The Effect of Music Listening on the
Cognitive Performance of Older Adults:
Does Familiarity Matter?
Research Report
MSc in Brain and Cognitive Sciences, Track Cognitive Science,
University of Amsterdam
Student: Jeannette van Ditzhuijzen (11116587) Number of ECTS: 26 ECTS
Time period: 18/01/2016 - 08/07/2016 Supervisor: Dr. Rebecca Schaefer
Co-assessor: Dr. Wery van den Wildenberg
UvA Representative: Dr. Wery van den Wildenberg
Research Institute: Institute for Psychology, Health, Medical & Neuropsychology Unit, Universiteit Leiden
The Effect of Music Listening on the Cognitive
Performance of Older Adults: Does Familiarity Matter?
ABSTRACT
Controversial evidence suggests that exposure to music may transiently increase cognitive performance in healthy older adults, possibly by impacting mood and arousal levels, as stated in the so-called arousal-mood hypothesis. In the current study, it is hypothesized that arousal-mood and arousal levels are influenced by the degree of familiarity with the music played, where greater familiarity is expected to lead to higher arousal levels, subsequently resulting in better cognitive performance relative to music with low familiarity. The impact of different types of music on cognitive performance was examined using cognitive tasks focused on working memory, memory search and visuospatial skills. Fifty-six healthy older adults (µ = 71.16 years) completed the Forward Digit Span, Letter Fluency and Mental Rotation task after they listened to a musical piece with differing degrees of familiarity between subjects. The obtained results indicated, surprisingly, that familiarity with the music stimulus negatively influenced both working memory and visuospatial performance slightly, when taking baseline performance into account as well as the subjective induced arousal response and the rating of importance of music in life. These findings do not confirm the predictions made on the basis of the arousal-mood hypothesis, therefore more research is recommended into the mechanism behind familiar and unfamiliar music listening and its effect on cognitive functioning, whilst comparing different age groups.
INTRODUCTION
Music, something so trivial and ordinary to everyone, has seen a surge of interest in the recent decade of research. Not only the effects of music on well-being (e.g., quality of life) are of interest (MacDonald, 2013), but also potential therapeutic effects on cognitive functioning, such as working memory, are attracting attention (Thompson, Moulin, Hayre, & Jones, 2005). A large amount of literature in the neuropsychology of music originally stems from the so-called Mozart
Effect, where better spatial reasoning skills
were reported in subjects after listening to Mozart’s Piano Sonata K 448 (Rauscher, Shaw, & Ky, 1993). The authors explain the effect by suggesting that exposure to musical compositions, that are structurally complex, excite certain cortical firing patters comparable to those activated when completing spatial-temporal tasks (Cassity, Henley, & Markley, 2007; Rauscher, Shaw, & Ky, 1995). However, failures to replicate the basic effect raised doubt about its reliability (Steele, Ball, & Runk, 1997; Steele, Bass, & Crook, 1999). Therefore, an alternative hypothesis was developed; the arousal-mood
hypothesis (Husain, Thompson, & Schellenberg, 2002), which states that improvement in cognitive abilities is caused by the impact of music on mood and arousal (Chabris, 1999; Schellenberg, 2014;
Thompson, Schellenberg, & Husain, 2001). Specifically, preferred stimuli, thus, not necessarily music (Nantais & Schellenberg, 1999), are believed to induce positive affect and heightened levels of arousal, resulting in better cognitive performance (Bottiroli, Rosi, Russo, Vecchi, & Cavallini, 2014; Cassity et al., 2007; Schellenberg & Weiss, 2013; Thompson et al., 2001). However, the potential role that subjects’ familiarity with the music may play in the emotional response, has not been a specific topic of investigation so far.
The effect of music-induced arousal and mood on cognitive performance has been studied mainly in younger subjects, using visuo-spatial tasks (e.g. Cassity et al., 2007; Nantais & Schellenberg, 1999; Rauscher et al., 1993; Thompson et al., 2001). Researchers in the same field investigated effects of music on other aspects of cognition, such as verbal episodic memory, working memory and processing speed (Ferreri et al., 2015; Schellenberg, Nakata, Hunter, & Tamoto, 2007). A smaller amount of research has investigated the effect of music listening on the cognitive performance of healthy ageing older adults (Bottiroli et al., 2014; Ferreri et al., 2014; Mammarella, Fairfield, & Cornoldi, 2007; Reaves, Graham, Grahn, Rabannifard, & Duarte, 2015; Thompson et al., 2005). Here, a beneficial effect of classical music listening was found for
3
working memory (Mammarella et al., 2007), processing speed and declarative memory (Bottiroli et al., 2014), and category fluency (Thompson et al., 2005). In addition, suggestive evidence exists for an improvement in episodic memory performance while upbeat acoustic jazz music is played in the background (Ferreri et al., 2014).
However, the aforementioned studies suffer from varied methodological weaknesses, by ignoring possible effects of arousal content of the music, music preference and subject’s familiarity with the music stimulus. In addition, the role that someone assigns to music in his or her life seem to influence the music-induced emotional response as well, where a greater importance of music in one’s life seems to result in a higher emotional arousal in response to music listening (Gingras, Marin, Puig-Waldmüller, & Fitch, 2015). Moreover, music has also been found to impair cognitive function in older adults when being presented alongside an associative memory task (Reaves et al., 2015), suggesting that music listening requires cognitive resources as well, disabling older adults to inhibit task-irrelevant information. These contrasting results indicate that more research is needed to be able to come to robust conclusions regarding the effect of music listening on specific cognitive tasks, and the mediating role of the music-induced emotional response. Therefore in the current study, it is investigated whether various types of music have differing effects on arousal and mood levels, which subsequently may influence cognitive performance on different tasks. Performance on several cognitive tasks is taken into account, based on previous research regarding working memory and memory search (Mammarella et al., 2007), and visuospatial skills (Rauscher et al., 1993). In addition, the current study attempts to contribute to the discussion by taking the methodological weaknesses of other studies into account, regarding music preference, the arousal content of the stimulus, and especially familiarity with the music that is listened to.
