toys and doodling exercises within many classroom settings. Such educators may have thereby put the cart before the horse by adopting methods that lack empirical support regarding their effectiveness (for review, see Kriescher et al., 2022). The present research tests competing hypotheses regarding the effect of doodling during episodes of boredom to get a better understanding of its impact on learners.
Boredom, mind-wandering, and attention
Boredom is a commonly experienced state that is thought to occur when we are unable to fully engage our attention with either external environmental stimuli or internal stimuli, such as our thoughts or feelings (Eastwood et al., 2012). Accordingly, when we experience difficulty becoming attentionally engaged, we typically attribute that experience and the corresponding feelings of boredom to the inadequacies of the environment. Attention is therefore a critical component of the experience of boredom (Eastwood et al., 2012). Indeed, the less attentive an individual is in an otherwise-relevant situation, the higher chance they have of perceiving that situation as boring (Gerritsen et al., 2014). Likewise, when the circumstance involves a task that is difficult to attend to, the attentional difficulty predicts levels of boredom (Carriere et al., 2008).
While attentional difficulty may lead to boredom, it is also worth considering how an activity that is perceived to be boring may, in turn, impact the ongoing focus of attention. Rather than struggling to stay focused on an external task that elicits boredom, switching attention to internal thoughts may allow your mind to save cognitive resources for information that is ultimately more useful (Pachai et al., 2016). The resulting mind-wandering may be either intentional (i.e., deliberately letting your thoughts drift from the task) or unintentional (i.e., your thoughts unintentionally drift from the task; Seli et al., 2015). Unintentional mind-wandering is thought to be a marker of boredom to the extent that when people realise their mind has involuntarily drifted, they take that as an indication that it occurred because they must have been bored (e.g., Critcher & Gilovich, 2010). Thus, any activity that acts to reduce boredom may also be expected to have a corresponding impact in reducing levels of mind-wandering.
Because the experience of boredom is quite common, it can be a persistent problem for educators trying to help students learn (Sinclair et al., 2018). This is underscored by the possibility that boredom-related failures of attention and increases in mind-wandering can impact performance (e.g., Eren & Coskun, 2016; Fritea & Fritea, 2013; Goetz et al., 2012; Pekrun et al., 2014; Tze et al., 2016). A study conducted by Pekrun et al. (2014), for example, found a reciprocal relation between boredom and exam performance for university students. Specifically, the level of course-related boredom was linked to worse exam scores, not only in that course (see also Fritea & Fritea, 2013), but in other classes as well. Moreover, Pekrun et al. (2014) observed that poor exam performance was predictive of subsequent levels of boredom in future scenarios. To evaluate the relationship between boredom and academic performance more directly, Tze et al. (2016) conducted a meta-analysis of results from 29 studies involving more than 19,000 students. Their analysis suggested that the negative impact of boredom on subsequent learning and achievement was greatest when experienced in class, rather than when studying elsewhere. Such links between in-class boredom and impaired learning emphasise the importance of educators being able to detect when a student is experiencing boredom. Having an easy-to-implement intervention that can then reduce boredom and the associated negative consequences would also be of value.
The possible functional role of fidgeting and doodling during boring episodes
In a learning environment, the onset and rise in the unnecessary body movements known as fidgets—defined by Mehrabian and Friedman (1986) as the engagement or manipulation of objects (e.g., pens, fidget toys) or your own body (e.g., tapping your foot)—is traditionally viewed as a critical marker for the onset of boredom (Seli et al., 2014). Fidgeting as an indication of one’s boredom became a noteworthy topic in the scientific literature as early as 1885 when Galton (1885, p. 174) observed fidgets in bored individuals during a lecture: When the audience is intent each person forgets his muscular weariness and skin discomfort, and he holds himself rigidly in the best position for seeing and hearing . . . But when the audience is bored the several individuals cease to forget themselves and begin to pay much attention to the discomforts attendant on sitting long in the same position.
What is interesting about Galton’s analysis is the notion that attention has a lot to do with the point at which people begin to fidget. It may be that when the individual’s task-related attention begins to decline, they become increasingly aware of their current level of boredom. As attention increasingly shifts from the task to unrelated thoughts, it seems the body increasingly begins to wander (Seli et al., 2014). This potential link between boredom, mind-wandering, and fidgeting suggests it may be possible to know when an individual is experiencing boredom and failures of attention based on their outward behaviour. Interventions aimed at reducing boredom and its negative cognitive consequences may therefore be most effective when targeted at periods showing increases in task-irrelevant movements.
Fidgeting has often been viewed in educational contexts as a marker of inattention and poor classroom behaviour. Research on naturalistic fidgets (e.g., swaying in your chair) has assessed this view by evaluating whether fidgeting is indeed an indication of inattention. Farley et al. (2013) tested this perspective using both self-report measures of attentiveness during lecture viewing and observations of participants’ fidgeting. Their results suggest that fidgeting impacted the ability to retain lecture material over and above the effect of inattention, thus implying that fidgeting was not merely a marker of inattention. They suggested that fidgeting may have helped to offset the negative distress of having to sustain attention for a prolonged period. However, while boredom could have been the form of distress thought to be offset by fidgeting here, it was not specifically measured or manipulated in this study, nor was the potential boredom-reducing benefit of fidgeting for helping students subsequently re-engage task-focused attention.
In addition to its characterisation in terms of attention, boredom has also been defined as being a “state of nonoptimal arousal” that occurs when there is a difference between the physiological activity required for an individual’s peak performance and the stimulation available within their environment (for review, see Eastwood et al., 2012). A state of boredom can thereby arise during either insufficient or excessive levels of arousal. Accordingly, boredom with low arousal may manifest as apathy and tiredness, whereas boredom with high arousal may manifest more as frustration and agitation. The optimal stimulation theory developed by Hebb (1955) was used by Kercood and Banda (2012) to explain how organisms achieve “stimulatory” homeostasis through stimulation-seeking activity (i.e., fidgeting, physical activity). Individuals are motivated to achieve an optimal level of arousal and will seek out ways to achieve that level. Fidgeting may thereby be functional in helping an individual regulate arousal levels to a point that allows them to re-engage their attention (Andrade, 2010; Hartanto et al., 2016; Kercood & Banda, 2012; Mead & Scibora, 2016; Stalvey & Brasell, 2006). Research on doodling published by Andrade (2010) provided preliminary support for such a theory.
Doodling is a commonly used form of task-irrelevant behaviour, which involves creating marks or drawings of things that are entirely unrelated to the information to which an individual is supposed to be attending (Meade et al., 2019). Interest in the potential usefulness of doodling as a learning technique has increased following Andrade’s (2010) demonstration of its ability to reduce boredom and increase recall of information. Participants were asked to listen to a 2.5-min voicemail that mentioned the names of multiple people and multiple cities. The doodling group was asked to shade in a set of shapes on a sheet of paper while listening to the voicemail, whereas the control group was asked to only listen. Both groups were asked to write down each of the people’s names mentioned in the voicemail as soon as they heard them. A surprise test was given afterward in which participants were asked to recall, not only all of the names they had written down but also all of the city names they had heard. Participants who doodled while listening to the voicemail recalled significantly more of the voicemail information than those in the control group. Because participants had just completed a much longer study before taking part in this experiment, Andrade speculated that the doodling-related improvement might have been due to the doodling increasing levels of arousal and thereby reducing boredom and mind-wandering.
