567933

research-article2015
POM0010.1177/0305735614567933Psychology of MusicOdendaal

Article

Psychology of Music

(Mis)matching perceptual
2016, Vol. 44(3) 353­–368
© The Author(s) 2015
Reprints and permissions:
learning styles and practicing sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0305735614567933
behavior in tertiary level pom.sagepub.com

Western Classical
instrumentalists

Albi Odendaal

Abstract
Perceptual learning style theory argues that humans have differing perceptual strengths and that
instruction and learning materials should be adapted to fit their visual, auditory and/or kinesthetic
preferences. The claims of various theorists that perceptual learning style theory also applies to
the learning of musical material when practicing were tested using a questionnaire that draws on
the formulations of these theorists. In total, 131 Western Classical instrument students in higher
music education were surveyed using a paper-based questionnaire regarding the frequency with
which they engaged in particular learning strategies and behaviors while practicing a large scale
composition from the Classical or Romantic era. No response patterns comparable to those suggested
by perceptual learning style theory could be identified through either principal components or cluster
analyses. The applicability of the theory to describe individual differences in learning is questioned.

Keywords
individual differences, instrumental learning and teaching, multisensory integration, perception, practice

There is considerable popular interest in the identification of learners’ preferences for learning
and teaching material in the visual, auditory or kinesthetic modalities. This interest is mirrored
in music education circles; for instance, perceptual learning style theory (PLS)1 has been advo-
cated as an instructional option in music teaching textbooks (e.g., Campbell, Scott-Kassner, &
Kassner, 2006; Jacobson, 2006), articles in music teaching periodicals (e.g., Garcia, 2002;

Department of Music Education, Jazz and Folk Music, Sibelius Academy of the University of the Arts, Finland
South African College of Music, University of Cape Town, South Africa

Corresponding author:
Albi Odendaal, Department of Music Education, Jazz and Folk Music, Sibelius Academy of the University of the Arts,
Finland; South African College of Music, University of Cape Town , Private Bag, Rondebosch 7701, South Africa.
Email: albi.odendaal@uct.ac.za
354 Psychology of Music 44(3)