According to Bharucha (1987), there are at least two processes which lie at the basis of an emotional response to music, namely
anticipation and resolution. Subsequently, these two processes are the result of structural knowledge and veridical knowledge (Bharucha, 1987). Structural knowledge consists of the implicit understanding of rules and regularities of music, due to mere exposure to music of a particular genre (Tillmann, 2005). Veridical knowledge on the other hand, refers to explicit knowledge on the following sequence of events within a musical piece (Bharucha, 1987). Consequently, either type, or a combination of the two lead to the creation of expectancies, which in turn may result in an emotional response via confirmation or violation of these expectations (Meyer, 1956; van den Bosch, Salimpoor, & Zatorre, 2013). Following from this, a familiar music song may lead to an increased emotional response relative to novel music, since the amount of veridical knowledge and related prediction generation will be higher. Pereira, Teixeira, Figueiredo, Xavier, Castro and Brattico (2011) reported a crucial role of familiarity with music-induced emotional engagement, where functional magnetic resonance imaging (fMRI) data showed that broad emotion-related limbic and paralimbic regions, as well as other reward circuitry, were significantly more active for familiar music than with unfamiliar music (Pereira et al., 2011). In addition, most emotion-related brain activity was triggered by familiar, either liked or disliked, rather than liked (familiar or unfamiliar) music (Pereira et al., 2011). In other words, the level of emotional engagement seems to be particularly influenced by familiarity, rather than preference, in contrast to the initial Mozart effect interpretation. Moreover, van den Bosch et al. (2013) report that familiarity mediates emotional arousal related to music listening, thus, only when a subject was exposed to a musical piece before, he or she may experience heightened arousal levels related to music listening.
In line with these findings, veridical knowledge of a specific music excerpt may contribute most to generated expectations, which in turn may influence emotional arousal. Results of previously mentioned studies regarding improvement in working memory in older adults after music listening (e.g. Mammarella et al., 2007), can thus
4
possibly be explained by the effect of familiarity on arousal, since all subjects were very familiar with the particular music piece (Mammarella et al., 2007). Although the mood-arousal hypothesis involves the effect of preferred stimuli, the current study examines the potential role of familiarity with the stimulus as aforementioned research have shown that familiarity seems to be an important aspect of the music-induced emotional response. In addition, it is well-known that liking for an unfamiliar musical item typically increases as a function of exposure (Peretz, Gaudreau, & Bonnel, 1998), raising the possibility that familiarity plays a potential role within the arousal-mood hypothesis.
In summary, based on previous results claiming cognitive improvement after music listening, the current research examines the effect of familiar music listening on working memory, memory search and visuospatial ability in a sample of healthy older adults, taking into account effects of music preference, emotional responses and arousal (subjectively reported). The role of familiarity is the focus of this study, as only little attention has been paid to this factor related to the effect of music listening on cognitive performance. It is hypothesized that familiarity of the music stimulus will lead to an increase in positive mood and heightened arousal response, which in turn may impact working memory, memory search and visuospatial skills. This hypothesis is based on van den Bosch et al.’s (2013) finding that listening to novel pieces may result in positive mood, though, without impacting arousal levels. Familiar music stimuli, on the other hand, are related to strong positive mood, as well as increased arousal levels (Pereira et al., 2011; van den Bosch et al., 2013). The combination of a positive mood and heightened arousal levels, in turn may positively influence cognitive performance, according to the arousal-mood hypothesis (Thompson et al., 2001).
If music can create a situation that potentially optimizes cognitive performance in older adults, this may have important practical implications for successful and healthy ageing. Moreover, positive findings regarding music familiarity will have implications concerning the development of music
interventions for patients with Alzheimer disease, since musical memory, thus music familiarity, is surprisingly preserved in the brain structure of these patients (Jacobsen, Fritz, Stelzer, & Turner, 2015).
METHOD
PARTICIPANTS
A final sample of fifty-four older adults between the age of 65 and 90 years (M = 71.04 years, SD = 4.24) were included in the study, after excluding two subjects due to medicinal issues or highly disruptive circumstances during the experiment. The participants were 40 females and 14 males. All subjects lived in the Netherlands, spoke Dutch and reported to be healthy and free from any psychiatric or neurological disorder. Exclusion criteria included the presence of a pacemaker or a former profession in the music industry (e.g. professional artist, music teacher, conductor). In addition, participants required a minimal score of 25 (Vertesi et al., 2001) on the Mini-Mental State Examination (MMSE) (M = 29.26, SD = .96). History of musical training and current regular musical activities were assessed, but were not used as exclusion criteria. Two participants were excluded from the analyses because the session got disrupted Prior to the beginning of the experiment, participants completed and accepted an informed consent form. At completion of the experiment, subjects received €10 as compensation for their time. Ethical consent was provided by the Ethics Committee of the Leiden University (CEP16-0303/109).
STIMULI
Ten, predicted to be highly familiar, music songs from the time period 1960-1980, were selected on the basis of Dutch pop charts (Veronica Top 40), playlists (e.g. Spotify), and music literature (Garofalo, 2010; Grijp, 2001). In addition ten, predicted to be unfamiliar, music songs were carefully chosen, originating from the same time period and from the same range of musical genres. Half of the stimulus set consisted of high arousal content, whereas the other ten
5
stimuli comprised low arousal content, as assessed through mutual agreement between three researchers. This yielded four conditions, namely every cross-factorial combination of high and low arousal content and high and low familiarity of the music stimulus (i.e. high arousal-high familiarity, high arousal-low familiarity, low arousal-high familiarity and low arousal-low familiarity), all consisting of five music stimuli. All stimuli
were preselected to span a range of genres, express positive valence affect, had a length between 2 and a half and 8 minutes, and sometimes included lyrics. These differences were kept intact, favoring a natural listening experience over constructed stimuli. The volume was adjusted for every individual participant. See Table 1 for an overview of the preselected stimuli and how often they were used.