Unfortunately, however, boredom was never intentionally induced or measured in Andrade’s (2010) study. Any differences in the level of attention or mind-wandering for participants in the different groups were also not formally considered. Moreover, the results of recent research (Boggs et al., 2017) that focused on the effects of doodling on mind-wandering are inconsistent with those reported by Andrade (2010). Specifically, Boggs et al. (2017) did not see any benefits of having participants do structured doodling (shading in shapes) when compared with performance when participants only listened in their task, while unstructured doodling (creating free-form images) hindered performance relative to the other conditions (see also Meade et al., 2019). These conflicting findings, when considered together with information sources that continue to promote the effectiveness of fidgeting and doodling for combatting boredom and enhancing attention (e.g., Rotz & Wright, 2021), underscore the importance of confirming whether doodling is indeed beneficial.
The current study
The present research tests competing hypotheses about the extent to which different methods of doodling may be helpful for cognitive functioning during an experience of boredom. The “fidgeting reduces boredom and increases attention” hypothesis posits that doodling is a beneficial form of fidgeting that can reduce boredom and increase attention to promote better learning (Andrade, 2010). In contrast, the “fidgeting reflects inattention” hypothesis maintains that doodling is merely an indication of the mind taking a mental break, thereby reflecting the absence of task-focused attention (mind-wandering; Boggs et al., 2017) and would therefore be linked to relatively poor learning. Distinguishing between these two possibilities is important for understanding how the mechanisms subserving attentional, behavioural, and affective engagement operate together and for resolving the conflict in existing evidence about the extent to which doodling is a viable method for educators and students to implement in the context of learning.
Experiment 1 is a replication and extension of the study conducted by Andrade (2010) to examine whether doodling can reduce feelings of boredom and levels of mind-wandering while improving task-focused attention and memory for associated information. Our main extension to this study was the addition of a manipulation to induce boredom. In Experiment 2, we build on the results of Experiment 1 by using a more ecologically valid approach in which participants listen to an actual lecture, and then complete a multiple-choice exam that tests retention of lecture material. Like Boggs et al. (2017), we included a listen-only control condition, as well as a note-taking condition, a structured-doodle condition, and an unstructured-doodle condition. To anticipate, our results—consistent with the fidgeting-reflects-inattention hypothesis—indicated that doodling neither reduced boredom or mind-wandering in either experiment, nor increased attention or retention of information when compared with the other conditions. None of our results were consistent with the fidgeting-reduces-boredom-and-increases-attention hypothesis.
Experiment 1
Experiment 1 examined the effect of adding a secondary task (structured doodling) compared with no additional task on subsequent recall ability. This experiment thereby represented an attempt to conceptually replicate and extend Andrade’s (2010) doodling study using similar methods and procedures. Participants reported their feelings of boredom both before and after listening to a boredom-inducing lecture. Participants listened to a voicemail recording immediately thereafter while either doodling and note-taking, or just note-taking, and were then asked to recall information from the voicemail. Our experiment was not an exact replication of the methods of Andrade (2010) due to our addition of the boredom-induction procedure and measure of state boredom. The fidgeting-reduces-boredom-and-increases-attention hypothesis, and the results reported by Andrade (2010), predicts that participants in the doodle condition should perform better on the subsequent memory test than those in the control condition. In contrast, the fidgeting-reflects-inattention hypothesis predicts that the addition of doodling would reduce attention to the voicemail and impair performance on the subsequent memory test when compared with that of the control condition.
Method
Participants
A total of 50 University of Guelph undergraduate students (age: M = 18.9, SD = 1.80; self-reported gender: 44 female, 6 male) were recruited from the Department of Psychology participant pool. They gave informed consent in writing and received course credit for participation. All materials, methods, and procedures used in the experiment were approved by the University of Guelph Research Ethics Board (REB protocol #16-12-398).
Materials
Recorded lecture
Participants listened to a 15-min recorded lecture titled “The Dark Ages” from an introductory Ancient Greek History course using a pair of over-ear noise-cancelling headphones. The audio was obtained from OpenYale Courses (https://oyc.yale.edu) and was used to induce boredom. Our previous use of this video in unrelated experiments has revealed that it typically elicits increases in feelings of boredom for our Psychology students.
Voicemail
Participants listened to a 2.5-min voicemail through the same headphones. The voicemail script was adopted from Andrade’s (2010) study and was slightly modified. The script was kept the same except for the names of places were altered to reflect the local geographical area. The speaker was a female who talked in a slow, monotone voice.
Multidimensional State Boredom Scale-8
Participants self-reported feelings of boredom before and after watching the lecture video by completing the Multidimensional State Boredom Scale-8 (MSBS-8; Hunter et al., 2015). The MSBS-8 includes items such as “I feel bored,” and “I wish I was doing something more exciting” that are presented with a sliding scale that ranges from 1 (strongly disagree) to 7 (strongly agree). Scores are summed across items to give a total boredom score, with higher scores indicating greater levels of subjective boredom (max score: 56).
Procedure
Participants were tested in-person in individual testing rooms in a laboratory setting. Before beginning the experiment, participants first provided informed consent and answered basic demographic questions. The first phase involved the boredom-induction procedure, consisting of the participants providing their pre-lecture state boredom rating (MSBS-8), listening to the first 15 min of the lecture audio, and then providing their post-lecture boredom ratings (MSBS-8). Immediately thereafter, participants were asked to listen to a short voicemail and pretend that the speaker is a friend who has called to invite them to a party. Participants were randomly assigned to either the doodle condition or the control. For the doodle condition, participants were provided with a regular sheet of white paper containing alternating rows of squares and circles. The right column of the page had a 4.5-cm-wide column of blank space. Participants in the doodle condition were asked to shade in the shapes while listening to the voicemail. They were asked to write on the right-hand side of the paper the names of the people mentioned in the voicemail as going to the party. In reference to the shading of the shapes, participants were told “It does not matter how neatly or how quickly you do this—it is just something to help relieve the boredom” (Andrade, 2010). The control condition was given a regular blank sheet of white paper and told to write the names of people going to the party, and not to write anything else. Participants in both conditions were given a regular pencil to write down names and were told they would not be tested on their ability to recall this information. Once completed, participants’ response sheets were collected, and they were asked to wait in the room until the experimenter returned (they had to wait for 1 min). Each participant was then asked whether they expected a memory test and to orally recall the names of people going to the party (attended information) as well as the names of any places mentioned in the voicemail (unattended information). The experimenter recorded each of the recalled names and places during the memory test. The ordering of the questions was counterbalanced across participants.
Results
An independent samples t-test was used to compare the summed MSBS-8 scores from before the lecture for the control group (M = 32.68, SD = 7.45) and the doodling group (M = 31.72, SD = 6.62). This confirmed that there was no significant difference between the groups in levels of state boredom at the beginning of the experiment, t(48) = .48, p = .63, d = .14. To ensure participants were bored before the voicemail listening task, summed MSBS-8 scores from before and after the lecture were analysed in a paired samples t-test. Results suggest that participants were significantly more bored after listening to the lecture video (M = 38.46, SD = 8.10), than they were prior to the video (M = 32.2, SD = 7); t(49) = −5.55, p < .001, d = −.79).
