- Email: [email protected]
Individual Differences in Working Memory Predict the Effect of Music on Student Performance
+Model
ARTICLE IN PRESS Journal of Applied Research in Memory and Cognition xxx (2017) xxx–xxx Contents lists available at ScienceDirect
Journal of Applied Research in Memory and Cognition journal homepage: www.elsevier.com/locate/jarmac
Individual Differences in Working Memory Predict the Effect of Music on Student Performance Eddie A. Christopher∗ University of Tennessee at Chattanooga, United States Purdue University, United States Jill Talley Shelton University of Tennessee at Chattanooga, United States Past research has demonstrated that music often negatively impacts performance on a variety of cognitive tasks, including academically relevant tasks. There are, however, discrepancies in the literature, including a handful of instances where no effect of music was observed. We tested the novel hypothesis that working-memory capacity moderated the detrimental effect of music on academic performance. Undergraduate students worked on readingcomprehension and math tasks under both music and silence conditions before completing a battery of workingmemory capacity assessments. Although music led to a significant decline in performance overall, working-memory capacity moderated this effect in the reading-comprehension tasks. These findings suggest that individuals who are better able to control their attention (as indexed by working-memory capacity) may be protected from music-related distraction when completing certain kinds of academically relevant tasks.
General Audience Summary Instructors of undergraduate psychology courses often inform their students of a finding that studying while listening to music hinders learning. Thus, the advice students often receive is that they ought not to attempt any sort of academic work while listening to music. Many students, however, profess a distrust of such a finding, and retain a belief that they in fact do better with music, despite evidence to the contrary. An important question then is whether the detrimental effect that music has on learning actually impacts everyone in the same way. The present study examined whether individual differences in people’s attention and memory abilities could predict the degree to which music impeded their performance of academic tasks. Interestingly, the higher an individual scored on the tasks measuring attention and memory abilities, the less they were affected by music while they worked on reading-comprehension questions. However, when solving math problems, the detrimental effect of music was similar, regardless of how individuals scored on the tasks measuring memory and attention. Keywords: Working memory capacity, Music, Auditory distraction, Academic performance
Many studies have demonstrated notable declines in performance on academic tasks completed in the presence of music relative to a silence condition (e.g., Anderson & Fuller, 2010; Henderson, Crews, & Barlow, 1945), although a subset of those studies yielded inconsistent results. For example, music
was detrimental to performance on some, but not all, tasks (Kantner, 2009; Tucker & Bushman, 1991), and the volume (Woo & Kanachi, 2005) or style of music (Henderson et al., 1945; Kantner, 2009; Mayfield & Moss, 1989; Woo & Kanachi, 2005) being played was a determining factor in whether or not
Author Note Corresponding author at: Correspondence concerning this article should be addressed to Eddie A. Christopher, Psychological Sciences, 703 Third Street, West Lafayette, IN 47907, United States. Contact: [email protected] ∗
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012
+Model
ARTICLE IN PRESS WORKING MEMORY, MUSIC, AND STUDENT PERFORMANCE
music affected performance. Other studies failed to observe any effect of music on performance (Abikoff, Courtney, Szeibel, & Koplewicz, 1996; Freeburne & Fleischer, 1952; Judde & Rickard, 2010; Mowsesian & Heyer, 1973; Pool, Koolstra, & Van Der Voort, 2003). Taken together, these studies suggest that there have been exceptions to the typical music-distraction effect that warrant further exploration. Some studies examined the effect of music on laboratorybased cognitive tasks such as the oddball task (Pacheco-Unguetti & Parmentier, 2014) and list learning (Judde & Rickard, 2010; Woo & Kanachi, 2005), but many others used more naturalistic tasks measuring academic skills. The most commonly used academically relevant tasks have been reading-comprehension assessments (Anderson & Fuller, 2010; Doyle & Furnham, 2012; Freeburne & Fleischer, 1952; Furnham & Bradley, 1997; Henderson et al., 1945; Perham & Currie, 2014; Pool et al., 2003; Tucker & Bushman, 1991). Only Tucker and Bushman (1991) and Pool et al. (2003) reported that reading-comprehension performance was unaffected by music. The impact of music on arithmetic performance has also yielded mixed results, with some studies revealing a decline in arithmetic performance during a music condition (Mayfield & Moss, 1989; Tucker & Bushman, 1991), and other studies yielding no music-distraction effect (Abikoff et al., 1996; Mowsesian & Heyer, 1973). A potential explanation for these discrepant findings is that the studies used different measures of a given construct (such as reading comprehension on the SAT vs. the GRE). It is also possible, however, that task-specific variations between studies did not fully explain why a music-distraction effect was not observed in select cases. We tested another potential explanation for the discrepant findings in the music-distraction literature: that individual differences in working-memory capacity (WMC) moderate the distracting effect of music. Performance on WMC tasks has successfully predicted individual differences in a variety of cognitive abilities including executive attention, which facilitates an individual’s ability to keep relevant items within conscious awareness while inhibiting irrelevant information (Engle, 2002; Kane, Bleckley, Conway, & Engle, 2001). In one such study, Conway, Cowan, and Bunting (2001) used a version of the dichotic-listening task wherein participants shadowed speech presented in one ear while ignoring speech presented in the other ear. During the task, the participant’s name was spoken in the stream of to-be-ignored speech. Afterwards, when participants were asked about hearing their name, individuals with lower WMC were much more likely to have noticed their name being presented. In failing to hear their name, participants who scored in the upper-quartile of WMC demonstrated an increased ability to inhibit the to-be-ignored speech. Music similarly represents a potentially distracting stimulus for students engaged in an academic task. Another avenue of study has been the frequency with which students choose to listen to music while studying or doing homework of their own volition. Robison and Unsworth (2015) reported that their participants believed they were being distracted when background noise simulating a noisy restaurant was introduced during a reading-comprehension task, but
2
students were less aware of a distracting effect of listening to music. Indeed, many students have reported that they prefer to listen to music at least some of the time while working on school assignments (Anderson & Fuller, 2010). Given the growing evidence that listening to music is not an optimal condition for completing academically relevant work (e.g. Anderson & Fuller, 2010), this behavior may represent a deficiency in metacognition. Metacognition can be generally understood as thinking about thinking, including the ability to regulate one’s own mental processes and activity (Flavell, 1979). In one study, Anderson and Fuller (2010) post-experimentally surveyed their participants’ preferences to assess how often they chose to listen to music while studying or doing homework. After controlling for the effect of condition (silence vs. music), participants who preferred to listen to music when studying did markedly worse on the reading-comprehension assessment. Thus, individuals who are less aware of how music is affecting their performance may be more susceptible to the detrimental effects of music. Replicating this finding was among the goals of the present study. We hypothesized that individuals who scored higher on tests of WMC would be less susceptible to the distracting effects of music when performing academic tasks, compared to those with lower WMC scores. Such a finding would facilitate a unifying explanation for inconsistencies within the music-distraction literature. Additionally, we sought to replicate the general finding of a relationship between WMC and metacognitive abilities (Dunlosky & Kane, 2007). Specifically, we predicted that lower WMC individuals would be more likely to choose to listen to music when completing academically relevant tasks. Method Participants