Table 1 – Overview of the preselected stimuli ordered by condition, including the name, artist or
composer, length of each song and the amount of times the stimulus was used
High familiarity – High arousal content Length (min) Times played
Bij ons in de Jordaan - Johnny Jordaan 3.29 1
Carmen Overture - Georges Bizet 3.28 2
(I Can’t Get No) Satisfaction - Rolling Stones 3.41 3
Think - Aretha Franklin 3.13 4
The Four Seasons: Spring - Antonio Vivaldi 6.02 3
High familiarity – Low arousal content
Air on G string - Johann Sebastian Bach 5.38 6
Imagine - John Lennon 3.03 4
Laat me alleen - Rita Hovink 4.44 1
Summertime - Ella Fitzgerald & Louis Armstrong 4.55 3
Unforgettable - Nat King Cole 3.10 -
Low familiarity – High arousal content
Boot um - Pro McClam & The Spaniels 2.23 5
Florentiner March - Julius Fucik 5.22 1
I’m a Man - Chicago 7.40 3
Symphony No.1 - 3. Scherzo, Allegro non troppo - Vasily Kalinnikov 8.12 3
Strand - Boudewijn de Groot 2.30 1
Low familiarity – Low arousal content
Sheep May Safely Graze - Leopold Stokowski 5.34 3
Creepin’ - Stevie Wonder 4.21 3
Idle Moments - Grant Green 6.55 4
Love me as I love you - Dave (Baby) Cortez 2.43 3
6
MEASUREMENTS
Cognitive performance
Cognitive performance was assessed using the Shepard-Metzler mental rotation task (Shepard & Metzler, 1971) in line with the original findings on visuospatial performance related to music (Rauscher et al., 1993). However, the parameter of reaction time was not taken into account given the age of the subjects, and only the total number of correct answers was used. In addition, in order to gain a more detailed assessment of cognitive performance, working memory was measured using forward Digit Span (WAIS IV) as well as memory search, via the Letter Fluency task (WAIS IV), following the study of Mammarella et al. (2007). In the following, these performance measures are referred to as MRT, DS and LF, respectively.
Mood, arousal, familiarity & preference
Emotional content that was perceived in as well as induced by the music was assessed using sliders together with the Self-Assessment Manikin (Bradley & Lang, 1994) as it includes independent assessments of arousal and valence in a nonverbal way, and can be applied to perceived and induced emotion. Familiarity, liking and the importance of music in their lives were measured using a slider as well, to produce a continuous scale with ratings from 0-100.
Heart rate & heart rate variability
Heart rate and heart rate variability (HR/HRV) were measured using a portable pulse oximeter (model PO 80, Beurer, Germany), attached to the non-dominant index finger, producing a photoplethysmogram (PPG) with a sample frequency of 60 Hz. Accompanying software, the ‘SpO2-Viewer’, recorded the PPG on a laptop which was connected to the pulse oximeter. Notably, these data were not included in the current report.
APPARATUS
The technical set-up consisted of two laptops. The presentation of task instructions and stimuli, as well as the recording of demographic information and all behavioral
responses, were running on one laptop, controlled by E-Prime software (Psychology Software Tools, Inc.). Meanwhile, another laptop was connected to the pulse oximeter and recorded changes in blood volume via the software ‘SpO2-Viewer’. Questionnaires regarding musical background, plus the MMSE, DS and LF were assessed by the experimenter using pencil and paper. Speakers (model A60, Creative Labs, Singapore) were used for the presentation of the music stimuli.
EXPERIMENTAL PROCEDURE
The experimental study took place in two sessions each lasting 30-45 minutes, at the participant’s home. In the first session, participants completed, in order, the informed consent form, questions concerning musical background, and the MMSE. When inclusion criteria were met, subjects completed some demographic questions and subsequently performed the three cognitive tests in randomized order, to measure baseline performance. Afterwards, a listening task was presented, where the participant listened to 30-seconds clips of five pre-selected stimuli that matched the condition the participant was assigned to. Here, participants were counterbalanced to match for gender over the different conditions before the first session. Ratings for all five excerpts were presented and collected regarding familiarity, liking, and induced emotional response in terms of arousal and valence. Based upon these ratings, the stimulus for the second session was selected to maximally fit the specific condition requirements regarding familiarity and induced arousal levels, meanwhile, keeping liking maximal and valence positive. Minimally one week, and maximally two weeks later, the second session took place, to avoid recognition of the excerpts classified as ‘unfamiliar’ and to avoid learning effects on the cognitive tasks. First, the participant listened to the full piece of music that was found to best fit the assigned condition for that specific participant, followed by the second cognitive assessment (MRT, DS, LF in randomized order). Next, ratings of familiarity, liking, and emotional response in terms of arousal and valence were assessed once more regarding the played music piece.
7
At the end of the second session, participants were asked to rate the role of music in their lives (Gingras et al., 2015). Both sessions included measurement of PPG, both starting with a 2-minute silent relaxation period to collect baseline HR/HRV, to account for inter-individual differences in physiological activity. Although these measurements are not part of the current report, the findings can thus be extended with a measure of physiological arousal at a later time.
DATA ANALYSIS
First, experimental manipulations of familiarity, emotional response and preference of the music stimulus were inspected by comparing the mean rating scores between the different groups using t-tests and analysis of variance (ANOVA). Afterwards, multiple bivariate and partial correlations were performed to investigate the hypothesis regarding the relationship between familiarity and emotional response. In addition, the relationships between liking, familiarity and the emotional response were explored with these analyses. Next, for every dependent variable of interest (DS, LF and MRT), a hierarchical multiple regression was performed using the second cognitive assessment score as dependent variable. Independent variables included the baseline performance score for the particular cognitive task and familiarity rating. Induced arousal
and the role of music in life were added to the analyses as covariates to address findings from previous literature. Local effect sizes of familiarity were calculated using Cohen’s ƒ2 as described in Selya, Rose, Dierker, Hedeker, & Mermelstein (2012). Prior to the hierarchical multiple regression analyses, outliers were removed on the basis of the frequently used Cook’s distance, applying the conventional cut-off score 4/n (0.0741) (Stevens, 2009). For all three predictive models four data points were removed, resulting in n’s of 50.