There were eight names of people going to the party for the participants to recall (while ignoring the names of people who were not going to the party) and eight names of places to recall. Any other names recalled were scored as a false alarm, and 1 point for every false alarm was deducted from their recall score for that category. This is the same approach used by Andrade (2010) to ensure that participants were not just recalling names of people who were mentioned in the voicemail but not as one of the people going to the party, or were not just generating random names. The average numbers of correct-recalls and false alarms for names and places, as well as the resulting memory scores are shown in Table 1 for the doodle and control groups. As can be seen in the table, and as detailed below, memory performance was no better for the participants who doodled than for those who did not.
Means and standard deviations for correct recall, false alarms, and overall memory scores for names and places in the control and doodle groups (n = 25 per group).
Submission of the correct-recall rates to a 2 (Group: doodling, control) × 2 (Stimulus-type: names, places) mixed-factors ANOVA, for example, revealed that there was no significant difference in the overall correct-recall performance of the doodlers compared with that of the control group, Group main-effect: F(1, 48) = 1.47, p = .23, ηp2 = .03. Although significantly more party-goer names were recalled than places, overall, Stimulus-type main effect: F(1, 48) = 59.07, p < .001, ηp2 = .55, the magnitude of this effect did not significantly depend on whether participants were in the doodle group or the control group, Group-by-Stimulus-type interaction: F(1, 48) = .34, p = .56, ηp2 = .007.
Submission of false-alarm numbers to a separate 2 (Group: doodling, control) × 2 (Stimulus-type: names, places) mixed-factors ANOVA, likewise revealed that there was no significant difference in the average numbers of false reports by the doodlers compared with that of the control group, Group main-effect: F(1, 48) = .25, p = .62, ηp2 = .005. Although significantly more party-goer names were incorrectly reported than places, overall,, Stimulus-type main effect: F(1, 48) = 12.23, p = .001, ηp2 = .20, the magnitude of this effect did not significantly depend on whether participants were in the doodle group or the control group, Group-by-Stimulus-type interaction: F(1, 48) = .40, p = .53, ηp2 = .008.
After subtracting false alarms from correct-recalls for reported party-goer names and places, the resulting total memory scores were submitted to another 2 (Group: doodling, control) × 2 (Stimulus-type: names, places) mixed-factors ANOVA. This revealed that the average total-memory performance for the doodle group was not significantly different from that of the control group, Group main-effect: F(1, 48) = .55, p = .46, ηp2 = .01. Although total-memory scores were again significantly higher for party-goer names than for places, overall, the magnitude of this effect did not significantly depend on whether participants were in the doodle group or the control group, Group-by-Stimulus-type interaction: F(1, 48) = .52, p = .48, ηp2 = .01.
Discussion
The fidgeting-reduces-boredom-and-increases-attention hypothesis and the results reported by Andrade (2010) predicted that participants in the doodle condition of Experiment 1 should perform better on the subsequent memory test than those in the control condition. They did not. Indeed, performance on the subsequent memory test was nominally lower for those in the doodle condition than for those in the control condition. While this memory impairment from doodling relative to that of the control condition was not statistically significant, the nominally lower performance is more in line with the fidgeting-reflects-inattention hypothesis, which predicted that the addition of doodling would—if anything—reduce attention to the voicemail when compared with that of the control condition. Our failure to observe a significant difference in memory performance between the doodle and control conditions in Experiment 1 represents a null result. However, our ability to draw conclusions from this is limited by the fact that a non-significant p-value in traditional null hypothesis significance testing does not itself provide evidence of no difference between groups, but instead simply indicates that any potential difference was not detectable. In response to a reviewer’s suggestion, we therefore used a Bayesian approach to more directly assess the evidence favouring the null hypothesis that doodling does not influence subsequent memory performance. A Bayes factor (BF10) was calculated using the default priors in JASP (JASP Team, 2023, version 0.17.3; see Morey & Rouder, 2011, for a review of BFs). BF10 values of less than 1 indicate different degrees of support for the null hypothesis depending on where they fall relative to key benchmarks, with values in the 0.33–1 range considered anecdotal (insufficient) evidence, those in the 0.1–0.33 range considered moderate evidence, and values less than 0.1 considered strong evidence. Regarding the main effect of Group (doodle vs. control) on the total memory score, we obtained a BF10 = 1.14 × 10−6, indicating strong support for the null hypothesis.
Previous studies have suggested that doodling may be a helpful intervention strategy in a boring situation by distracting the individual away from their boredom and thereby increasing the ability to recall information (Andrade, 2010). However, our results suggest that this may not be accurate, as there was no benefit of doodling (vs. not doodling) for subsequent memory performance under conditions in which participants were shown to be experiencing elevated levels of boredom. Our failure to replicate Andrade’s (2010) finding that doodling aids recall ability for voicemail material may be due to the possibility that their participants were not experiencing boredom. To the extent that elevated boredom is associated with difficulties with task-focused attention, the potential for low levels of boredom in Andrade’s (2010) study could mean that their participants had available attentional resources to be better able to engage with the material while doodling than our participants, who may have had boredom-related difficulties with attentional engagement. If this is true, it may indicate that doodling is not effective at offsetting boredom or its negative consequences on learning.
However, while we took steps to ensure there were elevated levels of boredom for participants going into the voicemail task in Experiment 1, this study alone cannot explain exactly how boredom may have played a role in determining the observed results. In a typical learning environment, boredom tends to occur during the actual task not prior. In addition, most educational settings require task-focused attention for much longer than 2.5 min. Thus, the length of the task we adapted from Andrade (2010) is far too short to be reflective of a real-life scenario and may therefore not have been long enough to provide an opportunity for doodling to be fully effective in offsetting boredom-related disengagement with the material. It is also impossible to know whether doodling could have been functional for offsetting boredom or any of its negative correlates in Experiment 1—beyond subsequent memory performance—because there was no measure of task-focused attention or off-task mind-wandering. We address this in Experiment 2 by manipulating doodling behaviour during a longer lecture-listening task in which we measured in-task levels of attention, mind-wandering, and boredom, as well as subsequent retention of the lecture material.
Experiment 2
Experiment 2 builds on Experiment 1 by assessing the impact of doodling on boredom, mind-wandering, attention, and retention in a more ecologically valid lecture-listening task. The importance of including in-task measures of mind-wandering and attention is highlighted by the findings of previous research showing both links between mind-wandering and fidgeting behaviour (Carriere et al., 2013), and how our body also becomes restless as our mind becomes restless (Seli et al., 2014). However, while these prior studies make clear a connection between fidgeting behaviours and difficulties maintaining task-focused attention, they did not include measures of boredom or associated memory performance, nor did they manipulate fidgeting behaviours to directly assess their impact on mind-wandering or attention. Experiment 2 therefore also extends this prior work by manipulating different types of fidgeting behaviours to directly assess their impact on boredom, mind-wandering, and attention during a lecture-listening task, as well as the associated effects on retention of lecture material.
Our decision to include a manipulation of different types of fidgeting behaviours—the same “shade in shapes” structured form of doodling used in Experiment 1, an “anything goes” unstructured form of doodling, a note-taking condition, and a “listen-only” control condition—was inspired by Boggs et al. (2017) who used these same conditions. Unstructured doodling more closely represents the type of doodling used in real-world settings, while note-taking represents another activity frequently used in real-world settings. Participants in Boggs et al.’s study were randomly assigned to one of these four conditions while they listened to a 5-min fictional conversation between two friends, after which they completed a quiz that tested their ability to recall information from that conversation. Boggs et al. found that unstructured doodling while listening led to significantly worse recollection of the conversation content than the structured doodling or note-taking. They interpreted this learning impairment as being due to the additional attentional demands during unstructured doodling of having to decide what to doodle, which may have thereby reduced participants’ capacity to encode details of the conversation. However, the authors urged caution regarding this interpretation because they did not include a measure of attention in their study. Experiment 2 therefore also extends this prior work by using the same doodling conditions as Boggs et al. along with in-task measures that allow a direct assessment of their impact on attention, mind-wandering, and boredom during a lecture-listening task, as well as the associated effects on retention of lecture material.