and design Participants (N = 151) consisted of University of Tennessee at Chattanooga undergraduate students, whose ages ranged from 18 to 30 (M = 19.03). The majority of participants (78%) identified themselves as being undecided or majoring in something other than psychology. Participants were recruited from undergraduate psychology courses and received extra credit for participation. Data from a few of these participants was not included for analysis due to a computer issue (n = 8), cell phone usage (n = 3), and failure to follow instructions (n = 2). A 2 (Task: math vs. reading comprehension) × 2 (Auditory Condition: music vs. silence) × 2 (Order: music condition first vs. silence condition first) mixed-model design was used. Within-participant manipulations of auditory condition and task required all participants to work through two academically relevant tasks during both the presence of music and silence. A between-participants manipulation of order was used to measure any potential effects of the order in which participants completed their work in the music and silence conditions. Because participants completed the experiment in groups, the sample size of each order condition (music condition first: n = 82 vs. silence condition first: n = 56) was allowed to vary so that all of the participants in a given experimental session would be presented with either music or silence first. During pilot testing the specific academic questions that were presented during the music or
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012
+Model
ARTICLE IN PRESS WORKING MEMORY, MUSIC, AND STUDENT PERFORMANCE
silence condition were varied, but no differences in performance were observed. For this reason, the specific academic questions that were presented during the music and silence conditions in the main experiment were not counterbalanced. Materials Academic materials. A list of questions consisting of arithmetic and reading-comprehension problems was acquired from a Scholastic Aptitude Test (SAT) practice book (Princeton Review, 2008) to test participants’ ability to perform academically relevant tasks under a music and a silence condition. SAT questions were used because: (a) the SAT consists of educationally relevant tasks, and (b) good performance variability could be expected. Consistent with what would typically be encountered on the SAT, arithmetic questions varied by type, including problems from both algebra and geometry. All questions were presented in a multiple-choice format. During the readingcomprehension task participants read passages that were either two or three paragraphs long, and then answered a series of multiple-choice questions following each passage. The reading passage itself remained available until all of the relevant questions had been answered. Working memory. Three separate computerized measures of WMC were employed to create a more reliable assessment, and a composite measure was created based on participant’s average z-scores across the three tasks. This composite score was derived from performance on the modified lag task (Shelton, Metzger, & Elliott, 2007), the automated operation-span task (Unsworth, Heitz, Schrock, & Engle, 2005), and the letternumber-sequencing task (Shelton, Elliott, Hill, Calamia, & Gouvier, 2009). All three tasks could be expected to reliably measure WMC and were weighted equally. Each WMC task required approximately 15 min to complete. In the modified lag task participants were shown 40 lists of words with either 6 or 8 items per list. At the end of each list, participants were asked to recall the word “n” back (last word, 1-back, 2-back, or 3-back). The task required participants to continually update a changing list of words and to be ready at any point to recall a given word from the list. A participant’s overall score was calculated by multiplying the total number of items that they recalled correctly at each lag (last word, 1-back, 2-back, or 3-back) by a weighting factor corresponding to that lag (1, 2, 3, or 4 respectively), and then adding the products (see Experiment 3 of Shelton et al., 2007). During the automated operation-span task participants solved arithmetic equations, with each equation being immediately followed by a to-be-remembered letter. At the end of the series of equations, participants were asked to identify all of the letters in the order that they were presented. All participants completed 15 trials with 3–7 equations/letters per trial. The participant’s score was the proportion of letters correctly recalled in order across all trials in which they answered at least 80% of the arithmetic questions correctly. All participants successfully answered 80% of these arithmetic questions. The letter-number-sequencing task exposed participants to lists of varying length comprising both letters and numbers. All participants saw 21 lists consisting of 3–9 items, with each participant
3
receiving 3 lists at each list length. Participants were asked to respond at the end of the lists with the letters and then numbers in alphabetical and numerical order. A participant’s score was the total number of items correctly recalled in order across all trials. Metacognitive measures. Metacognition was assessed in two ways. First, participants were asked to predict how well they would perform on the arithmetic and reading-comprehension tasks on a scale of 1–100. All participants were informed of the impending type of trial (music vs. silence) before they made their metacognitive predictions. Second, a post-experimental questionnaire was administered to assess study habits and music listening preferences. Students were asked to describe the percentage (0–100%) of time they typically spent listening to music while performing academic activities such as homework, and to describe on a scale of 1–9 how much they believe music helps/hurts their performance. Music. The music played during the music condition consisted of songs randomly selected from the Billboard Top 100, and represented currently popular songs that were familiar to most participants. We reasoned that using currently popular music with vocals would be more applicable to the everyday study habits of students and should maximize the distracting effect of the music. Instrumental music has been shown in previous research to be less distracting compared to music with words because of the varying degrees of information one has to selectively inhibit (Freeburne & Fleischer, 1952). Similarly, popular music has been found to be increasingly distracting for persons when compared to older, less popular music (Henderson et al., 1945; Woo & Kanachi, 2005). Procedure All participants completed the experiment in groups of 2–10 while seated at individual computer stations with Dell desktop computers. The experimental tasks were constructed using E-prime 2.0 (Psychology Software Tools, Pittsburgh, PA). Participants filled out the informed consent using pencil and paper prior to the first trial. Next, participants completed arithmetic and reading-comprehension problems from the SAT under both silence and music conditions while wearing Sony noise-canceling headphones. All participants were presented with the arithmetic problems first, followed by the reading-comprehension problems. Approximately half of the participants completed the SAT problems in the silence followed by the music condition, and the other participants first completed problems in the music condition followed by a silence condition. Each block of either music or silence lasted for 20 min (10 min per academic task). Within each section, participants were allowed to return to previous questions until they had exhausted their time. No participant finished the questions early. Once participants finished the academic tasks under music and silence conditions they immediately progressed to the three WMC tests. All participants completed the modified lag task, followed by letter number sequencing, and finally operation span. After finishing all experimental trials, participants were given a survey to assess their own music
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012
+Model
ARTICLE IN PRESS WORKING MEMORY, MUSIC, AND STUDENT PERFORMANCE
4
Table 1 Pearson Correlations and Cronbach’s Alpha for All Tasks Task
MM
MRC
SM
SRC
MM MRC SM SRC O-Span Mod. Lag. LNS
.64 .29 .38 .27 .25 .43 .22
.57 .32 .22 .21 .32 .30
.57 .38 .25 .43 .27
.72 .11 .29 .19
O-Span
.79 .34 .33
Mod. Lag.
.63 .39
LNS
.74
Note. N = 138. Cronbach’s alpha is provided along the diagonal in boldface, and correlation coefficients are provided throughout the rest of the table in regular print. MM = music: math; MRC = music: reading comprehension; SM = silence: math; SRC = silence: reading comprehension; O-Span = operation span; Mod. Lag. = modified lag task; LNS = letter number sequencing. Correlations > .2 are significant at p < .05.