RESULTS
The mean score regarding familiarity with the music stimulus, together with the subjective induced arousal rating are presented in Figure 1. Here, two conditions are shown indicating low and high familiarity with the music, therefore, not taking arousal content of the stimuli into account. When performing an independent-samples t-test, a significant difference between the two groups in low familiarity (n = 27, M = 45.07, SD = 34.34) and high familiarity rating (n = 27, M = 85.07, SD = 25.50, t(52) = -4.86, p <.01) was observed, even though the mean familiarity score for the low familiarity group was higher than expected. No significant difference between these two groups in subjective rating of induced arousal was shown (M = 42.59, SD = 29.69, M = 49.63, SD = 37.12, t(52) = -.77, p
Table 2 – Bivariate and partial correlations
between familiarity, preference, arousal and emotional response Measure 1 2 3 1. Familiarity - 2. Liking .33* - 3. Induced arousal .33* -.12 - 4. Induced valence .30* .87** -.11 Partial correlations controlling for preference, arousal and emotional response
2. Liking .21 -
3. Induced arousal .39** -.14 - 4. Induced valence .001 .85** .004 *p < .05. **p < .01.
Figure 1 - Mean scores of familiarity and arousal
ratings, ordered by familiarity group. The error bars represent standard deviations
0 20 40 60 80 100 120
High familiarity Low familiarity
M EAN RAT ING SCO RE
8
= .45). However, a significant correlation between familiarity and induced arousal was observed (r(52) = .33, p = .02), implying that the degree of familiarity with the presented music influences subjective arousal levels by heightening them.
Familiarity also positively correlated with the preference score (r(52) = .33, p = .01), suggesting that preference for a musical piece increases with previous knowledge of the stimulus. Consequently, the rating scale regarding preference score was significantly higher on average for the high familiar group (M = 85.48, SD = 14.32) relative to low familiar music (M = 64.78, SD = 24.25, t(52) = -3.82, p < .01). In addition, the induced emotion ratings suggest that familiar music induces relatively more positive affect (M = 85.48, SD = 14.60) than less familiar music (M = 69.04, SD = 25.50, t(52) = -2.91, p = .01). A positive significant correlation between familiarity and induced emotion (r(52) = .30, p = .03) indicates this as well. A summary of the bivariate and partial correlations between familiarity and the other factors can be found in Table 2. When considering the partial correlations, it should be noted that familiarity only retained a significant correlation with induced arousal, when statistically controlling for preference and valence response (r(50) = .39, p < .01). Again, suggesting that the level of familiarity with the music heightens arousal levels. In addition, a highly positive significant bivariate correlation and partial correlation was observed between the variables of preference and induced valence, when
controlling for induced arousal and familiarity, indicating that listening to preferred music stimuli resulted in a stronger positive mood. When examining the four groups, therefore taking arousal content of the music into account as well (i.e. high arousal-high familiarity (n = 13), high arousal-low familiarity (n = 13), low arousal-high familiarity (n = 14) and low arousal-low familiarity (n = 14)), a significant difference in familiarity rating (F(3, 50) = 8.92, p < .01) and induced arousal rating (F(3, 50) = 13.73, p < .01) regarding the music stimulus was found, between the four groups, indicating that the requirements regarding familiarity and arousal ratings were relatively met between the four groups, as can be seen in Figure 2, although the low familiarity-high arousal group only fits this manipulation relative to the other groups.
However, both the familiarity ratings as well the induced arousal scores of the different groups are observed to have large standard deviations, indicating a wide spread of rating scores. Together with the correlations that were detected between familiarity and emotional responses, this has led to the decision to use the familiarity rating score as a continuous variable in the regression analyses, instead of using the groups to represent different familiarity levels in the analyses. The induced subjective arousal variable was also used as a continuous variable.
Figure 2 - Mean scores of familiarity and arousal ratings, ordered by familiarity-arousal group. The
error bars represent standard deviations.
0 10 20 30 40 50 60 70 80 90 100 High familiarity High arousal High familiarity Low arousal Low familiarity High arousal Low familiarity Low arousal M EAN RAT ING SCO RE
9
COGNITIVE PERFORMANCE
Digit Span
To investigate whether familiarity with the music is playing a facilitating role within DS performance, a hierarchical multiple regression was performed, using the baseline DS score as a control variable as well as subjective induced arousal and importance of music in life variables. Table 3 presents the regression coefficients of the performed hierarchical multiple regression. A significant model (R2 = .63, F
(2, 47)= 39.87, p < .01) was observed (Model 2 in Table 3). The data was checked and met the assumptions regarding
normality, multicollinearity, linearity and homoscedasticity. The model suggested, surprisingly, a significant negative effect of familiarity on DS assessment after music listening (β= -.21, p = .03). In addition, when the two control variables of induced arousal levels and importance of music were statistically controlled for (Model 4 in Table 3, R2 = .66, F
(4, 45)= 22.14, p < .01), the negative influence of familiarity on DS remained significant (β = -.25, p = .01). This suggests that performance on DS decreases when the familiarity with the music played increases. No significant effects of the control variables induced arousal (β = .17, p =.07) and
Table 4 - Summary of Hierarchical Regression Analysis for variables predicting Letter Fluency performance
after music listening (N = 50)
Model 1 Model 2 Model 3 Model 4 Variable B SE B β B SE B β B SE B β B SE B β LF baseline .70 .07 .82** .70 .07 .81** .74 .07 .86** .73 .07 .85** Familiarity -.02 .02 -.07 -.01 .02 -.02 -.01 .02 -.02 Arousal -.05 .03 -.17 -.05 .03 -.16 Importance .07 .06 .10 R2 .66 .67 .69 .70 F for change in R2 94.94** .70 3.30 1.62 *p < .05. **p < .01.
Table 5 - Summary of Hierarchical Regression Analysis for variables predicting Mental Rotation Task
performance after music listening (N = 50)
Model 1 Model 2 Model 3 Model 4 Variable B SE B β B SE B β B SE B β B SE B β MRT baseline .13 .13 .14 .10 .13 .11 .09 .13 .10 .03 .14 .04 Familiarity -.02 .01 -.31* -.03 .01 -.35* -.02 .01 -.34* Arousal .01 .01 .13 .01 .01 .12 Importance -.03 .02 -.17 R2 .02 .11 .13 .15 F for change in R2 .92 4.95* .75 1.34 *p < .05. **p < .01.