Swanson, 2005), and articles in academic journals (e.g., Beheshti, 2009; R. E. Dunn, 2008).
This support exists despite long-standing research-based opposition to the theory (Krätzig &
Arbuthnott, 2006; Pashler, McDaniel, Rohrer, & Bjork, 2009; Sharpley 1984, 1987). The
claims of PLS have far-reaching pedagogical implications, and are usually accompanied by
guarantees of successful learning or teaching, should the theory be conscientiously imple-
mented. These are important claims to investigate.
PLS was probably derived from an aspect of neuro-linguistic programming (NLP) (see
Fleming, 2006), rather than from the theories of multiple intelligences (Gardner, 1985; as
claimed by Behesthi, 2009) or representational theory (Bruner, 1964, 1966; as claimed by
Gault, 2005). Neither Gardner nor Bruner classified individuals according to modality strengths
nor did they insist on modality specific instruction. In the seminal work of NLP, however,
Grinder and Bandler (1976) maintained that individuals receive information and mentally rep-
resent the world around them through three major ‘input channels’ (visual, auditory, and kin-
esthetic). They noted that while none of the channels are better than the others, a person might
have one or more dominant channels, and might alternate between channels. The preferred
representational system (PRS) in individuals can be identified through their use of language,
through eye movements, and through self-report. In the therapy situation, the authors main-
tained that therapists should match the PRS of their clients in order to communicate more
effectively. This is almost identical to the main formulations of PLS, except that PLS is typically
applied to learning or teaching situations with the dictum that teachers should attempt to pre-
sent material in the strongest modality of a learner (Barbe & Swassing, 1979; R. Dunn & Griggs,
2003; Fleming, 2006; Hannaford, 1995).
Several theoretical perspectives on how PLS should be applied to the music studio or class-
room have been published in academic and trade journals (Beheshti, 2009; Everett, 1997;
Garcia, 2002; Gault, 2005; Miller, 2002; Swanson, 2005). Despite this, no consensus exists on
how the modalities, and behaviors linked to them, should be defined. This can also be said for the
main formulations of the theory outside of music education (compare, for example, Fleming,
2006, with R. Dunn & Griggs, 2003). For the purpose of this investigation, the various formula-
tions were compared and the most common shared aspects taken to define each of the modali-
ties. Thus, visual was taken to pertain to any information that is seen, including scores, markings,
instruments, hands, or movements (of self or others), together with mental visualizations of
these things; auditory was taken to pertain to any sonic material, whether made by oneself
(including vocalizations or self-talk), made by someone else (including musical sounds or verbal
instructions), or imagined; kinesthetic was taken to pertain to any aspect of movement, includ-
ing imagined movement, involved in the learning situation. The kinesthetic modality is typically
used as a combination term for the proprioceptive, tactioceptive and kinesthetic senses.
The majority of empirical research studies on PLS in musical learning investigated the phe-
nomenon in young children. R. E. Dunn (2008) found that presentation with either visual,
kinesthetic or no reinforcement influenced the music listening behaviors of primary school
children. The children’s attention to the music was aided, distracted or not affected by these
reinforcements in a clear pattern over repeated sessions. Persellin and Pierce (1988) found that
primary-school children remembered a short rhythm better when it was presented to them in
their preferred modality, either aurally, through flashing light or by tapping on their shoulder.
Zikmund and Nierman (1992) presented primary school children with melodic conservation-
type exercises and found that modality-specific reinforcement had a positive effect on children’s
performance. In a year-long grounded-theory study of pre-school children learning to play vio-
lin in a group, Calissendorff (2006, p. 89) found that learning was ‘highly individual, varying
from child to child’. One of the areas of variance was that of perception, with children
Odendaal 355

responding differently to the same stimulus, and finding individual strategies to learn to play,
either preferring touch or sight or sound as a primary sense.
When investigating the relationship between music memorization strategy and PLS in col-
lege level musicians, Mishra (2007) found only weak correlations between learning style pref-
erence and memorization strategy as defined by two self-report questionnaires, with only
self-identified visual learners seeming to prefer visual memorization strategies. Korenman and
Peynircioglu (2007), however, when investigating the effect of presentation mode (visual or
auditory) and learning style preference on the memorization of musical and verbal material in
an experimental setting, found that learning style had a significant effect on memorization of
short sentences and musical phrases for college level students.
While there is empirical support for PLS in some aspects of musical learning, especially in
younger children, the research also leaves several questions unaddressed. Little research has
been conducted focusing on the practicing of complete works, and there is also little research
on adults. The definitions of appropriate modality-specific behaviors given by authors who
apply the theory to musical learning have not been rigorously tested. For this reason, a mixed-
methods study of the relationship between PLS and musical practicing and learning in music
students in higher music education, involving both questionnaire and observation data, was
launched. This article reports on the questionnaire data; complementary data can be found in
Odendaal (2013, 2014).

Aim of the study

The aim of the study was to identify whether there are patterns in the ways that Western clas-
sical music instrument students in higher education describe their practicing behaviors and
strategies for learning a new work from sheet music, and to relate these (possible) patterns to
PLS. For this reason, descriptions of practicing behaviors and strategies were taken from litera-
ture on PLS and practicing. It was hypothesized, based on readings of PLS theory in musical
learning, that a visual learner would indicate using more visual strategies, an auditory learner
would indicate using more auditory strategies and a kinesthetic learner would indicate using
more kinesthetic strategies, and that they would do so even when not prompted about PLS or its
claims.