The fidgeting-reduces-boredom-and-increases-attention hypothesis predicts that participants in both the structured-doodling and unstructured-doodling conditions should show lower levels of in-task boredom and mind-wandering, and higher levels of in-task attention and subsequent memory for lecture content, than those in the control condition. In contrast, the fidgeting-reflects-inattention hypothesis predicts that the addition of either doodling condition would reduce attention to the lecture, increase in-task boredom and mind-wandering, and—if anything—impair performance on the subsequent memory test when compared with that of the control condition. Furthermore, consider Boggs et al.’s (2017) speculation that unstructured doodling places additional demands on attention, due to the extra thought and effort required to generate doodle content, relative to structured doodling. If this is correct, then the unstructured-doodling condition in our experiment should also show reduced in-task attention, more mind-wandering, and worse memory for lecture content relative to the structured doodling condition. In terms of in-task boredom, however, it may be possible that unstructured doodling is relatively interesting and engaging when compared with the experience of structured doodling or passive listening. If so, it might produce lower levels of in-task boredom and mind-wandering than these other conditions, regardless of whether it results in less task-focused attention or relatively poor memory for lecture content.
As for the note-taking condition, writing down information as you hear it may aid the encoding of that information. For example, summarising material—restating information in your own written words—has been shown to be an effective learning strategy when used to aid the recall of text-based information and lecture material (e.g., Bretzing & Kulhavy, 1979; King, 1992). This type of note-taking is believed to aid the encoding process by requiring the individual to use generative processing. More specifically, it requires the individual to identify the most important points, organising the information in the process until they can restate the content more concisely in their own words. This process is thought to help the individual make stronger connections between existing knowledge and the newly encountered information (see Fiorella & Mayer, 2014 for review). This could explain both Boggs et al.’s (2017) finding that note-taking leads to better memory performance than passive listening or unstructured doodling, and Meade et al. (2019) finding that writing down words as they were heard also resulted in better memory for those words than for words that were heard during unstructured doodling. However, the extent to which any difference in memory performance for note-taking compared with the other conditions is associated with differences in boredom, mind-wandering, or attention remains unclear as these prior studies did not include measures of these factors.
Participants in Experiment 2 completed several questionnaires to allow us to examine whether there is a relation between an individual’s tendency to doodle (DSAQ: Doodle Spontaneous Activity Questionnaire; developed for this study based on Carriere et al., 2013), their tendency to fidget (SAQ: Spontaneous Activity Questionnaire; Carriere et al., 2013), and the extent to which they routinely experience boredom (BPS: Boredom Proneness Scale; Farmer & Sundberg, 1986) or attention-related difficulties in terms of mind-wandering (MWQ: Mind-Wandering Questionnaire; Mrazek et al., 2013), attention-related cognitive errors (ARCES; Carriere et al., 2008), or lapses in mindful attention (MAAS-LO: Mindful Attention Awareness Scale-Lapses Only; Carriere et al., 2008, cf. Brown & Ryan, 2003). Of the potential relations between individual-difference measures, we were particularly interested in the extent to which those who self-report being high in fidgeting behaviour (SAQ), also report being high in doodling behaviour (DSAQ). If doodling is just a specific form of the more general category of fidgeting, then DSAQ scores and SAQ scores should be positively correlated.
Obtaining a measure of trait mind-wandering (MWQ) for each participant, along with a measure of self-reported doodling frequency (DSAQ) was intended to help test our competing hypotheses and thereby address the discrepancies between prior work, implying that doodlers mind-wander less (Andrade, 2010) and subsequent results suggesting doodlers mind-wander more (Boggs et al., 2017). Specifically, the fidgeting-reduces-boredom-and-increases-attention hypothesis predicts negative correlations between the self-reported tendency to doodle (DSAQ) and measures of boredom proneness (BPS), the tendency to experience attentional failures (ARCES), and lapses in attention (MAAS-LO). In contrast, the fidgeting-reflects-inattention hypothesis predicts positive correlations between DSAQ scores and both ARCES scores and MAAS-LO scores. The questionnaires we included to look at individual differences in self-reported doodling, fidgeting, attentional lapses/cognitive errors, boredom, and mind-wandering also allowed us to assess whether there are any particular subsets of individuals for whom fidgeting/doodling may be especially effective, as has previously been suggested (e.g., those who routinely experience attention-related difficulties; Kercood & Banda, 2012; Rotz & Wright, 2005; Zentall, 1975). Including these measures also allows an assessment of the speculation by Boggs et al. (2017) that doodling might have differing cognitive impacts based on whether a given individual already has a tendency to doodle.
Method
Participants
Participants were 172 (43 per group, age: M = 19.08, SD = 2.46; self-reported gender: female = 145, male = 27) University of Guelph undergraduate students. They were recruited using the Department of Psychology participant pool, provided written informed consent, and received course credit for their participation. All materials and procedures for this study were approved by the University of Guelph Research Ethics Board (REB protocol #16-12-398).
Materials: mass testing
SAQ
The SAQ was completed by participants during mass testing to get a measure of an individual overall tendency to fidget (Carriere et al., 2013). The questionnaire has eight questions with excellent internal consistency (α = .94). An example question from the SAQ is “I fidget while I am in deep thought.” Answers range from 1 (never) to 7 (always), with higher summed scores (max score is 56) indicating an individual has a greater tendency to fidget.
BPS
The BPS was completed during mass testing to measure an individual’s propensity to experience boredom (Farmer & Sundberg, 1986). The BPS measures how often the individual experiences boredom in his or her daily life rather than at a specific time point. A sample item from the BPS includes “It takes me more stimulation to get me going than most people.” The BPS has 28 questions with acceptable reliability (α = .79), that are answered as True (1 point), False (0 points), or not answer at all. (Note: items 1, 7, 8, 11, 13, 15, 18, 22, 23, and 24 are reversed scored.) The higher the sum of the responses (max score is 28) is indicative that the individual has a higher trait tendency to experience boredom.
MWQ
The MWQ is completed during mass testing to measure an individual’s overall tendency to mind-wander (Mrazek et al., 2013). The questionnaire consists of five items with good internal consistency (α = .85). The MWQ includes items such as “I mind-wander during lectures or presentations” with responses ranging from 1 (almost never) to 6 (almost always). Higher scores (max score of 30) indicate a higher propensity to mind-wander. Previous research has shown that there is a relation between the tendency to experience boredom measured by the BPS and instances of mind-wandering, suggesting that higher boredom proneness is related to more off-task thought (Isacescu et al., 2017).
Materials: in-session
Recorded lecture
Participants listened to the first 45 min of the recorded lecture titled “The Dark Ages” from an introductory Ancient Greek History course (same as Experiment 1).