Table 3 - Summary of Hierarchical Regression Analysis for variables predicting Digit Span performance after
music listening (N = 50)
Model 1 Model 2 Model 3 Model 4 Variable B SE B β B SE B β B SE B β B SE B β DS baseline .76 .09 .77** .80 .09 .81** .77 .09 .78** .78 .09 .79** Familiarity -.01 .004 -.20* -.01 .01 -.25** -.01 .01 -.25* Arousal .01 .01 .18 .01 .01 .17 Importance -.01 .01 -.07 R2 .59 .63 .66 .66 F for change in R2 68.93** 5.03* 3.86 .68 *p < .05. **p < .01.
10
importance (β = -.07, p = .41) were found on DS performance. The significant effect of familiarity is also being confirmed when examining the R2 change of the four models. Only a significant increase in explained variance is found when familiarity is added to the model (R2 = .63, F
(1, 47) = 5.03, p = .03) in comparison to the addition of the two control variables of arousal and importance, as can be seen in Table 3. According to Cohen’s ƒ2
, familiarity had a medium local effect size in all three models (ƒ2Model 2 = .11, ƒ2Model 3= .18, ƒ2 Model 4= .15).
Letter Fluency
A hierarchical multiple regression was performed to examine whether familiarity with the music is playing a facilitating role within LF performance, as well, using the baseline LF performance as a control variable just as subjective induced arousal and the role of music in life variable. The results of the regression analysis can be found in Table 4. A significant model (R2 = .70, F
(4, 45)= 26.50, p < .01) was observed (Model 4 in Table 4), which included both familiarity as the control variables and resulted in the biggest proportion of explained variance. The data was checked and met assumptions regarding normality, multicollinearity, linearity and homoscedasticity. However, no significant impact of familiarity on the second LF assessment was observed (β = -.02, p = .79), when induced arousal levels and importance of music were statistically controlled. Also, the model suggested no significant effect of the control variables induced arousal (β = -.16, p = .09) and importance (β = .10, p = .21). In addition, little to no local effect size (ƒ2
Model 2 = .03, ƒ2Model 3 = .00, ƒ2 Model 4 = .00) was observed for familiarity.
Mental Rotation task
To examine whether familiarity with the music is playing a possible facilitating role within MRT performance, a hierarchical multiple regression was performed again, using the baseline MRT performance as a control variable as well as subjective induced arousal and importance of music in life variable. Table 5 presents the regression coefficients of this analysis. Here, only one significant model (R2 = .11, F
(2, 47)= 2.98, p = .05) was
observed (Model 2 in Table 5). The data was
checked and met assumptions regarding normality, multicollinearity, linearity and homoscedasticity. The model suggested, remarkably, a significant negative influence of familiarity on the MRT assessment after listening to music (β = -.31, p = .01). Meaning that when familiarity with the music increased, performance on the MRT decreases. The significant effect of familiarity is also being confirmed when examining the explained varianceof the four models. Only a significant increase in explained variance is namely found when familiarity is added to the model (R2 = .11, F
(1, 47) = 4.95, p = .03) in comparison to the addition of the two control variables of arousal and importance, as presented in Table 5. According to Cohen’s ƒ2
, familiarity had a small to medium local effect size on MRT performance in the three models (ƒ2Model 2 = .10, ƒ2Model 3= .13, ƒ2 Model 4= .12).
DISCUSSION
The present study examined the impact of different types of music on tasks indicating working memory (DS), memory search (LF) and visuospatial skills (MRT), in a sample of healthy older adults. From the obtained results it appeared that performance in visuospatial abilities, as well as working memory, was significantly influenced by the level of familiarity with the music played. In contrast to the proposed hypothesis, this impact was negative for both tasks, meaning that when familiarity with the heard music increased, performance on the Mental Rotation Task and Digit Span decreased, when the influence of induced arousal levels and the role of music in life were statistically controlled for. Subjects’ performance on the memory search on the other hand, was not influenced by the level of familiarity.
Possibly, an alternative explanation for the observed, negative effect of familiarity involves the degree of distraction that music listening may contribute to, by claiming attentional resources in older adults (Reaves et al., 2015). Perhaps, less familiar music may relatively enhance concentration by demanding somewhat less attention than highly known music. However, this distinction is probably highly person-specific and this
11
argument should therefore be read with caution. Nevertheless, Avila, Furnham, & McClelland (2011), and Etaugh & Michals, (1975) have suggested that younger adults’ verbal task performance (Avila et al., 2011) and reading comprehension (Etaugh & Michals, 1975) worsened whilst listening to familiar music relative to quiet surroundings, by an increase in distraction. Or perhaps, the explanation lies within the complexity of the music played, as music with a low information load, thus being highly repetitive and within a narrow tonal range, seems to benefit cognitive performance more than music with a high information load, which is more dissonant, rhythmically varied and highly dynamic (Kiger, 1989; Avila, 2011). Plausibly the preselected familiar pieces in this study were relatively higher in complexity, or differed in other musical aspects such as rhythm or pitch, relative to the unfamiliar songs, which implicitly may have influenced cognitive performance. However, this seems somewhat unlikely, since the stimuli were carefully chosen and generally derived from the same musical genres. In addition, it should be noted that these aforementioned studies played music during the task, whilst the current study played the music before task performance. Moreover, no comparison was made between familiar and unfamiliar music in these studies, but rather the influence of music was tested, which is not the case in the current study. To our knowledge, the differing effects of familiar and unfamiliar music on cognitive performance have not previously been explicitly investigated. Therefore, future studies into the concept of music exposure and its effect on working memory and visuospatial skills are highly recommended, whilst systematically comparing outcomes regarding music listening before or during task execution. This may shed more light on the specific issue.