listening habits and preferences while completing homework and studying. Results Scoring WMC was treated as a continuous variable and was based on a composite score created from three separate measures of WMC. Performance on each WMC task was standardized and then averaged together to create a measure of WMC. This approach allowed for a measurement of the effect of individual differences in WMC across a continuous range of WMC scores. The degree to which participants were accurate in predicting their performance on a given task (e.g., proportion correct on math questions during the playing of music) was calculated with a calibration score. Correlations between measures, as well as reliabilities for all measures, are presented in Table 1. Music Effect and Working Memory For all analyses the significance level was set at .05. A 2 (Task: math vs. reading comprehension) × 2 (Auditory condition: music vs. silence) × 2 (Order: music condition first vs. silence condition first) mixed-model Analysis of Covariance (ANCOVA) was used with performance on the WMC tasks included as a covariate to predict accuracy (proportion correct) on the academically relevant tasks. There was no main effect of order nor did it interact with any variables (all F’s < 1.1). A main effect of auditory condition was observed, F(1, 136) = 12.38, p < .001, MSE = 0.29, η2 = .03. Furthermore, there was a main effect of WMC, F(1, 136) = 25.73, p < .001, MSE = 1.37, η2 = .03. However, the interaction between auditory condition and WMC was not statistically significant, F(1, 136) = 2.74, p = .101, MSE = 2.74, η2 < .01. The main effect of task on performance was also not statistically significant, F(1, 136) = 1.23, p = .269, MSE = 0.03, η2 < .01. Moreover, there was no significant interaction between task and auditory condition, or task and WMC (both F’s < 2). Importantly, there was a qualifying three-way interaction between auditory condition, task, and WMC, F(1, 136) = 4.00, p = .048, MSE = 0.08, η2 = .01. Although performance was worse in the presence of music in both the reading-comprehension (M = .38, SD = .18) and math (M = .38,
SD = .19) tasks relative to when the reading-comprehension (M = .41, SD = .16) and math (M = .44, SD = .20) tasks were performed in silence, the degree to which WMC impacted the effect of auditory condition needed to be evaluated in each task separately (performance across tasks for music: M = .38, SD = .15 and silence: M = .43, SD = .15). Two separate repeated measures ANCOVAs were conducted for the reading-comprehension and math tasks to follow-up this three-way interaction. In the following ANCOVAs auditory condition varied within participants, and performance on the WMC tasks was included as a covariate. The first analysis revealed that the difference in performance on the reading-comprehension task between the music and silence conditions was less pronounced for participants with higher WMC scores, F(1, 136) = 6.72, p = .011, MSE = 0.14, η2 = .05. A separate ANCOVA revealed that this was not the case on the math task, F(1, 136) = 0.04, p = .836, MSE < 0.01, η2 < .01. Thus, the observed three-way interaction was driven by WMC composite scores moderating the music distraction effect for the reading-comprehension but not the math task (Figure 1). Metacognitive Knowledge Paired-samples t-tests were used to compare participants’ predictions for their performance across conditions, and none of these comparisons approached statistical significance. On both tasks, participants’ predictions of how well they would perform, and the degree to which they were accurate in those predictions, did not vary between the music and silence conditions (all t’s < 1.3). Furthermore, the accuracy of performance predictions across academic tasks was not significantly correlated with performance on WMC tasks during the playing of music, r(136) = −.02, p = .815 or during silence, r(136) = .03, p = .726. The self-reported percentage of time that participants listen to music while doing academically relevant tasks (M = 56%, SD = 34%), such as homework, was not significantly related to any potentially relevant factors such as the effect of music on the math, r(136) = −.03, p = .726 or reading-comprehension, r(136) = .10, p = .242 tasks, or WMC, r(136) = −.01, p = .916. Similarly, the degree to which participants believed that music affected them was not significantly correlated to the effect that music had on their math, r(136) = −.07, p = .413 or readingcomprehension, r(136) = .01, p = .907 performance.
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012
+Model
ARTICLE IN PRESS WORKING MEMORY, MUSIC, AND STUDENT PERFORMANCE
5
Figure 1. The moderating role of WMC in the effect of music on performance. For graphical purposes, the effect of music was calculated by subtracting the proportion correct during silence from the proportion correct during music on a given task. Therefore, negative scores represent lower accuracy during the playing of music.
Discussion The present study replicated previous research demonstrating that performance of academically relevant tasks suffered under a music condition compared to a silence condition (e.g., Anderson & Fuller, 2010; Henderson et al., 1945). Importantly, the present findings offered an exciting extension to this line of research by providing a novel explanation for why some have failed to observe the music distraction effect, as individual differences in WMC had not previously been considered as a moderator. It may be that in previous studies where an effect of music was not observed, a sufficiently large proportion of the sample had high WMC, which would have hindered any attempt to observe the typically detrimental effect of music. The implication of the present study is that, indeed, some students are correct in their assumption that it is relatively safe for them to listen to music while doing their homework. Specifically, our data suggest that individuals who score higher on WMC tests perform reading-comprehension tasks just as well when they are listening to music relative to performing the task in silence. However, individual differences in WMC did not provide a buffer against music-related distraction in math tasks. The finding that individuals with higher WMC scores could safely perform reading-comprehension tasks while listening to music dovetails with previous observations of higher-WMC individuals better employing executive attention skills (Conway et al., 2005; Kane et al., 2001). We had predicted that high WMC would similarly facilitate performance on arithmetic tasks, regardless of the auditory condition. The failure to observe this finding could reflect the fact that arithmetic tasks are supported by different cognitive processes than reading-comprehension tasks. For example, reading-comprehension tasks involve encoding information into long-term memory (Atkinson & Shiffrin, 1968), a process not necessarily inherent in arithmetic tasks. Notably, all of the WMC tasks used in the present study were verbal in nature, which could have diminished their predictive utility for the math task. Consistent with this premise, children with specific arithmetic deficits have produced significantly lower scores on the spatial, but not the verbal, component of WMC relative to children with no specific arithmetic deficits (McLean
& Hitch, 1999). One explanation then, is that higher WMC will most effectively moderate a distracting environmental stimulus (e.g., music) when there is a match between the modality of the distractor and the WMC system being measured. However, there is also research suggesting that WMC is a unitary construct, and that separate stores for verbal and visuo-spatial information are not necessary (Cowan & Morey, 2007; Kane et al., 2004). Thus, a fruitful direction for future research would be to investigate whether individual differences in spatial WMC could, indeed, moderate the music-distraction effect that has been observed in math tasks. Contrary to what Anderson and Fuller (2010) found, our results suggest that although some individuals can do their homework and listen to music at the same time, knowledge of one’s own ability may be tenuous. The current study had an older sample than did Anderson and Fuller (2010) who used junior high school students. It may be that by the time students reach college they believe they have developed sufficient strategies for working on academically relevant tasks while listening to music. Furthermore, the way in which beliefs regarding the effect of music were measured in the present study varied from Anderson and Fuller (2010) who questioned their participants regarding whether or not they listened to music while doing homework using a 5-point Likert scale. Conversely, participants in the present study were asked to report their music listening habits as a percentage, which allowed for a more exact representation of predicted behavior. It is also worth noting that WMC scores in the present study were surprisingly not associated with participants’ ability to accurately predict their performance on the arithmetic or reading-comprehension tasks across auditory conditions. Given that the presence of music was associated with a drop in performance, a lack of awareness of the detrimental effect of music likely represented a deficiency in metacognition. Greater metacognitive abilities should lead individuals to detect and avoid the effect of music whenever possible. Typically, increased WMC has been associated with better metacognitive abilities (Dunlosky & Kane, 2007; Thomas, Bonura, Taylor, & Brunyé, 2012), but in the present study the way in which music impacted performance represented a blind-spot for participants across the