Method
Materials
The questionnaire (see the Appendix) was constructed, primarily, by drawing on theoretical
descriptions of the practicing strategies and behaviors of ideal visual, auditory and kinesthetic
music learners (Beheshti, 2009; Garcia, 2002; Mixon, 2004; Swanson, 2005). However, sec-
ondarily, published lists or examples of specific strategies and learning behaviors that musi-
cians use were scanned for identifiable aspects of perceptual modality use that conforms to the
theoretical descriptions given by the authors listed above (among others: Hallam, 2001a;
Holmes, 2005; Hultberg, 2008; Miklaszewski, 1995; Nielsen, 2008; Sullivan & Cantwell,
1999). Additionally, some items from Mishra’s (2007) Musical Memorisation Inventory, which
attempts to identify modality use in memorization, were adapted and included.
The questionnaire was refined through three pilot studies, undertaken with doctoral stu-
dents at the main research site (n = 11), with participants of an intensive international course
for masters students (n = 4), and with an opportunistic sample from an institute of higher
356 Psychology of Music 44(3)

music education in the same city as the main research site (n = 23) (Odendaal, 2011). This
process resulted in 47 items being included in the final version of the questionnaire. These
items were balanced in the ways that they seemed to focus on visual, auditory, or kinesthetic
responses, with 15 or 16 items in each category. A professional translation company translated
the questionnaire to the home language of the majority students at the research site, and this
version was used in the third pilot study and the final data collection.
Because PLS was assumed to exist independently of the respondents’ self-awareness of their
learning style, the purpose of the questionnaire was obscured. PLS was not mentioned on the
questionnaire, nor in introductory talks to taking the questionnaire. Nine distractor items
regarding organization of time and resources were added, giving the questionnaire the appear-
ance that it was about practicing in general rather than about PLS in order to limit confirma-
tion bias. The questionnaire was paper-based, and the 56 items were presented in randomized
order over two pages in order to further obscure the organization of the questions. On a third
page, some demographic information was also requested.
In the questionnaire, respondents were asked to imagine that they are starting to learn a new
large-scale work from the Classical or Romantic era. Respondents were asked to indicate the fre-
quency with which they would use a strategy or behavior using a 6-point rating scale (never, very
rarely, rarely, occasionally, frequently, always). The responses were scored between 1 for never and
6 for always by an independent coder, and coding was checked for consistency by the researcher.
Because the questionnaire does not posit clear dependent and independent variables, and
because the data were ordinal, analyses of variance and regressions were deemed inappropriate
analytical tools. Instead, because the theory assumed that respondents would answer according
to the patterns predicted by PLS, principal components and cluster analyses were employed to
explore these response patterns. While principal components analysis and factor analysis are
distinct approaches to data reduction, they often yield similar results, especially when there are
many variables involved (> 30) (Gorsuch, 1983, pp. 122–123). Principal components analysis
was selected because it ‘makes no assumptions about an underlying causal model’ (Hatcher,
1994, p. 10) and is thus useful as an exploratory method (Gorsuch, 1983, p. 124) – appropriate
in a case such as this where the questionnaire has not been previously tested on large samples.
Principal components provided a means of reducing the number of variables in order to explain
the variance in the data set, while remaining theoretically neutral.

Participants
Targeted samples within a large music university in northern Europe were employed in data
collection. Instrumental pedagogy classes, orchestral and ensemble rehearsals and an orienta-
tion day at the beginning of the academic year were used to present the questionnaire to groups
of students. In all, the questionnaire was presented to an estimated 390 students, but several of
these might have been duplicates as some students were involved in more than one of these
activities. Through this process, 165 responses were gathered, and a response rate of about
40% was thus achieved. Respondents who did not identify themselves as classical instrumen-
talists were removed from the response pool. This left 131 students from four departments, who
together represented 15% of the total number of students from those departments (see Table
1). A confidence interval of 7.9 at the 95% confidence level was calculated from this dataset.

Results
Alpha reliability coefficients were calculated from both Pearson’s correlation matrix
(Cronbach’s alpha) and the polychoric correlation matrix (Ordinal alpha) (Gadermann,
Odendaal 357

Table 1.  Number of responses to the questionnaire received and analysed, separated by department.