As mentioned, the obtained results contradict the proposed expectation of the study regarding the facilitating role of familiarity with music on cognitive performance. Nevertheless, when considering the subjective ratings regarding the induced emotional arousal response, the hypotheses regarding the relationships between familiarity, mood and arousal, based on
Pereira et al. (2011) and van den Bosch et al. (2013), are observed to some extent. Music stimuli with a greater degree of familiarity namely did lead to relatively higher subjective ratings of induced arousal and positive affect. However, when considering the partial correlations, an interesting dissociation between the presumably related concepts of preference and familiarity can be observed. Liking especially impacts mood positively, whereas familiarity influences subjective arousal levels, indicating that the degree of familiarity with the music played does influence subjective arousal levels by heightening them, as suggested by van den Bosch et al. (2013). However, indirectly, this does not confirm the predictions made by the arousal-mood hypothesis (Husain et al., 2002; Thompson et al., 2001), since the arousal levels plausibly induced by familiarity, did not positively influence cognitive performance, but even resulted in a contradicting outcome regarding working memory and visuospatial performance. Possibly the objective assessments of physiological arousal, namely HR and HRV, which were measured in the current study but not used for the analyses, can shed more light on the exact relationships between familiarity, emotional arousal and cognitive performance.
Thus, no support for the arousal-mood hypothesis has been found, also when taking merely the obtained results regarding the subjective induced arousal levels into consideration. No influence of different arousal levels was observed on all tasks performances, even though a large stimulus set was used in this study to induce widely ranging levels of arousal between subjects. Therefore, the hypothesis that music listening results in improved performance only if it increases positive mood and arousal levels, is not supported by the current data.
An alternative explanation for the observed positive effect of music on working memory, memory search and visuospatial skills, as suggested by Mammarella et al., (2007) and Rauscher et al. (1993), could involve the claim that music listening results in a reduction of anxiety (Västfjäll, 2012; Lee et al., 2012; Sung & Chang, 2005), which in turn may positively influence cognitive
12
performance in older adults. It is also plausible that different age-related mechanisms are at work since it seems that supportive evidence for the arousal-mood hypothesis has mainly been observed in younger subjects (e.g. Husain et al., 2002; Thompson et al., 2001). Other research into healthy older adults, using a mood or arousal measure, found no direct evidence for arousal-mood hypothesis as well regarding free recall, letter fluency (Bottiroli et al., 2014) and working memory performance (Hirokawa, 2004). Future research is highly recommended where the mechanisms underlying music listening and its effect on cognitive performance are investigated, whilst comparing the different age groups. The question why only working memory and visuospatial skills benefitted from less familiar music listening relative to familiar music, and memory search did not, is intriguing. Since both tasks showed similar results regarding familiarity of the stimulus, this suggests that verbal working memory and visuospatial processes may function on a mere domain-general construct. However this is not in accordance with findings of Hale et al. (2011) which suggests that these cognitive processes operate on distinct mechanisms. A systematic exploration of the effect of familiar and unfamiliar music listening on the performance of several tasks indicating the cognitive processes of working memory and visuospatial ability, may shed more light on this issue.
When solely considering working memory, the obtained results somewhat contradict the popular, but controversial, idea of a plausible positive relationship between familiar music and memory enhancement, as studied in Alzheimer’s disease (Baird & Samson, 2009; El Haj, Fasotti, & Allain, 2012; Jacobsen et al., 2015; Simmons-Stern et al., 2012). Thompson et al. (2005) argue that music may have a specific effect on memory performance in healthy older adults also, by holding specific personal memories for an individual, through which memory recall may enhance. However, this relation would not account for any effect of unfamiliar music. Our findings highlight the possibility that unfamiliar music may facilitate working memory performance even more, though, as
long as it is music which is preferred by the listener and induces positive affect. Whether music listening actually has a beneficial effect on working memory, whilst taking into account familiarity, is worth studying more thoroughly, since working memory seems to have more profound age-related decline than other cognitive domains as executive functioning and information processing speed (Charlton, Schiavone, Barrick, Morris, & Markus, 2010).
Whether music has a facilitating effect on visuospatial abilities has also been studied in older adults with mild cognitive impairment (Cacciafesta, Ettorre, Amici, Cicconetti, Martinelli, Linguanti, Barrata, Verrusio, & Marigliano, 2010), and in patients with Alzheimer’s disease (Johnson, Shaw, Vuong, Vuong, & Cotman, 2002), where the results suggest that these skills can improve directly after listening to Mozart’s piano sonata for both clinical groups. Cacciafesta et al. (2010) discuss why Mozart’s composition K448, and other compositions with the same index of periodicity, may particularly result in visuospatial improvement, thus, arguing that this potential music effect only occurs with particular types of compositions, undervaluing the inter-individual differences in a musical experience, as argued in the current study. Ultimately, evidence that particular music improves visuospatial abilities can have important implications for rehabilitation in clinical practice, as for the concept of healthy ageing, since it seems that white matter integrity reduces with age which predicts a decline in visuospatial functions (Antonenko & Flöel, 2014).
Limitations of the current study should be noted. By using this particular study design, the influence of music itself on cognitive performance could not be examined. This leaves the question open whether there was actual improvement in cognitive performance within subjects, after listening to less familiar music. A repeated-measures ANOVA design, whilst counterbalancing silent and music conditions differing in familiarity over subjects, would be advisable to investigate the existence of any effect of music, whilst taking into account the possible mediating role of familiarity. In addition, the choice to take individual differences into account
13
regarding aspects as preference and familiarity, led to twenty different music stimuli being included. Therefore, not every participant listened to the same music piece, disabling the investigation concerning the effects of certain musical aspects such as rhythm, melody and dynamics on performance. Finally, by presenting the listening task at the end of the first session through which the music stimulus was chosen for the second session, participants were exposed to a small clip of the music piece already, influencing the degree of familiarity implicitly. To make more conclusive statements about the specific effect of familiarity follow-up studies are recommended where subjects are exposed to entirely novel music stimuli, compared to well-known musical pieces.