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012
+Model
ARTICLE IN PRESS WORKING MEMORY, MUSIC, AND STUDENT PERFORMANCE
spectrum of WMC. However, it is possible that participants in the present study listened to music while working on academic material outside of the lab for reasons other than poor metacognition. In particular, it may be that many students in a university setting face a limited availability of quiet places to do homework and study. The participants in this study may have believed (and perhaps correctly) that music is less detrimental than the potentially chaotic noises that they encounter in their campus environment. Future research should investigate the degree to which students believe music to be preferable to silence, as opposed to other noisy conditions. In conclusion, for most students it would be a mistake to listen to music while working on academically relevant tasks. However, higher-WMC individuals may be less vulnerable to performance deficits traditionally associated with listening to music. Unfortunately, no evidence was revealed to suggest that these individuals know who they are, so many students may listen to music despite the detrimental effects this choice has on their learning. The findings from the present study also have important theoretical implications as they provide one potential explanation for why inconsistencies have been observed in the musicdistraction literature. Future research is needed to elucidate which other academically relevant tasks are affected by music, and to what extent individual difference factors other than WMC contribute to one’s susceptibility to music-related distractions. Author Contributions Eddie A. Christopher conceived of the project, and both authors contributed equally to the planning of the experiment. Eddie A. Christopher was primarily responsible for data collection and analysis. Eddie A. Christopher was primarily responsible for writing the manuscript. Conflict of Interest Statement The authors declare no conflict of interest. Acknowledgements This manuscript served as the Master’s thesis for Eddie A. Christopher. Portions of these data were presented at the annual meeting of the Psychonomic Society held in Chicago, IL (November, 2015). We would like to thank Iain Scott and Emily Van Zandbergen for their assistance with data collection. Furthermore, we would like to thank Dr. Mike Biderman and Dr. Amanda Clark for their valuable contributions as committee members. References Abikoff, H., Courtney, M. E., Szeibel, P. J., & Koplewicz, H. S. (1996). The effects of auditory stimulation on the arithmetic performance of children with ADHD and nondisabled children. Journal of Learning Disabilities, 29(3), 238–246. http://dx.doi.org/10.1177/002221949602900302 Anderson, S. A., & Fuller, G. B. (2010). Effect of music on reading comprehension of junior high school students. School Psychology Quarterly, 25(3), 178–187. http://dx.doi.org/10.1037/a0021213
6
Atkinson, R. C., & Shiffrin, R. M. (1968). Human memory: A proposed system and its control processes. In The psychology of learning and motivation. Oxford, England: Academic Press. http://dx.doi.org/10.1016/S007967421(08)6042263 Conway, A. A., Cowan, N., & Bunting, M. F. (2001). The cocktail party phenomenon revisited: The importance of working memory capacity. Psychonomic Bulletin & Review, 8(2), 331–335. http://dx.doi.org/10.3758/BF03196169 Conway, A. A., Kane, M. J., Bunting, M. F., Hambrick, D., Wilhelm, O., & Engle, R. W. (2005). Working memory span tasks: A methodological review and user’s guide. Psychonomic Bulletin & Review, 12(5), 769–786. http://dx.doi.org/10.3758/BF03196772 Cowan, N., & Morey, C. C. (2007). How can dual-task working memory retention limits be investigated? Psychological Science, 18(8), 686–688. http://dx.doi.org/10.1111/j.1467-9280.2007.01960.x Doyle, M., & Furnham, A. (2012). The distracting effects of music on the cognitive test performance of creative and noncreative individuals. Thinking Skills and Creativity, 7(1), 1–7. http://dx.doi.org/10.1016/j.tsc.2011.09.002 Dunlosky, J., & Kane, M. J. (2007). The contributions of strategy use to working memory span: A comparison of strategy assessment methods. The Quarterly Journal of Experimental Psychology, 60, 1227–1245. http://dx.doi.org/10.1080/17470210600926075 Engle, R. W. (2002). Working memory capacity as executive attention. Current Directions in Psychological Science, 11(1), 19–23. http://dx.doi.org/10.1111/1467-8721.00160 Flavell, J. H. (1979). Metacognition and cognitive monitoring: A new area of cognitive–developmental inquiry. American Psychologist, 34(10), 906–911. http://dx.doi.org/10.1037/0003-066X.34.10.906 Furnham, A., & Bradley, A. (1997). Music while you work: The differential distraction of background music on the cognitive test performance of introverts and extroverts. Applied Cognitive Psychology, 11(5), 445–455. http://dx.doi.org/10.1002/(SICI)10990720(199710)11:5<445::AID-ACP472>3.0.CO;2-R Freeburne, C. M., & Fleischer, M. S. (1952). The effect of music distraction upon reading rate and comprehension. Journal of Educational Psychology, 43(2), 101–109. http://dx.doi.org/10.1037/h0054219 Henderson, M. T., Crews, A. A., & Barlow, J. J. (1945). A study of the effect of music distraction on reading efficiency. Journal of Applied Psychology, 29(4), 313–317. http://dx.doi.org/10.1037/h0056128 Judde, S., & Rickard, N. (2010). The effect of post- learning presentation of music on long-term word- list retention. Neurobiology of Learning and Memory, 94(1), 13–20. http://dx.doi.org/10.1016/ j.nlm.2010.03.002 Kane, M. J., Bleckley, M., Conway, A. A., & Engle, R. W. (2001). A controlled-attention view of working- memory capacity. Journal of Experimental Psychology: General, 130(2), 169–183. http://dx.doi.org/10.1037/0096-3445.130.2.169 Kane, M. J., Hambrick, D. Z., Tuholski, S. W., Wilhelm, O., Payne, T. W., & Engle, R. W. (2004). The generality of working memory capacity: A latent-variable approach to verbal and visuospatial memory span and reasoning. Journal of Experimental Psychology: General, 133(2), 189–217. http://dx.doi.org/10.1037/ 0096-3445.133.2.189 Kantner, J. (2009). Studying with music: Is the irrelevant speech effect relevant? In M. R. Kelley (Ed.), Applied memory (pp. 19–40). Hauppauge, NY, US: Nova Science Publishers. Mayfield, C., & Moss, S. (1989). Effect of music tempo on task performance. Psychological Reports, 65(3, Pt 2), 1283–1290. http://dx.doi.org/10.2466/pr0.1989.65.3f.1283 McLean, J. F., & Hitch, G. J. (1999). Working memory impairments in children with specific arithmetic learning difficulties.
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012
+Model
ARTICLE IN PRESS WORKING MEMORY, MUSIC, AND STUDENT PERFORMANCE
Journal of Experimental Child Psychology, 74(3), 240–260. http://dx.doi.org/10.1006/jecp.1999.2516 Mowsesian, R., & Heyer, M. R. (1973). The effect of music as a distraction on test-taking performance. Measurement & Evaluation in Guidance, 6(2), 104–110. Pacheco-Unguetti, A. P., & Parmentier, F. R. (2014). Sadness increases distraction by auditory deviant stimuli. Emotion, 203–213. http://dx.doi.org/10.1037/a0034289 Perham, N., & Currie, H. (2014). Does listening to preferred music improve reading comprehension performance? Applied Cognitive Psychology, 28(2), 279–284. http://dx.doi.org/10.1002/acp.2994 Pool, M. M., Koolstra, C. M., & Van Der Voort, T. A. (2003). Distraction effects of background soap operas on homework performance: An experimental study enriched with observational data. Educational Psychology, 23(4), 361–380. http://dx.doi.org/ 10.1080/01443410303211 Princeton Review. (2008). 11 practice tests for the SAT and PSAT. Natick, MA: Random House. Robison, M. K., & Unsworth, N. (2015). Working memory capacity offers resistance to mind-wandering and external distraction in a context-specific manner. Applied Cognitive Psychology, 680–690. http://dx.doi.org/10.1002/acp.3150 Shelton, J. T., Elliott, E. M., Hill, B. D., Calamia, M. R., & Gouvier, W. D. (2009). A comparison of laboratory and clinical working memory tests and their prediction of fluid intelligence. Intelligence, 37(3), 283–293. http://dx.doi.org/10.1016/j.intell.2008.11.005
7
Shelton, J. T., Metzger, R. L., & Elliott, E. M. (2007). A group-administered lag task as a measure of working memory. Behavior Research Methods, 39(3), 482–493. http://dx.doi.org/ 10.3758/BF03193017 Thomas, A. K., Bonura, B. M., Taylor, H. A., & Brunyé, T. T. (2012). Metacognitive monitoring in visuospatial working memory. Psychology and Aging, 27(4), 1099–1110. http://dx.doi.org/ 10.1037/a0028556 Tucker, A., & Bushman, B. J. (1991). Effects of rock and roll music on mathematical, verbal, and reading comprehension performance. Perceptual and Motor Skills, 72(3, Pt 1), 942. http://dx.doi.org/10.2466/PMS.72.3.942-942 Unsworth, N., Heitz, R. P., Schrock, J. C., & Engle, R. W. (2005). An automated version of the operation span task. Behavior Research Methods, 37, 498–505. http://dx.doi.org/10.3758/BF03192720 Woo, E. W., & Kanachi, M. (2005). The effects of music type and volume on short-term memory. Tohoku Psychologica Folia, 64, 68–76.