Thought probes
To obtain in-task measures of state boredom, mind-wandering, and task-focused attention, we presented thought probes at four different points throughout the recorded lecture (9, 18, 27, and 36 min). For each probe, the lecture audio would stop playing and participants would be asked to use a Likert-type scale to rate their level of boredom (“How bored were you prior to the probe?”: 1 = not at all bored to 7 = extremely bored), mind-wandering (“Where was your attention focused just before the probe?”: 1 = not at all on task to 7 = completely on task), and attention (How much attention were you paying prior to the probe?”: 1 = not paying attention at all to 7 = full attention).
MSBS-8
Participants indicated the level of boredom they were experiencing before and after listening to the recorded lecture by completing the MSBS-8 (Hunter et al., 2015; same as Experiment 1).
MAAS-LO
To assess individual differences in participants’ propensity to experience attentional lapses, participants were asked to complete the 12-question MAAS-LO (Carriere et al., 2008). The MAAS-LO has good reliability (α = .83) (e.g., Carriere et al., 2008, 2013; Cheyne et al., 2006). Responses are made using a Likert-type scale ranging from 1 (almost never) to 6 (always) on questions such as “I rush through activities without really being attentive to them.” Responses are summed with higher scores indicating greater tendencies to experience attentional lapses (max score of 72). Previous research has shown this measure is positively correlated with the tendency to fidget (measured via SAQ; Carriere et al., 2013) and experience boredom (measured via BPS; Carriere et al., 2008).
ARCES
To assess individual differences in participants’ propensity to experience cognitive failures due to attention lapses, participants were asked to complete the 12-item ARCES (Cheyne et al., 2006). An example question includes, “I have gone to the fridge to get one thing (e.g., milk) and taken something else (e.g., juice).” Responses are made using a Likert-type scale ranging from 1 (never) to 5 (very often). The sum of the responses provides a score that indexes the cognitive consequences of attentional lapses, with higher scores indicating greater tendencies to experience ARCES (max score of 60). ARCES scores have been shown to be positively correlated with MAAS-LO scores (Carriere et al., 2008, 2013; Cheyne et al., 2006), SAQ scores (Carriere et al., 2013), and BPS scores (Carriere et al., 2008).
DSAQ
To assess individual differences in participants’ propensity to doodle, we modified the SAQ (Carriere et al., 2013) to focus specifically on doodling behaviour. We simply changed each instance of “I fidget” in the questions to “I doodle.” An example of a modified question includes “I doodle when I am worried about something.” Responses are made using a Likert-type scale ranging from 1 (never) to 7 (always). The sum of the responses provides a score that indexes the propensity to engage in doodling, with higher scores indicating greater tendencies to doodle (max score is 49). If doodling is just a specific form of the more general category of fidgeting, then DSAQ scores and SAQ scores should be positively correlated.
Retention exam
Participants were asked to complete a set of 28 multiple-choice questions to measure their retention of lecture material. The questions asked had to do with specific lecture content (i.e., “Where do we see civilisation for the first time in the Aegean Sea area?”).
Procedure
Participants were required to complete an online mass-testing series of questionnaires to be eligible for participation. Once in the lab, the participant was asked to complete a demographic survey and the MSBS-8. Participants were told they would be listening to the audio of a university lecture through headphones and that the headphones were not to be removed. They were also told that questions would occasionally appear on the screen during the lecture audio that would require them to read the question and make an honest response using the numbers on the keyboard. Participants were informed that there would be a memory test at the end of the lecture, so they should pay close attention to the material. At this point, they were given instructions based on their randomly assigned condition and then the audio was started.
The control condition was not given any secondary task to do while listening. The note-takers were instructed to take notes while listening to the lecture anytime they were not answering questions on the screen (thought probes). Participants were given five sheets of blank white paper with two pencils and a pen to use. The structured doodlers were instructed to shade in shapes while listening to the lecture anytime they were not answering questions on the screen (thought probes). They were not to do anything else or write notes. Participants were given five sheets of paper with alternating shapes on them, with two pencils and a pen to use. Finally, the unstructured doodlers were instructed to doodle while listening to the lecture anytime they were not answering questions on the screen (thought probes). Participants were given five sheets of blank white paper with two pencils and a pen and told they could doodle anything, except lecture material or write notes, to use. Upon completion, the participants completed the MSBS-8, DSAQ, MAAS-LO, ARCES, and the retention questions. Participants were then debriefed on the purpose of the study and thanked for their participation. All of the questionnaires were administered using Qualtrics online survey software.
Results
Pre-experiment levels of boredom
To confirm that there were no pre-existing differences in the level of boredom being experienced by our different experimental groups, we submitted the summed pre-task MSBS-8 scores to a one-way ANOVA with Group (Listen-only, Note-taking, Structured-doodle, and Unstructured-doodle) as the between-subjects factor. The analysis confirmed that there were no significant group differences in levels of state boredom prior to the experiment, F(3, 168) = 0.27, p = .85, η2 = .005.
In-task measures of boredom, mind-wandering, and attention
To assess the effects of the different types of fidgeting behaviours on in-task levels of boredom, mind-wandering, and attention throughout the lecture-listening task, we conducted a separate 4 (Group: Listen-only, Note-taking, Structured-doodle, and Unstructured-doodle) × 4 (Time: 9, 18, 27, and 36 min) mixed-factors ANOVA for each of these measures. As detailed below, there was no benefit on any of these measures for either type of doodling, relative to the listen-only control. There was, however, a distinct advantage for the note-takers relative to all other groups that was specific to levels of mind-wandering and task-focused attention.
Boredom
As shown in Figure 1, self-reported state boredom increased over time for all groups. This main effect of Time was significant, F(3, 504) = 22.07, p < .001, ηp2 = .12. Figure 1 shows that the overall levels of boredom over time were similar for the different groups, which was reflected in the lack of a significant main effect of Group, F(3, 168) = .94, p = .42, ηp2 = .02, and the lack of a significant Time-by-Group interaction, F(9, 504) = 1.24, p = .27, ηp2 = .02. As a further test of the possibility that different fidgeting behaviours might differentially influence the experience of boredom, we submitted the post-task MSBS-8 scores to a one-way ANOVA with Group (Listen-only, Note-taking, Structured-doodle, and Unstructured-doodle) as the between-subjects factor. This confirmed that there were also no significant group differences in levels of state boredom after the listening task, F(3, 168) = 1.03, p = .38, η2 = .018.
Average mid-task ratings of boredom (1 = not at all bored to 7 = extremely bored), mind-wandering (1 = not at all on task to 7 = completely on task), attention (1 = not paying attention at all to 7 = full attention) at different timepoints of the lecture-listening task (9, 18, 27, and 36 min) for participants in the Listen-only control condition, the Structured-doodle condition, the Unstructured-doodle condition, and the Note-taking condition (n = 43 per condition). (Error bars = 95% confidence intervals.)
Mind-wandering
As shown in Figure 1, self-reported levels of mind-wandering increased over time for all groups. This main effect of Time was significant, F(3, 504) = 36.94, p < .001, ηp2 = .18. However, Figure 1 also shows that the overall levels of mind-wandering over time were noticeably lower for the Note-taking group than for the other groups, which was reflected in a significant main effect of Group, F(3, 97) = 11.49, p < .001, ηp2 = .17. Post hoc tests using the Bonferroni correction revealed that the Note-taker group mind-wandered significantly less than the Listen-only control group (t = −4.96, p < .001, d = −.83), Structured-doodle group (t = −4.60, p < .001, d = −.77), and Unstructured-doodle group (t = −4.79, p < .001, d = −.81), none of which significantly differed from each other (for each comparison, t < .37, p > .99, d < .06). Moreover, the extent to which note-taking reduced levels of mind-wandering relative to the other conditions did not change over time, as reflected by the lack of a significant Time-by-Group interaction, F(9, 504) = .55, p = .84, ηp2 = .02.