In conclusion, the current study examined whether familiarity is playing a beneficial role in the effect of music listening on cognitive performance in healthy older adults. Contrary to the proposed predictions, familiarity with the music seemed to deteriorate both working memory and visuospatial skills. Performance on memory search on the other hand did not seem to be affected by the degree of familiarity, indicating that the influence of familiarity is not homogenous. The effect of music on cognitive performance is a complex issue given the many intervening variables and individual differences in the experience of music listening. However, from the obtained results it seems that the degree of familiarity has a particular role in its outcome, in contrast to music-induced arousal levels, of which no impact was observed, thus not confirming the mood-arousal hypothesis. Finally, more research is recommended with the focus on the particular mechanism behind familiar and unfamiliar music listening and its possible effect on cognitive performance, whilst comparing different age groups.
REFERENCES
Appelhans, B. M., & Luecken, L. J. (2006). Heart rate variability as an index of regulated emotional responding. Review of General
Psychology, 10(3), 229–240.
Antonenko, D., & Flöel, A. (2014). Healthy Aging
by Staying Selectively Connected: A Mini-Review. Gerontology, 60, 3–9.
http://doi.org/10.1159/000354376
Avila, C., Furnham, A., & McClelland, A. (2011). The influence of distracting familiar vocal music on cognitive performance of introverts and extraverts. Psychology of
Music, 40(1), 84–93.
http://doi.org/10.1177/0305735611422672 Baird, A., & Samson, S. (2009). Memory for
music in Alzheimer’s disease:
Unforgettable? Neuropsychology Review,
19(1), 85–101.
http://doi.org/10.1007/s11065-009-9085-2 Bharucha, J. J. (1987). Music Cognition and
Perceptual Faciliation: A Connectionist Framework. Music Perception, 5(1), 1–30. Bottiroli, S., Rosi, A., Russo, R., Vecchi, T., &
Cavallini, E. (2014). The cognitive effects of listening to background music on older adults: processing speed improves with upbeat music, while memory seems to benefit from both upbeat and downbeat music. Frontiers in Aging Neuroscience,
6(10), 1–7.
http://doi.org/10.3389/fnagi.2014.00284 Bradley, M., & Lang, P. J. (1994). Measuring
Emotion: The Self-Assessment Manikin and the Semantic Differential. Journal of
Behavior Therapy and Experimental
Psychiatry, 25(1), 49–59.
http://doi.org/10.1016/0005-7916(94)90063-9
Cacciafesta, M., Ettorre, E., Amici, A., Cicconetti, P., Martinelli, V., Linguanti, A., Barrata, A., Verrusio, W. & Marigliano, V. (2010). New frontiers of cognitive rehabilitation in geriatric age: the Mozart Effect (ME). Archives of Gerontology and
Geriatrics, 51(3), 79-82.
Cassity, H. D., Henley, T. B., & Markley, R. P. (2007). The Mozart Effect: Musical Phenomenon or Musical Preference? A More Ecologically Valid Reconsideration.
Journal of Instructional Psychology, 34(1),
13–18. Retrieved from
http://www.eric.ed.gov/ERICWebPortal/reco rdDetail?accno=EJ761251
Chabris, C. F. (1999). Prelude or requiem for the “Mozart effect”? Nature, 400, 826–827. http://doi.org/10.1038/23608
Charlton, R. a, Schiavone, F., Barrick, T. R., Morris, R. G., & Markus, H. S. (2010). Diffusion tensor imaging detects age
14 related white matter change over a 2 year follow-up which is associated with working memory decline. Journal of Neurology,
Neurosurgery, and Psychiatry, 81(1), 13–
19. http://doi.org/10.1136/jnnp.2008.167288 El Haj, M., Fasotti, L., & Allain, P. (2012). The
involuntary nature of music-evoked autobiographical memories in Alzheimer’s disease. Consciousness and Cognition,
21(1), 238–246.
http://doi.org/10.1016/j.concog.2011.12.005 Etaugh, C., & Michals, D. (1975). Effects on
reading comprehension of preferred music and frequency of studying to music.
Perceptual and Motor Skills, 41, 553–554.
Ferreri, L., Bigand, E., & Bugaiska, A. (2015). The positive effect of music on source memory. Musicae Scientiae, 19(4), 402– 411.
http://doi.org/10.1177/1029864915604684 Ferreri, L., Bigand, E., Perrey, S., Muthalib, M.,
Bard, P., & Bugaiska, A. (2014). Less Effort, Better Results: How Does Music Act on Prefrontal Cortex in Older Adults during Verbal Encoding? An fNIRS Study.
Frontiers in Human Neuroscience, 8(5), 1–
11.
http://doi.org/10.3389/fnhum.2014.00301 Garofalo, R. (2010). Rockin'out: Popular Music in
the U.S.A. (5th ed). Massachusetts:
Pearson.
Gingras, B., Marin, M. M., Puig-Waldmüller, E., & Fitch, W. T. (2015). The Eye is Listening: Music-Induced Arousal and Individual Differences Predict Pupillary Responses.
Frontiers in Human Neuroscience, 9(11),
619.
http://doi.org/10.3389/fnhum.2015.00619 Grijp, L.P. (2001). Een muziekgeschiedenis der
Nederlanden. Amsterdam: Amsterdam
University Press-Salomé.
Hale, S., Rose, N. S., Myerson, J., Strube, M. J., Sommers, M., Tye-Murray, N., & Spehar, B. (2011). The structure of working memory abilities across the adult life span.
Psychology and Aging, 26(1), 92–110.
http://doi.org/10.1037/a0021483
Hirokawa, E. (2004). Effects of Music Listening and Relaxation Instructions on Arousal Changes and the Working Memory Task in Older Adults. Journal of Music Therapy,
41(2), 107–127.
Husain, G., Thompson, W. F., & Schellenberg, E.
G. (2002). Effects of Musical Tempo and Mode on Arousal, Mood, and Spatial Abilities. Music Perception: An
Interdisciplinary Journal, 20(2), 151–171.
Jacobsen, J., Fritz, T., Stelzer, J., & Turner, R. (2015). Why musical memory can be preserved in advanced Alzheimer’s disease. Brain, 138, 2438–2450. http://doi.org/10.1093/awv148
Johnson, J. K., Shaw, G. L., Vuong, M., Vuong, S., & Cotman, C. W. (2002). Short-Term Improvement on a Visual-Spatial Task After Music Listening in Alzheimer’s Disease.