Department Students in Questionnaires Questionnaires

department received analysed
Church music 101 13 (12.9%) 13 (12.9%)
Classical 463 101 (21.8%) 101 (21.8%)
Composition and theory 56 3 (5.4%) 3 (5.4%)
Education 254 26 (10.2%) 14 (5.5%)
Folk music 85 6 (7.1%) 0
Jazz 55 4 (7.3%) 0
Music technology 36 4 (11.1%) 0
Singing 76 8 (10.5%) 0

Guhn, & Zumbo, 2012), and were deemed acceptable, with Cronbach’s alpha in the range of
.82–.84 and Ordinal alpha in the range of .84–.85 for each of the 47 main variables. No
items were thus excluded from analysis on the basis of reliability. The dataset was fairly com-
plete, with only a few missing variables, and thus listwise replacement of missing data was
used in all the analyses.
None of the items had normal distribution (Kolmogorov-Smirnov test of normality, df 115,
p < .001). Excluding outlying answers did not alter the non-normal distribution, which was
mainly related to the highly skewed nature of responses. Outliers should only be excluded on
firm theoretical grounds (Hill & Lewicki, 2007, p. 21), and PLS would predict a range of answers
on each item, thus this was not deemed a worthwhile exercise. While exploratory principal
components analysis assumes normal distribution, it does not depend on it, since it does not
calculate a p score.2 Although non-normal distribution affects generalizability negatively, pat-
terns that exist in the data can be explored regardless.

Principal component analysis

A principal components analysis was conducted on the 47 main variables. Components that
had an eigenvalue ⩾ 1.0 were considered, resulting in 29 components: too many to interpret
and with many cross loadings. A scree plot was calculated (Figure 1) and showed ‘elbows’ at the
3rd, 5th, and 9th components (Hill & Lewicki, 2007, p. 269). These were extracted, and orthog-
onal (Varimax) and oblique (Promax) rotations were also calculated for each.
The obliquely rotated 9-component solution provided the strongest component structure of
the options analysed, accounting for 50.49% of the variation in the questionnaire, and is the
solution reported in this article. The principal component structures of both the initial extrac-
tions and the rotated solutions in the other solutions were weak (with loadings below .7) and
accounted for considerably less variance than the 9-component solution. Considering only
component loadings ⩾ .5 in the obliquely rotated 9-component solution resulted in a clear and
interpretable component structure. This decision was confirmed when compared with compo-
nents of the orthogonal rotation, which reported the same contributing variables above .5 but
variations below .5 (see Gorsuch, 1983, p. 206).
Communalities averaged .51, which is moderate. The 9 components were not well defined,
with between two and four indicators per component. Accordingly, this study would probably
need between 100 and 200 respondents, but closer to 200 (MacCallum, Widaman, Preacher, &
Hong, 2001; MacCallum, Widaman, Zhang, & Hong, 1999). The slightly small sample size also
had a negative effect on generalizability.
358 Psychology of Music 44(3)

Figure 1.  A scree plot of the questionnaire data showing elbows at the 3rd, 5th and 9th components.

Components.  None of the components (see Table 2 for contributing variables and loadings for
each of the 9 components) displayed a single modality focus. Component 1 suggests that visual
imagery of the hands and fingers are integrally linked with the experience of movement, and
should thus not be understood as visual only. Component 2 includes some predominantly visual
activities, although comparing editions and analysis may involve some auditory imagery
together with visual activity. Component 3 includes each of the modalities. Component 4
includes a mixture of auditory and visual modalities. Component 5 involves each of the modali-
ties. Both the contributing variables of Component 6 involve listening, although there is also a
kinesthetic and possibly visual aspect to both. Component 7 includes each of the modalities.
Component 8 includes the auditory and kinesthetic modalities. Component 9 includes items
that all describe some form of vocalization, which is typically treated as an auditory behavior in
the literature, although it also has a kinesthetic aspect.