Attention
Whereas boredom and mind-wandering increased over time, Figure 1 shows that the self-perceived amount of attention paid to the lecture-listening task decreased over time. This main effect of Time was significant, F(3, 504) = 37.81, p < .001, ηp2 = .18. Moreover, as with levels of mind-wandering, Figure 1 also shows that the levels of attention paid over time were noticeably different for the Note-taking group than for the other groups, which was reflected in a significant main effect of Group, F(3, 168) = 6.09 p = .001, ηp2 = .10. Post hoc tests using the Bonferroni correction revealed that the Note-taker group reported paying significantly more attention than the Listen-only control group (t = 3.20, p < .001, d = .55), Structured-doodle group (t = 2.95, p = .022, d = .51), and Unstructured-doodle group (t = 3.98, p < .001, d = .68), none of which significantly differed from each other (for each comparison, t < 1.04, p > .99, d < .18). Moreover, the extent to which note-taking increased levels of attention relative to the other conditions did not change over time, as reflected by the lack of a significant Time-by-Group interaction, F(9, 504) = 1.24, p = .27, ηp2 = .02.
Retention
The notion that fidget behaviours might reduce boredom and mind-wandering, and thereby boost task-focused attention and aid learning (or, conversely that fidget behaviours are an index of a lack of engagement and may thereby be linked to less task-focused attention and impaired learning), is based on the possibility that higher levels of boredom and mind-wandering and lower levels of attention all lead to impairments in the ability to encode and retain information encountered during a learning experience. To test this, we examined the extent to which levels of state boredom, mind-wandering, and attention were correlated, overall, with subsequent performance on the multiple-choice test of memory. This showed that the number of correct memory-test answers was indeed negatively correlated with overall levels of in-task boredom (average of the four ratings), r = −.21, p = .006, post-task boredom (MSBS-8 score obtained after the lecture), r = −.28, p < .001, and overall levels of mind-wandering (average of the four in-task ratings, r = −.36, p < .001), while memory-test performance was positively correlated with levels of self-perceived attention paid to the lecture-listening task (average of the four in-task ratings, r = .34, p < .001). This is consistent with prior evidence that boredom and its corresponding difficulties with attentional engagement have clear negative consequences for academic outcomes (Fritea & Fritea, 2013; Pekrun et al., 2014; Tze et al., 2016). It also underscores the potential value for learning contexts of any fidgeting-based intervention that may be effective in reducing boredom and mind-wandering, and increasing task-focused attention.
As shown in Table 2, while there was little difference among the average numbers of correct multiple-choice answers for the Structured-doodle, Unstructured-doodle, and Listen-only control groups, the memory performance of the Note-taking group was notably higher. A one-way ANOVA with Group (Listen-only, Structured-doodle, Unstructured-doodle, and Note-taking) as the between-subjects factor confirmed that this main effect of Group was significant, F(3, 168) = 9.55, p < .001, ηp2 = .15. Post hoc tests using the Bonferroni correction revealed that the Note-taker group remembered significantly more of the lecture content than the Listen-only control group (t = 3.47, p < .001, d = .75), the Structured-doodle group (t = 4.42, p < .001, d = .95), and the Unstructured-doodle group (t = 4.79, p < .001, d = 1.03). There were no significant differences among the Structured-doodle, Unstructured-doodle, and Listen-only control groups (for each comparison, t < 1.32, p > .56, d < .29). In other words, neither type of doodling led to any better retention of the lecture content than passively listening.
Average retention score (number correct answers on the 28-question multiple-choice test of memory for lecture content) for the Listen-only, Structured-doodle, Unstructured-doodle, and Note-taking groups (n = 43 per group).
Individual differences
We analysed our individual-difference measures to assess whether there are any particular subsets of individuals for whom fidgeting/doodling may be especially effective and to test our competing hypotheses regarding the cognitive-affective correlates of self-reported tendencies to doodle and fidget (see Table 3 for descriptive statistics).
Descriptive statistics (SD = Standard deviation) for all individual-difference measures, including trait mind-wandering (MWQ), tendency to experience attentional lapses (MAAS-LO) and attention-related cognitive errors (ARCES), boredom proneness (BPS), doodling behaviour (DSAQ), and fidgeting behaviour (SAQ).
MWQ: Mind-wandering Questionnaire; MAAS-LO: Mindful Attention Awareness Scale-Lapses Only; ARCES: Attention-Related Cognitive Errors Scale; BPS: Boredom Proneness Scale; DSAQ: Doodle Spontaneous Activity Questionnaire; SAQ: Spontaneous Activity Questionnaire.
Participant subsets
Perhaps our failure to observe an overall benefit of doodling for mid-task levels of boredom, mind-wandering, and task-focused attention, or for subsequent memory performance is due to the possibility that doodling and other forms of fidgeting only have cognitive–affective benefits for certain types of people. Fidgeting and doodling have been purported, for example, to be particularly helpful for individuals with attention-related difficulties (e.g., Kercood & Banda, 2012; Rotz & Wright, 2005). The possibility that fidgeting/doodling may help to reduce boredom suggests that such behaviours might also be of particular value to individuals who are prone to experience boredom, or to those who, through experience, have come to routinely engage in fidgeting or doodling. We therefore identified the subsets of participants who had scores higher than the group median on the individual-difference measures of mind-wandering (MWQ), attentional lapses (MAAS-LO), ARCES, BPS, doodling behaviours (DSAQ), and fidgeting behaviours (SAQ). For each of these participant-subsets, we then conducted separate 3 (Group: Listen-only, Structured-doodle, and Unstructured-doodle) × 4 (Time: 9, 18, 27, and 36 min) mixed-factors ANOVAs to assess the effect of the different types of doodling behaviours on in-task levels of boredom, mind-wandering, and attention. Note that, to focus solely on the potential benefits of doodling, per se, we omitted all note-taking participants from these additional analyses. The effect of doodling condition on subsequent retention of lecture content was also assessed for each subset of participants using a one-way ANOVA with Group (Listen-only, Structured-doodle, and Unstructured-doodle) as the between-subjects factor. The results for the main effect of Group from these ANOVAs are shown in Table 4.
Average mid-task ratings of Boredom, Attention, Mind-wandering, and subsequent Retention scores (# correct out of 28) for the subsets of participants with scores that were higher than median of all participants in their tendencies to experience Mind-wandering, Attentional lapses, Attention-related cognitive errors, Boredom, Doodling, and Fidgeting. F-values and ηp2-values are reported for the main effect of Group (Listen-only control, Structured-doodle, Unstructured-doodle) for each measure obtained from each participant subset.
CI: confidence interval; MWQ: Mind-wandering Questionnaire; MAAS-LO: Mindful Attention Awareness Scale-Lapses Only; ARCES: Attention-Related Cognitive Errors Scale; BPS: Boredom Proneness Scale; DSAQ: Doodle Spontaneous Activity Questionnaire; SAQ: Spontaneous Activity Questionnaire.
p < .01.