Received 8 July 2016; accepted 30 January 2017 Available online xxx
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012
ARTICLE IN PRESS Journal of Applied Research in Memory and Cognition xxx (2017) xxx–xxx Contents lists available at ScienceDirect
Journal of Applied Research in Memory and Cognition journal homepage: www.elsevier.com/locate/jarmac
Individual Differences in Working Memory Predict the Effect of Music on Student Performance Eddie A. Christopher∗ University of Tennessee at Chattanooga, United States Purdue University, United States Jill Talley Shelton University of Tennessee at Chattanooga, United States Past research has demonstrated that music often negatively impacts performance on a variety of cognitive tasks, including academically relevant tasks. There are, however, discrepancies in the literature, including a handful of instances where no effect of music was observed. We tested the novel hypothesis that working-memory capacity moderated the detrimental effect of music on academic performance. Undergraduate students worked on readingcomprehension and math tasks under both music and silence conditions before completing a battery of workingmemory capacity assessments. Although music led to a significant decline in performance overall, working-memory capacity moderated this effect in the reading-comprehension tasks. These findings suggest that individuals who are better able to control their attention (as indexed by working-memory capacity) may be protected from music-related distraction when completing certain kinds of academically relevant tasks.
General Audience Summary Instructors of undergraduate psychology courses often inform their students of a finding that studying while listening to music hinders learning. Thus, the advice students often receive is that they ought not to attempt any sort of academic work while listening to music. Many students, however, profess a distrust of such a finding, and retain a belief that they in fact do better with music, despite evidence to the contrary. An important question then is whether the detrimental effect that music has on learning actually impacts everyone in the same way. The present study examined whether individual differences in people’s attention and memory abilities could predict the degree to which music impeded their performance of academic tasks. Interestingly, the higher an individual scored on the tasks measuring attention and memory abilities, the less they were affected by music while they worked on reading-comprehension questions. However, when solving math problems, the detrimental effect of music was similar, regardless of how individuals scored on the tasks measuring memory and attention. Keywords: Working memory capacity, Music, Auditory distraction, Academic performance
Many studies have demonstrated notable declines in performance on academic tasks completed in the presence of music relative to a silence condition (e.g., Anderson & Fuller, 2010; Henderson, Crews, & Barlow, 1945), although a subset of those studies yielded inconsistent results. For example, music
was detrimental to performance on some, but not all, tasks (Kantner, 2009; Tucker & Bushman, 1991), and the volume (Woo & Kanachi, 2005) or style of music (Henderson et al., 1945; Kantner, 2009; Mayfield & Moss, 1989; Woo & Kanachi, 2005) being played was a determining factor in whether or not
Author Note Corresponding author at: Correspondence concerning this article should be addressed to Eddie A. Christopher, Psychological Sciences, 703 Third Street, West Lafayette, IN 47907, United States. Contact: [email protected] ∗
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012
+Model
ARTICLE IN PRESS WORKING MEMORY, MUSIC, AND STUDENT PERFORMANCE
music affected performance. Other studies failed to observe any effect of music on performance (Abikoff, Courtney, Szeibel, & Koplewicz, 1996; Freeburne & Fleischer, 1952; Judde & Rickard, 2010; Mowsesian & Heyer, 1973; Pool, Koolstra, & Van Der Voort, 2003). Taken together, these studies suggest that there have been exceptions to the typical music-distraction effect that warrant further exploration. Some studies examined the effect of music on laboratorybased cognitive tasks such as the oddball task (Pacheco-Unguetti & Parmentier, 2014) and list learning (Judde & Rickard, 2010; Woo & Kanachi, 2005), but many others used more naturalistic tasks measuring academic skills. The most commonly used academically relevant tasks have been reading-comprehension assessments (Anderson & Fuller, 2010; Doyle & Furnham, 2012; Freeburne & Fleischer, 1952; Furnham & Bradley, 1997; Henderson et al., 1945; Perham & Currie, 2014; Pool et al., 2003; Tucker & Bushman, 1991). Only Tucker and Bushman (1991) and Pool et al. (2003) reported that reading-comprehension performance was unaffected by music. The impact of music on arithmetic performance has also yielded mixed results, with some studies revealing a decline in arithmetic performance during a music condition (Mayfield & Moss, 1989; Tucker & Bushman, 1991), and other studies yielding no music-distraction effect (Abikoff et al., 1996; Mowsesian & Heyer, 1973). A potential explanation for these discrepant findings is that the studies used different measures of a given construct (such as reading comprehension on the SAT vs. the GRE). It is also possible, however, that task-specific variations between studies did not fully explain why a music-distraction effect was not observed in select cases. We tested another potential explanation for the discrepant findings in the music-distraction literature: that individual differences in working-memory capacity (WMC) moderate the distracting effect of music. Performance on WMC tasks has successfully predicted individual differences in a variety of cognitive abilities including executive attention, which facilitates an individual’s ability to keep relevant items within conscious awareness while inhibiting irrelevant information (Engle, 2002; Kane, Bleckley, Conway, & Engle, 2001). In one such study, Conway, Cowan, and Bunting (2001) used a version of the dichotic-listening task wherein participants shadowed speech presented in one ear while ignoring speech presented in the other ear. During the task, the participant’s name was spoken in the stream of to-be-ignored speech. Afterwards, when participants were asked about hearing their name, individuals with lower WMC were much more likely to have noticed their name being presented. In failing to hear their name, participants who scored in the upper-quartile of WMC demonstrated an increased ability to inhibit the to-be-ignored speech. Music similarly represents a potentially distracting stimulus for students engaged in an academic task. Another avenue of study has been the frequency with which students choose to listen to music while studying or doing homework of their own volition. Robison and Unsworth (2015) reported that their participants believed they were being distracted when background noise simulating a noisy restaurant was introduced during a reading-comprehension task, but
2
students were less aware of a distracting effect of listening to music. Indeed, many students have reported that they prefer to listen to music at least some of the time while working on school assignments (Anderson & Fuller, 2010). Given the growing evidence that listening to music is not an optimal condition for completing academically relevant work (e.g. Anderson & Fuller, 2010), this behavior may represent a deficiency in metacognition. Metacognition can be generally understood as thinking about thinking, including the ability to regulate one’s own mental processes and activity (Flavell, 1979). In one study, Anderson and Fuller (2010) post-experimentally surveyed their participants’ preferences to assess how often they chose to listen to music while studying or doing homework. After controlling for the effect of condition (silence vs. music), participants who preferred to listen to music when studying did markedly worse on the reading-comprehension assessment. Thus, individuals who are less aware of how music is affecting their performance may be more susceptible to the detrimental effects of music. Replicating this finding was among the goals of the present study. We hypothesized that individuals who scored higher on tests of WMC would be less susceptible to the distracting effects of music when performing academic tasks, compared to those with lower WMC scores. Such a finding would facilitate a unifying explanation for inconsistencies within the music-distraction literature. Additionally, we sought to replicate the general finding of a relationship between WMC and metacognitive abilities (Dunlosky & Kane, 2007). Specifically, we predicted that lower WMC individuals would be more likely to choose to listen to music when completing academically relevant tasks. Method Participants