Activities, Adaptation & Aging, 26(3), 37– 50. http://doi.org/10.1300/J016v26n03 Kiger, D. M. (1989). Effects of music information
load on a reading comprehension task.
Perceptual and Motor Skills, 69, 531–534.
Lee, K.-C., Chao, Y.-H., Yiin, J.-J., Hsieh, H.-Y., Dai, W.-J., & Chao, Y.-F. (2012). Evidence That Music Listening Reduces Preoperative Patients’ Anxiety. Biological Research For
Nursing, 14(1), 78–84.
http://doi.org/10.1177/1099800410396704 MacDonald, R. A. R. (2013). Music, health, and well-being: A review. International Journal
of Qualitative Studies on Health and Well-Being, 8, 1–14.
Mammarella, N., Fairfield, B., & Cornoldi, C. (2007). Does music enhance cognitive performance in healthy older adults? The Vivaldi effect. Aging Clinical and
Experimental Research, 19(5), 394–399.
http://doi.org/10.1007/BF03324720 Meyer, L. B. (1956). Emotion and Meaning in
Music. Chicago, IL: University of Chicago
Press.
Nantais, K. M., & Schellenberg, E. G. (1999). The Mozart Effect: An Artifact of
Preference. Psychological Science, 10(4), 370–373. http://doi.org/10.1111/1467-9280.00170
Pereira, C. S., Teixeira, J., Figueiredo, P., Xavier, J., Castro, S. L., & Brattico, E. (2011). Music and Emotions in the Brain: Familiarity Matters. PLoS ONE, 6(11), 1–9. http://doi.org/10.1371/journal.pone.0027241 Peretz, I., Gaudreau, D., & Bonnel, A.-M. (1998).
Exposure effects on music preference and recognition. Memory & Cognition, 26(5), 884–902.
http://doi.org/10.3758/BF03201171
15 Music and spatial task performance.
Nature, 365, 611.
Rauscher, F. H., Shaw, G. L., & Ky, K. N. (1995). Listening to Mozart enhances spatial-temporal reasoning: towards a
neurophysiological basis. Neuroscience
Letters, 185(1), 44–47.
http://doi.org/10.1016/0304-3940(94)11221-4
Reaves, S., Graham, B., Grahn, J., Rabannifard, P., & Duarte, A. (2015). Turn Off the Music! Music Impairs Visual Associative Memory Performance in Older Adults. The
Gerontologist, 00(00), 1–10.
http://doi.org/10.1093/geront/gnu113 Schellenberg, E. G. (2014). Mozart Effect. In W.
F. Thompson (Ed.), Music in the Social and
Behavioral Sciences: An Encyclopedia (pp.
717–718). Sage.
Schellenberg, E. G., Nakata, T., Hunter, P. G., & Tamoto, S. (2007). Exposure to music and cognitive performance: tests of children and adults. Psychology of Music, 35(1), 5–19. http://doi.org/10.1177/0305735607068885 Schellenberg, E. G., & W. Weiss, M. (2013).
Music and Cognitive Abilities. In The
Psychology of Music (Third Edition) (pp.
499–550). http://doi.org/10.1016/B978-0-12-381460-9.00012-2
Schellenberg, E. G. (2012). Cognitive performance after listening to music: A review of the Mozart effect. In R.
MacDonald, G. Kreutz, & L. Mitchell (Eds.),
Music, health, and wellbeing (324–338).
Oxford, UK: Oxford University Press. Selya, A. S., Rose, J. S., Dierker, L. C., Hedeker,
D., & Mermelstein, R. J. (2012). A practical guide to calculating Cohen’s f2, a measure of local effect size, from PROC
MIXED. Frontiers in Psychology, 3, 1-6. Simmons-Stern, N. R., Deason, R. G., Brandler,
B. J., Frustace, B. S., O’Connor, M. K., Ally, B. A., & Budson, A. E. (2012). Music-based memory enhancement in Alzheimer’s Disease: Promise and limitations.
Neuropsychologia, 50(14), 3295–3303.
http://doi.org/10.1016/j.neuropsychologia.2 012.09.019
Steele, K. M., Ball, T. N., & Runk, R. (1997). Listening to Mozart does not improve Backwards Digit Span Performance.
Perceptual and Motor Skills, 84, 1179–
1184.
Steele, K. M., Bass, K. E., & Crook, M. D. (1999). The mystery of the Mozart Effect: Failure to Replicate. Psychological Science, 10(4), 366–369.
Stevens, J.P. (2009). Applied multivariate
statistics for the social sciences (5th ed).
New York: Routledge.
Sung, H. C., & Chang, A. M. (2005). Use of preferred music to decrease agitated behaviours in older people with dementia: A review of the literature. Journal of Clinical
Nursing, 14(9), 1133–1140.
http://doi.org/10.1111/j.1365-2702.2005.01218.x
Thompson, R. G., Moulin, C. J. a, Hayre, S., & Jones, R. W. (2005). Music Enhances Category Fluency in Healthy Older Adults and Alzheimer’s Disease Patients.
Experimental Aging Research, 31(1), 91–
100.
http://doi.org/10.1080/03610730590882819 Thompson, W. F., Schellenberg, E. G., & Husain,
G. (2001). Arousal, Mood, and The Mozart Effect. Psychological Science, 12(3), 248– 251.
http://doi.org/10.1111/1467-9280.00345
Tillmann, B. (2005). Implicit investigations of tonal knowledge in nonmusician listeners.
Annals of the New York Academy of
Sciences, 1060, 100–110.
http://doi.org/10.1196/annals.1360.007 van den Bosch, I., Salimpoor, V. N., & Zatorre,
R. J. (2013). Familiarity mediates the relationship between emotional arousal and pleasure during music listening. Frontiers in
Human Neuroscience, 7(9), 1–10.
http://doi.org/10.3389/fnhum.2013.00534 Vertesi, A., Lever, J. A., Molloy, D. W.,
Sanderson, B., Tuttle, I., Pokoradi, L., & Principi, E. (2001). Standardized Mini-Mental State Examination. Use and interpretation. Canadian Family Physician.
Le Médecin de Famille Canadien, 47,