As can be seen in Table 4, there was a significant main effect of doodling condition on the mid-task level of boredom experienced by the subset of participant who self-reported high levels of ARCES. Inspection of the corresponding means and confidence intervals shows that boredom was lower for these high-ARCES participants that engaged in structured doodling, compared with those who engaged in unstructured doodling or just passively listened. The high-ARCES participants that engaged in structured doodling also reported nominally lower levels of mind-wandering and nominally higher levels of self-perceived attention, compared with those who engaged in unstructured doodling or just passively listened, although they also had nominally lower retention scores than either of these other conditions. None of the other measures for any of the other subsets of participants showed a significant effect of doodling condition. Thus, the lower level of mid-task boredom for high-ARCES participants is the only indication from our additional individual-difference analyses that any type of doodling could have any type of beneficial effect. To the extent that structured doodling was effective in reducing mid-task boredom for this specific subset of participants, it is noteworthy that this was not accompanied by a corresponding benefit for learning/remembering lecture content.
Individual differences: cognitive-affective correlates of fidgeting/doodling
To test our competing hypotheses regarding the cognitive-affective correlates of self-reported tendencies to doodle and fidget, we calculated the correlation between each of our individual-difference measures, including trait mind-wandering (MWQ), attentional lapses (MAAS-LO), ARCES, BPS, doodling behaviour (DSAQ), and fidgeting behaviour (SAQ), as well as between our self-reported measures of cognitive-affective states, including boredom before the listening tasks (pre-task MSBS1), during the task (average of Boredom ratings obtained at each of the four timepoints) and after the task (post-task MSBS2), self-perceived levels of attention (average of Attention ratings obtained at each of the four timepoints), mind-wandering (average of Mind-wandering ratings obtained at each of the four timepoints), plus the score on the subsequent lecture-retention exam. These are reported in Table 5.
Pearson product-moment correlations for all individual-difference measures, including trait mind-wandering (MWQ), attentional lapses (MAAS-LO), attention-related cognitive errors (ARCES), boredom proneness (BPS), doodling behaviour (DSAQ), and fidgeting behaviour (SAQ), as well as self-reported cognitive-affective states, including boredom before the listening tasks (pre-task MSBS1), during the task (average of Boredom ratings obtained at each of the four timepoints) and after the task (post-task MSBS1), self-perceived levels of attention (average of Attention ratings obtained at each of the four timepoints), mind-wandering (average of Mind-wandering ratings obtained at each of the four timepoints), plus the score on the subsequent lecture-retention exam.
MWQ: Mind-wandering Questionnaire; MAAS-LO: Mindful Attention Awareness Scale-Lapses Only; ARCES: Attention-Related Cognitive Errors Scale; BPS: Boredom Proneness Scale; DSAQ: Doodle Spontaneous Activity Questionnaire; SAQ: Spontaneous Activity Questionnaire.
p < .05. **p < .01. ***p < .001.
As shown in Table 5, our measures of individual differences in the tendency to doodle (DSAQ) and the tendency to fidget (SAQ) were positively correlated, which is consistent with the view that doodling is a specific form of the more general category of fidgeting. However, the fact that these measures were only moderately correlated makes clear that these measures do not merely tap into the exact same behavioural traits. Indeed, whereas the tendency to fidget was significantly positively correlated with the tendency to mind-wander, the tendency to doodle was not.
The fidgeting-reduces-boredom-and-increases-attention hypothesis predicts negative correlations between the self-reported tendency to doodle (DSAQ) and measures of boredom proneness (BPS), the tendency to experience lapses in attention (MAAS-LO), and the cognitive consequences of attentional errors (ARCES). In contrast, we found that DSAQ was not correlated with BPS, but was moderately positively correlated with both MAAS-LO and ARCES. Inspection of Table 5 also shows that DSAQ scores were positively correlated with the levels of state boredom reported both before (pre-task MSBS1) and after (post-task MSBS2) the lecture-listening task. These significant positive correlations between the tendency to doodle, levels of state boredom, and tendencies to experience attentional lapses/errors is therefore more consistent with the fidgeting-reflects-inattention hypothesis. Likewise, we observed that the tendency to fidget (SAQ) was also positively correlated with both MAAS-LO and ARCES (as well as MWQ and the levels of state boredom reported both before and after the lecture-listening task) and negatively correlated with the average level self-perceived attention paid to the listening task, providing added converging support for the fidgeting-reflects-inattention hypothesis. These results are not consistent with the fidgeting-reduces-boredom-and-increases-attention hypothesis.
Discussion
Experiment 2 aimed to further examine the competing fidgeting-related hypotheses. Recall that the fidgeting-reduces-boredom-and-increases-attention hypothesis predicts that participants in both the structured-doodling or unstructured-doodling conditions should show lower levels of in-task boredom and mind-wandering, and higher levels of in-task attention and subsequent memory for lecture content, than those in the control condition. However, the results failed to support this hypothesis. In terms of boredom, group allocation made no difference to boredom scores at any time point. In fact, boredom increased significantly over time for all groups, importantly including those who doodled. Similar results were seen for mind-wandering and attention self-reports over time, such that all groups reported increased mind-wandering and decreased attention as the lecture progressed. On top of that, the memory performance of doodlers was nominally worse than those who did nothing and significantly worse than those who took notes. In addition, this hypothesis would predict negative correlations between the DSAQ, BPS, ARCES, and MAAS-LO, yet, upon examination, this was not the case. The tendency to doodle was not significantly associated with boredom proneness and was positively associated with attentional failures/lapses.
The counter-hypothesis suggesting that fidgeting-reflects-inattention predicts that either doodling condition would reduce attention to the lecture, increase in-task boredom and mind-wandering and impair performance on the subsequent memory test when compared with that of the control condition. Specifically, those who are in the unstructured doodle should show reduced in-task attention, more mind-wandering, and worse memory for lecture content relative to the structured doodling condition. The results did not support this hypothesis either. Although attention did decrease, and boredom and mind-wandering increased for those in the doodle conditions, these results did not significantly differ from those who were in the control condition. Beyond that, the memory test findings did not support this hypothesis as scores were slightly worse than the control, but not enough to be significant. For individual differences, the hypothesis would suggest positive correlations between DSAQ scores and both ARCES scores and MAAS-LO scores. As stated previously, DSAQ was in fact positively correlated with attentional lapses and failures. The individual difference measures suggest that a higher tendency to doodle is associated with a higher tendency to experience lapses and/or failures in attention. As well as a higher tendency to doodle is related to higher state boredom. Thus, by examining individuals’ own tendency to doodle outside of a forced condition, we are better able to see how it relates to attentional engagement and state boredom.
General discussion
While traditionally viewed in educational contexts as markers of inattention and poor classroom behaviour doodling and fidgeting have more recently been considered as possible routes to improve performance. Across two experiments, we directly tested the competing hypotheses about the extent to which different methods of doodling may indeed be helpful for cognitive functioning: the “fidgeting reduces boredom and increases attention” hypothesis (e.g., Andrade, 2010), which posits that doodling is a beneficial form of fidgeting that can reduce boredom and increase attention to promote better learning, and the “fidgeting reflects inattention” hypothesis, which maintains that doodling is merely an indication of the mind taking a mental break, thereby reflecting the absence of task-focused attention (mind-wandering; Boggs et al., 2017) and is therefore linked to relatively poor learning. Across our studies, doodling neither reduced boredom or mind-wandering nor increased attention or retention of information when compared with conditions without doodling.