and design Participants (N = 151) consisted of University of Tennessee at Chattanooga undergraduate students, whose ages ranged from 18 to 30 (M = 19.03). The majority of participants (78%) identified themselves as being undecided or majoring in something other than psychology. Participants were recruited from undergraduate psychology courses and received extra credit for participation. Data from a few of these participants was not included for analysis due to a computer issue (n = 8), cell phone usage (n = 3), and failure to follow instructions (n = 2). A 2 (Task: math vs. reading comprehension) × 2 (Auditory Condition: music vs. silence) × 2 (Order: music condition first vs. silence condition first) mixed-model design was used. Within-participant manipulations of auditory condition and task required all participants to work through two academically relevant tasks during both the presence of music and silence. A between-participants manipulation of order was used to measure any potential effects of the order in which participants completed their work in the music and silence conditions. Because participants completed the experiment in groups, the sample size of each order condition (music condition first: n = 82 vs. silence condition first: n = 56) was allowed to vary so that all of the participants in a given experimental session would be presented with either music or silence first. During pilot testing the specific academic questions that were presented during the music or
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012
+Model
ARTICLE IN PRESS WORKING MEMORY, MUSIC, AND STUDENT PERFORMANCE
silence condition were varied, but no differences in performance were observed. For this reason, the specific academic questions that were presented during the music and silence conditions in the main experiment were not counterbalanced. Materials Academic materials. A list of questions consisting of arithmetic and reading-comprehension problems was acquired from a Scholastic Aptitude Test (SAT) practice book (Princeton Review, 2008) to test participants’ ability to perform academically relevant tasks under a music and a silence condition. SAT questions were used because: (a) the SAT consists of educationally relevant tasks, and (b) good performance variability could be expected. Consistent with what would typically be encountered on the SAT, arithmetic questions varied by type, including problems from both algebra and geometry. All questions were presented in a multiple-choice format. During the readingcomprehension task participants read passages that were either two or three paragraphs long, and then answered a series of multiple-choice questions following each passage. The reading passage itself remained available until all of the relevant questions had been answered. Working memory. Three separate computerized measures of WMC were employed to create a more reliable assessment, and a composite measure was created based on participant’s average z-scores across the three tasks. This composite score was derived from performance on the modified lag task (Shelton, Metzger, & Elliott, 2007), the automated operation-span task (Unsworth, Heitz, Schrock, & Engle, 2005), and the letternumber-sequencing task (Shelton, Elliott, Hill, Calamia, & Gouvier, 2009). All three tasks could be expected to reliably measure WMC and were weighted equally. Each WMC task required approximately 15 min to complete. In the modified lag task participants were shown 40 lists of words with either 6 or 8 items per list. At the end of each list, participants were asked to recall the word “n” back (last word, 1-back, 2-back, or 3-back). The task required participants to continually update a changing list of words and to be ready at any point to recall a given word from the list. A participant’s overall score was calculated by multiplying the total number of items that they recalled correctly at each lag (last word, 1-back, 2-back, or 3-back) by a weighting factor corresponding to that lag (1, 2, 3, or 4 respectively), and then adding the products (see Experiment 3 of Shelton et al., 2007). During the automated operation-span task participants solved arithmetic equations, with each equation being immediately followed by a to-be-remembered letter. At the end of the series of equations, participants were asked to identify all of the letters in the order that they were presented. All participants completed 15 trials with 3–7 equations/letters per trial. The participant’s score was the proportion of letters correctly recalled in order across all trials in which they answered at least 80% of the arithmetic questions correctly. All participants successfully answered 80% of these arithmetic questions. The letter-number-sequencing task exposed participants to lists of varying length comprising both letters and numbers. All participants saw 21 lists consisting of 3–9 items, with each participant
3
receiving 3 lists at each list length. Participants were asked to respond at the end of the lists with the letters and then numbers in alphabetical and numerical order. A participant’s score was the total number of items correctly recalled in order across all trials. Metacognitive measures. Metacognition was assessed in two ways. First, participants were asked to predict how well they would perform on the arithmetic and reading-comprehension tasks on a scale of 1–100. All participants were informed of the impending type of trial (music vs. silence) before they made their metacognitive predictions. Second, a post-experimental questionnaire was administered to assess study habits and music listening preferences. Students were asked to describe the percentage (0–100%) of time they typically spent listening to music while performing academic activities such as homework, and to describe on a scale of 1–9 how much they believe music helps/hurts their performance. Music. The music played during the music condition consisted of songs randomly selected from the Billboard Top 100, and represented currently popular songs that were familiar to most participants. We reasoned that using currently popular music with vocals would be more applicable to the everyday study habits of students and should maximize the distracting effect of the music. Instrumental music has been shown in previous research to be less distracting compared to music with words because of the varying degrees of information one has to selectively inhibit (Freeburne & Fleischer, 1952). Similarly, popular music has been found to be increasingly distracting for persons when compared to older, less popular music (Henderson et al., 1945; Woo & Kanachi, 2005). Procedure All participants completed the experiment in groups of 2–10 while seated at individual computer stations with Dell desktop computers. The experimental tasks were constructed using E-prime 2.0 (Psychology Software Tools, Pittsburgh, PA). Participants filled out the informed consent using pencil and paper prior to the first trial. Next, participants completed arithmetic and reading-comprehension problems from the SAT under both silence and music conditions while wearing Sony noise-canceling headphones. All participants were presented with the arithmetic problems first, followed by the reading-comprehension problems. Approximately half of the participants completed the SAT problems in the silence followed by the music condition, and the other participants first completed problems in the music condition followed by a silence condition. Each block of either music or silence lasted for 20 min (10 min per academic task). Within each section, participants were allowed to return to previous questions until they had exhausted their time. No participant finished the questions early. Once participants finished the academic tasks under music and silence conditions they immediately progressed to the three WMC tests. All participants completed the modified lag task, followed by letter number sequencing, and finally operation span. After finishing all experimental trials, participants were given a survey to assess their own music
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012
+Model
ARTICLE IN PRESS WORKING MEMORY, MUSIC, AND STUDENT PERFORMANCE
4
Table 1 Pearson Correlations and Cronbach’s Alpha for All Tasks Task
MM
MRC
SM
SRC
MM MRC SM SRC O-Span Mod. Lag. LNS
.64 .29 .38 .27 .25 .43 .22
.57 .32 .22 .21 .32 .30
.57 .38 .25 .43 .27
.72 .11 .29 .19
O-Span
.79 .34 .33
Mod. Lag.
.63 .39
LNS
.74
Note. N = 138. Cronbach’s alpha is provided along the diagonal in boldface, and correlation coefficients are provided throughout the rest of the table in regular print. MM = music: math; MRC = music: reading comprehension; SM = silence: math; SRC = silence: reading comprehension; O-Span = operation span; Mod. Lag. = modified lag task; LNS = letter number sequencing. Correlations > .2 are significant at p < .05.