In Experiment 1 we sought to replicate and extend Andrade’s (2010) observation that doodling can reduce feelings of boredom and levels of mind-wandering while improving task-focused attention and memory for associated information. Instead, we found no evidence that doodling was any better than solely listening when it came to remembering task-relevant information. Indeed, participants who doodled did nominally worse on the memory assessment. The main difference between our study and Andrade’s (2010) study is that we used a boredom-induction procedure to ensure our participants were experience significantly elevated levels of state boredom prior to the listening task that contained the doodling manipulation. Thus, a potential reason why our results differ from those of Andrade’s (2010) study could be that participants were not experiencing the same levels of boredom prior to the task and thus they did not have the same difficulty staying engaged in the task in the first place. Boredom is thought to be a pervasive affective state that arises when we want to, but are unable to, engage attention in a satisfying activity (Eastwood et al., 2012). A study conducted by Sinclair et al. (2018) looked at positive and negative emotions across time during a computer-based learning program. The results from their study suggest that of all the emotions measured (e.g., boredom, frustration, pride, etc.) boredom was the only state that caused concern for students because of the relatively small chance of being able to escape from that emotion. Thus, it could also be that a more effective way to alleviate and prevent boredom would be to change the task altogether, rather than to add doodling.
In Experiment 2, we sought to further contrast the fidgeting-reduces-boredom-and-increases-attention hypothesis against the fidgeting-reflects-inattention hypothesis using a more ecologically valid task in which we manipulated different types of fidgeting behaviours to directly assess their impact on boredom, mind-wandering, and attention during a lecture-listening task, as well as the associated effects on retention of lecture material. We found that doodling neither reduced boredom or mind-wandering nor increased attention or retention of information compared with other conditions. In contrast, attention and test performance were highest (and boredom and mind-wandering lowest) for those focused solely on note-taking. Moreover, our inclusion of self-report measures of the tendency to doodle, fidget, experience boredom, and attentional lapses and failures revealed that higher levels of self-reported doodling behaviours were associated with higher levels of attentional lapses, attentional failures, and state boredom, which resonates more with the fidgeting-reflects-inattention hypothesis than the fidgeting-reduces-boredom-and-increases-attention hypothesis.
We formally considered the possibility that our failure to observe an overall benefit of doodling for mid-task levels of boredom, mind-wandering, and task-focused attention, or for subsequent memory performance in Experiment 2 was due to doodling and other forms of fidgeting only having cognitive-affective benefits for certain types of people, such as those with attention-related difficulties, those prone to experience boredom, or those who routinely engage in fidgeting or doodling. We found no evidence that doodling reduced boredom or mind-wandering, or increased task-focused attention or subsequent memory performance for individuals who are relatively high in routinely experiencing attention-related difficulties, including mind-wandering or attentional lapses, or who are relatively high in boredom proneness. Doodling also provided no observable benefits for those who are relatively high in the extent to which they routinely engage in doodling or fidgeting behaviours. Indeed, the only group for whom we found any benefit of doodling was for participants with relatively high scores on the ARCES scale: participants in this subset that engaged in structured doodling experienced less boredom than those that engaged in unstructured doodling or who solely listened to the lecture. The doodling-related benefit for this participant subset, however, did not lead to better retention of the lecture material.
Experiment 2 provided additional evidence confirming the benefits of taking notes for retaining crucial information. These findings replicate Boggs et al. (2017) and Meade et al. (2019) findings showing individuals are better able to retain attended content if they take notes rather than simply listening or doodling while listening. Boggs et al. hypothesised that this finding was because taking notes would enhance the encoding process. This seems plausible as previous research has shown that note-taking has an instant positive effect on memory providing individuals with a deeper level of processing (i.e., the encoding effect; Di Vesta & Gray, 1972; Peper & Mayer, 1978). Boggs et al. speculated that this finding was also due to note-taking maintaining arousal levels and thus reducing boredom, which is consistent with the result of our study. Note-takers throughout the lecture reported paying more attention and experiencing mind-wandering less than the other groups. Although taking notes could not eliminate boredom altogether, it did appear to mitigate the negative consequences associated with boredom (e.g., reduced performance, reduced attention, increases in mind-wandering). Therefore, promoting note-taking in situations in which it is important for students to be able to later recall associated information appears be a more effective strategy than promoting doodling.
It is important to note that doodling in the lab may not produce the same effects as doodling in real-world situations. In the lab, we take away the naturalism of doodling by placing participants in controlled conditions and then asking them to do a specific doodling-type task throughout the entire learning activity. Thus, not only may the type of doodling imposed upon them by an experimenter be different from the type of doodling they may spontaneously do on their own, but the ability to control when—and for how long—one doodles may be critical in determining the type of effects it may have on attention, boredom, and the ability to learn. Accordingly, the use of doodling manipulations such as those employed in our experiments may not be optimal operationalisations of the most important features of real-world doodling or fidgeting. This underscores the potential value of future research that not only investigates differences between experimenter-imposed versus self-chosen types of doodling, as well as the potential benefits of allowing participants to stagger doodling across shorter periods, or to be able to delay it to moments when they know that the presented information is not as useful. Future research might also benefit from considering the role that differences in motivation may play. In the lab, for example, there are no real consequences of failing to pay attention to the lecture material; the student completes the study and can receive their course credit regardless of how they performed. In a real-life lecture, however, the consequences of failing to maintain attentional engagement can include missing important information about an upcoming assignment or content for an exam. Thus, students in the lab may lack the motivation that happens for them in a classroom. Examining the causes and consequences of doodling in the classroom, rather than a lab, while probing students’ motivation to learn the material being presented may help to address the role of motivation in the link between doodling, boredom, mind-wandering, and the retention of lecture material.
It is also important to consider that while participants were monitored in both of our experiments to ensure they stayed on task (e.g., doodle if you are in the doodle condition and take notes if you are in the note-taking condition), they were not monitored for other forms of fidgeting behaviour (e.g., foot tapping, using the pencil to tap, and rocking in their chair). Future research would benefit from observing and recording any additional fidgeting behaviour that may occur across the different experimental conditions. In addition, although we examined fidgeting more generally in terms of differences in prevalence across individuals (e.g., SAQ), this study was primarily focused on doodling. Our results may therefore not generalise to other forms of fidgeting. With that said, it is noteworthy that the DSAQ and SAQ both positively correlated with the MAAS-LO, ARCES, MSBS1, MSBS2, as well as with each other. This suggests that while there may be intriguing things to learn about how different specific types of fidgeting may be associated with attention and boredom, there is nevertheless similarity in the general relation between attention, boredom and fidgeting, broadly speaking, and that between attention, boredom, and doodling, specifically.
Contrasting the fidgeting-reduces-boredom-and-increases-attention hypothesis and the fidgeting-reflects-inattention hypothesis is important for understanding how the mechanisms subserving attentional, behavioural, and affective engagement operate together and for resolving the conflict in prior evidence about the extent to which doodling is a viable method for educators and students to implement in the context of learning. With the overall consistency of our findings with the fidgeting-reflects-inattention hypothesis, our results suggest that the adoption of doodling exercises or other fidgeting-based interventions within classroom settings may not be an effective strategy for increasing classroom engagement or promoting learning.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Natural Science and Engineering Research Council of Canada (Discovery Grant #401526).