listening habits and preferences while completing homework and studying. Results Scoring WMC was treated as a continuous variable and was based on a composite score created from three separate measures of WMC. Performance on each WMC task was standardized and then averaged together to create a measure of WMC. This approach allowed for a measurement of the effect of individual differences in WMC across a continuous range of WMC scores. The degree to which participants were accurate in predicting their performance on a given task (e.g., proportion correct on math questions during the playing of music) was calculated with a calibration score. Correlations between measures, as well as reliabilities for all measures, are presented in Table 1. Music Effect and Working Memory For all analyses the significance level was set at .05. A 2 (Task: math vs. reading comprehension) × 2 (Auditory condition: music vs. silence) × 2 (Order: music condition first vs. silence condition first) mixed-model Analysis of Covariance (ANCOVA) was used with performance on the WMC tasks included as a covariate to predict accuracy (proportion correct) on the academically relevant tasks. There was no main effect of order nor did it interact with any variables (all F’s < 1.1). A main effect of auditory condition was observed, F(1, 136) = 12.38, p < .001, MSE = 0.29, η2 = .03. Furthermore, there was a main effect of WMC, F(1, 136) = 25.73, p < .001, MSE = 1.37, η2 = .03. However, the interaction between auditory condition and WMC was not statistically significant, F(1, 136) = 2.74, p = .101, MSE = 2.74, η2 < .01. The main effect of task on performance was also not statistically significant, F(1, 136) = 1.23, p = .269, MSE = 0.03, η2 < .01. Moreover, there was no significant interaction between task and auditory condition, or task and WMC (both F’s < 2). Importantly, there was a qualifying three-way interaction between auditory condition, task, and WMC, F(1, 136) = 4.00, p = .048, MSE = 0.08, η2 = .01. Although performance was worse in the presence of music in both the reading-comprehension (M = .38, SD = .18) and math (M = .38,
SD = .19) tasks relative to when the reading-comprehension (M = .41, SD = .16) and math (M = .44, SD = .20) tasks were performed in silence, the degree to which WMC impacted the effect of auditory condition needed to be evaluated in each task separately (performance across tasks for music: M = .38, SD = .15 and silence: M = .43, SD = .15). Two separate repeated measures ANCOVAs were conducted for the reading-comprehension and math tasks to follow-up this three-way interaction. In the following ANCOVAs auditory condition varied within participants, and performance on the WMC tasks was included as a covariate. The first analysis revealed that the difference in performance on the reading-comprehension task between the music and silence conditions was less pronounced for participants with higher WMC scores, F(1, 136) = 6.72, p = .011, MSE = 0.14, η2 = .05. A separate ANCOVA revealed that this was not the case on the math task, F(1, 136) = 0.04, p = .836, MSE < 0.01, η2 < .01. Thus, the observed three-way interaction was driven by WMC composite scores moderating the music distraction effect for the reading-comprehension but not the math task (Figure 1). Metacognitive Knowledge Paired-samples t-tests were used to compare participants’ predictions for their performance across conditions, and none of these comparisons approached statistical significance. On both tasks, participants’ predictions of how well they would perform, and the degree to which they were accurate in those predictions, did not vary between the music and silence conditions (all t’s < 1.3). Furthermore, the accuracy of performance predictions across academic tasks was not significantly correlated with performance on WMC tasks during the playing of music, r(136) = −.02, p = .815 or during silence, r(136) = .03, p = .726. The self-reported percentage of time that participants listen to music while doing academically relevant tasks (M = 56%, SD = 34%), such as homework, was not significantly related to any potentially relevant factors such as the effect of music on the math, r(136) = −.03, p = .726 or reading-comprehension, r(136) = .10, p = .242 tasks, or WMC, r(136) = −.01, p = .916. Similarly, the degree to which participants believed that music affected them was not significantly correlated to the effect that music had on their math, r(136) = −.07, p = .413 or readingcomprehension, r(136) = .01, p = .907 performance.
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012
+Model
ARTICLE IN PRESS WORKING MEMORY, MUSIC, AND STUDENT PERFORMANCE
5
Figure 1. The moderating role of WMC in the effect of music on performance. For graphical purposes, the effect of music was calculated by subtracting the proportion correct during silence from the proportion correct during music on a given task. Therefore, negative scores represent lower accuracy during the playing of music.
Discussion The present study replicated previous research demonstrating that performance of academically relevant tasks suffered under a music condition compared to a silence condition (e.g., Anderson & Fuller, 2010; Henderson et al., 1945). Importantly, the present findings offered an exciting extension to this line of research by providing a novel explanation for why some have failed to observe the music distraction effect, as individual differences in WMC had not previously been considered as a moderator. It may be that in previous studies where an effect of music was not observed, a sufficiently large proportion of the sample had high WMC, which would have hindered any attempt to observe the typically detrimental effect of music. The implication of the present study is that, indeed, some students are correct in their assumption that it is relatively safe for them to listen to music while doing their homework. Specifically, our data suggest that individuals who score higher on WMC tests perform reading-comprehension tasks just as well when they are listening to music relative to performing the task in silence. However, individual differences in WMC did not provide a buffer against music-related distraction in math tasks. The finding that individuals with higher WMC scores could safely perform reading-comprehension tasks while listening to music dovetails with previous observations of higher-WMC individuals better employing executive attention skills (Conway et al., 2005; Kane et al., 2001). We had predicted that high WMC would similarly facilitate performance on arithmetic tasks, regardless of the auditory condition. The failure to observe this finding could reflect the fact that arithmetic tasks are supported by different cognitive processes than reading-comprehension tasks. For example, reading-comprehension tasks involve encoding information into long-term memory (Atkinson & Shiffrin, 1968), a process not necessarily inherent in arithmetic tasks. Notably, all of the WMC tasks used in the present study were verbal in nature, which could have diminished their predictive utility for the math task. Consistent with this premise, children with specific arithmetic deficits have produced significantly lower scores on the spatial, but not the verbal, component of WMC relative to children with no specific arithmetic deficits (McLean
& Hitch, 1999). One explanation then, is that higher WMC will most effectively moderate a distracting environmental stimulus (e.g., music) when there is a match between the modality of the distractor and the WMC system being measured. However, there is also research suggesting that WMC is a unitary construct, and that separate stores for verbal and visuo-spatial information are not necessary (Cowan & Morey, 2007; Kane et al., 2004). Thus, a fruitful direction for future research would be to investigate whether individual differences in spatial WMC could, indeed, moderate the music-distraction effect that has been observed in math tasks. Contrary to what Anderson and Fuller (2010) found, our results suggest that although some individuals can do their homework and listen to music at the same time, knowledge of one’s own ability may be tenuous. The current study had an older sample than did Anderson and Fuller (2010) who used junior high school students. It may be that by the time students reach college they believe they have developed sufficient strategies for working on academically relevant tasks while listening to music. Furthermore, the way in which beliefs regarding the effect of music were measured in the present study varied from Anderson and Fuller (2010) who questioned their participants regarding whether or not they listened to music while doing homework using a 5-point Likert scale. Conversely, participants in the present study were asked to report their music listening habits as a percentage, which allowed for a more exact representation of predicted behavior. It is also worth noting that WMC scores in the present study were surprisingly not associated with participants’ ability to accurately predict their performance on the arithmetic or reading-comprehension tasks across auditory conditions. Given that the presence of music was associated with a drop in performance, a lack of awareness of the detrimental effect of music likely represented a deficiency in metacognition. Greater metacognitive abilities should lead individuals to detect and avoid the effect of music whenever possible. Typically, increased WMC has been associated with better metacognitive abilities (Dunlosky & Kane, 2007; Thomas, Bonura, Taylor, & Brunyé, 2012), but in the present study the way in which music impacted performance represented a blind-spot for participants across the
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012
+Model
ARTICLE IN PRESS WORKING MEMORY, MUSIC, AND STUDENT PERFORMANCE
spectrum of WMC. However, it is possible that participants in the present study listened to music while working on academic material outside of the lab for reasons other than poor metacognition. In particular, it may be that many students in a university setting face a limited availability of quiet places to do homework and study. The participants in this study may have believed (and perhaps correctly) that music is less detrimental than the potentially chaotic noises that they encounter in their campus environment. Future research should investigate the degree to which students believe music to be preferable to silence, as opposed to other noisy conditions. In conclusion, for most students it would be a mistake to listen to music while working on academically relevant tasks. However, higher-WMC individuals may be less vulnerable to performance deficits traditionally associated with listening to music. Unfortunately, no evidence was revealed to suggest that these individuals know who they are, so many students may listen to music despite the detrimental effects this choice has on their learning. The findings from the present study also have important theoretical implications as they provide one potential explanation for why inconsistencies have been observed in the musicdistraction literature. Future research is needed to elucidate which other academically relevant tasks are affected by music, and to what extent individual difference factors other than WMC contribute to one’s susceptibility to music-related distractions. Author Contributions Eddie A. Christopher conceived of the project, and both authors contributed equally to the planning of the experiment. Eddie A. Christopher was primarily responsible for data collection and analysis. Eddie A. Christopher was primarily responsible for writing the manuscript. Conflict of Interest Statement The authors declare no conflict of interest. Acknowledgements This manuscript served as the Master’s thesis for Eddie A. Christopher. Portions of these data were presented at the annual meeting of the Psychonomic Society held in Chicago, IL (November, 2015). We would like to thank Iain Scott and Emily Van Zandbergen for their assistance with data collection. Furthermore, we would like to thank Dr. Mike Biderman and Dr. Amanda Clark for their valuable contributions as committee members. References Abikoff, H., Courtney, M. E., Szeibel, P. J., & Koplewicz, H. S. (1996). The effects of auditory stimulation on the arithmetic performance of children with ADHD and nondisabled children. Journal of Learning Disabilities, 29(3), 238–246. http://dx.doi.org/10.1177/002221949602900302 Anderson, S. A., & Fuller, G. B. (2010). Effect of music on reading comprehension of junior high school students. School Psychology Quarterly, 25(3), 178–187. http://dx.doi.org/10.1037/a0021213
6
Atkinson, R. C., & Shiffrin, R. M. (1968). Human memory: A proposed system and its control processes. In The psychology of learning and motivation. Oxford, England: Academic Press. http://dx.doi.org/10.1016/S007967421(08)6042263 Conway, A. A., Cowan, N., & Bunting, M. F. (2001). The cocktail party phenomenon revisited: The importance of working memory capacity. Psychonomic Bulletin & Review, 8(2), 331–335. http://dx.doi.org/10.3758/BF03196169 Conway, A. A., Kane, M. J., Bunting, M. F., Hambrick, D., Wilhelm, O., & Engle, R. W. (2005). Working memory span tasks: A methodological review and user’s guide. Psychonomic Bulletin & Review, 12(5), 769–786. http://dx.doi.org/10.3758/BF03196772 Cowan, N., & Morey, C. C. (2007). How can dual-task working memory retention limits be investigated? Psychological Science, 18(8), 686–688. http://dx.doi.org/10.1111/j.1467-9280.2007.01960.x Doyle, M., & Furnham, A. (2012). The distracting effects of music on the cognitive test performance of creative and noncreative individuals. Thinking Skills and Creativity, 7(1), 1–7. http://dx.doi.org/10.1016/j.tsc.2011.09.002 Dunlosky, J., & Kane, M. J. (2007). The contributions of strategy use to working memory span: A comparison of strategy assessment methods. The Quarterly Journal of Experimental Psychology, 60, 1227–1245. http://dx.doi.org/10.1080/17470210600926075 Engle, R. W. (2002). Working memory capacity as executive attention. Current Directions in Psychological Science, 11(1), 19–23. http://dx.doi.org/10.1111/1467-8721.00160 Flavell, J. H. (1979). Metacognition and cognitive monitoring: A new area of cognitive–developmental inquiry. American Psychologist, 34(10), 906–911. http://dx.doi.org/10.1037/0003-066X.34.10.906 Furnham, A., & Bradley, A. (1997). Music while you work: The differential distraction of background music on the cognitive test performance of introverts and extroverts. Applied Cognitive Psychology, 11(5), 445–455. http://dx.doi.org/10.1002/(SICI)10990720(199710)11:5<445::AID-ACP472>3.0.CO;2-R Freeburne, C. M., & Fleischer, M. S. (1952). The effect of music distraction upon reading rate and comprehension. Journal of Educational Psychology, 43(2), 101–109. http://dx.doi.org/10.1037/h0054219 Henderson, M. T., Crews, A. A., & Barlow, J. J. (1945). A study of the effect of music distraction on reading efficiency. Journal of Applied Psychology, 29(4), 313–317. http://dx.doi.org/10.1037/h0056128 Judde, S., & Rickard, N. (2010). The effect of post- learning presentation of music on long-term word- list retention. Neurobiology of Learning and Memory, 94(1), 13–20. http://dx.doi.org/10.1016/ j.nlm.2010.03.002 Kane, M. J., Bleckley, M., Conway, A. A., & Engle, R. W. (2001). A controlled-attention view of working- memory capacity. Journal of Experimental Psychology: General, 130(2), 169–183. http://dx.doi.org/10.1037/0096-3445.130.2.169 Kane, M. J., Hambrick, D. Z., Tuholski, S. W., Wilhelm, O., Payne, T. W., & Engle, R. W. (2004). The generality of working memory capacity: A latent-variable approach to verbal and visuospatial memory span and reasoning. Journal of Experimental Psychology: General, 133(2), 189–217. http://dx.doi.org/10.1037/ 0096-3445.133.2.189 Kantner, J. (2009). Studying with music: Is the irrelevant speech effect relevant? In M. R. Kelley (Ed.), Applied memory (pp. 19–40). Hauppauge, NY, US: Nova Science Publishers. Mayfield, C., & Moss, S. (1989). Effect of music tempo on task performance. Psychological Reports, 65(3, Pt 2), 1283–1290. http://dx.doi.org/10.2466/pr0.1989.65.3f.1283 McLean, J. F., & Hitch, G. J. (1999). Working memory impairments in children with specific arithmetic learning difficulties.
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012
+Model
ARTICLE IN PRESS WORKING MEMORY, MUSIC, AND STUDENT PERFORMANCE
Journal of Experimental Child Psychology, 74(3), 240–260. http://dx.doi.org/10.1006/jecp.1999.2516 Mowsesian, R., & Heyer, M. R. (1973). The effect of music as a distraction on test-taking performance. Measurement & Evaluation in Guidance, 6(2), 104–110. Pacheco-Unguetti, A. P., & Parmentier, F. R. (2014). Sadness increases distraction by auditory deviant stimuli. Emotion, 203–213. http://dx.doi.org/10.1037/a0034289 Perham, N., & Currie, H. (2014). Does listening to preferred music improve reading comprehension performance? Applied Cognitive Psychology, 28(2), 279–284. http://dx.doi.org/10.1002/acp.2994 Pool, M. M., Koolstra, C. M., & Van Der Voort, T. A. (2003). Distraction effects of background soap operas on homework performance: An experimental study enriched with observational data. Educational Psychology, 23(4), 361–380. http://dx.doi.org/ 10.1080/01443410303211 Princeton Review. (2008). 11 practice tests for the SAT and PSAT. Natick, MA: Random House. Robison, M. K., & Unsworth, N. (2015). Working memory capacity offers resistance to mind-wandering and external distraction in a context-specific manner. Applied Cognitive Psychology, 680–690. http://dx.doi.org/10.1002/acp.3150 Shelton, J. T., Elliott, E. M., Hill, B. D., Calamia, M. R., & Gouvier, W. D. (2009). A comparison of laboratory and clinical working memory tests and their prediction of fluid intelligence. Intelligence, 37(3), 283–293. http://dx.doi.org/10.1016/j.intell.2008.11.005
7
Shelton, J. T., Metzger, R. L., & Elliott, E. M. (2007). A group-administered lag task as a measure of working memory. Behavior Research Methods, 39(3), 482–493. http://dx.doi.org/ 10.3758/BF03193017 Thomas, A. K., Bonura, B. M., Taylor, H. A., & Brunyé, T. T. (2012). Metacognitive monitoring in visuospatial working memory. Psychology and Aging, 27(4), 1099–1110. http://dx.doi.org/ 10.1037/a0028556 Tucker, A., & Bushman, B. J. (1991). Effects of rock and roll music on mathematical, verbal, and reading comprehension performance. Perceptual and Motor Skills, 72(3, Pt 1), 942. http://dx.doi.org/10.2466/PMS.72.3.942-942 Unsworth, N., Heitz, R. P., Schrock, J. C., & Engle, R. W. (2005). An automated version of the operation span task. Behavior Research Methods, 37, 498–505. http://dx.doi.org/10.3758/BF03192720 Woo, E. W., & Kanachi, M. (2005). The effects of music type and volume on short-term memory. Tohoku Psychologica Folia, 64, 68–76.
Received 8 July 2016; accepted 30 January 2017 Available online xxx
Please cite this article in press as: Christopher, E. A., & Shelton, J.T. Individual Differences in Working Memory Predict the Effect of Music on Student Performance. Journal of Applied Research in Memory and Cognition (2017), http://dx.doi.org/10.1016/j.jarmac.2017.01.012