Relationship between instrument type and extracted components.  The variation found in, for exam-
ple, Component 1 could be attributable to differences in the instruments played, especially
when there is a difference between whether the hands and the instrument can be readily seen,
or are typically looked at while playing, and therefore whether they are likely to form an inte-
gral part of movement imagery. A Kruskal-Wallis test, with primary instrument sub-groups as
grouping variable, was used to investigate inter-instrumental variation. Three instrument sub-
groups were used: 1) piano or harpsichord, 2) violin or viola, and 3) woodwinds and brass (see
Table 3 for a distribution of instruments). While the two instruments paired in each of the first
two sub-groups have very similar posture and movement patterns, the third group is much
more disparate and results from this analysis should be treated with caution. Each of these sub-
groups had 30 or more members, although listwise replacement of missing variables reduced
this to n = 32 for sub-group 1, n = 26 for sub-group 2, and n = 23 for sub-group 3.
 Odendaal

Table 2.  Components listing the contributing variables over .5, each described with keywords.
1. Movement 2. Research (7.70% 3. Simplification 4. Reflection apart 5. Scouting 6. Musicality 7. Distraction 8. Metronome 9. Vocalization
imagery (12.32% of variance) (5.36% of from playing out (4.41% of (4.24% of variance) (4.07% of variance) use (3.96% of (3.49% of variance)
of variance) variance) (4.93% of variance) variance) variance)

K6 .75 V15 Reading .73 A6 Speak .83 V5 Cerebral .76 V10 Video .65 A10 Phrasing .75 V2 Distracted .75 A13 Use .78 A7 Sing .68
Movement information rhythm analysis and dynamics by vision metronome
imagery
V4 .60 V9 Compare .69 A15 .67 V14 Study .52 A4 Audio .63 A16 Aural .57 A12 .63 K3 .50 A9 Self-talk .52
Visualization editions Count out apart from imagery Distracted by Automation
loud instrument music aim
K4 .58 V13 Graphic .57 V8 Colors .55 A3 Aural .50 K10 Finding .59 K9 Technical −.57 K12 .56 A8 Hum .51
Movement representation imagery difficulties work Distracted by while playing
imagery aim environment
V11 .56 K1 Silent .51 V6 Practical −.72  
Visualization rehearsal analysis
aim
359
360 Psychology of Music 44(3)

Table 3.  The distribution of primary and secondary instruments of respondents.

Instrument Primary % Secondary %

instrument instrument
Piano/Harpsichord 34 26.0 48 36.6
Violin/Viola 30 22.9 7 5.3
Cello/Double Bass 14 10.7 1 0.8
Flute 8 6.1 0 0
Clarinet/Oboe/ Saxophone 7 5.3 1 0.8
Bassoon 5 3.8 0 0
Trumpet 2 1.5 0 0
Horn 6 4.6 0 0
Trombone 1 0.8 0 0
Percussion 2 1.5 1 0.8
Guitar/Bass 4 3.1 7 5.3
Kantele 4 3.1 0 0
Organ 9 6.9 1 0.8
Tuba 1 0.8 1 0.8
Accordion 3 2.3 1 0.8
Conducting 1 0.8 0 0
Voice 0 0 18 13.7
Drumkit 0 0 1 0.8
Total 131 100.0 87 66.4

Note. The Piano/Harpsichord and Violin/Viola grouping were big enough to use as sub-groups for comparative purposes,
and a third group was created using all woodwind and brass players.
The discrepancy between the numbers and the totals of the columns are due to the effect of rounding. The numbers
are all correct, as is the underlying mathematics.

In Component 1, there was a significant difference by instrument group for item K4 χ2(2, N
= 81) 8.532, p = .014 and item V11 χ2(2, N = 81) 6.532, p = .038. In Component 2, there was
a significant difference by instrument group for item V13 χ2(2, N = 81) 9.438, p = .009. In
Component 3, there was a significant difference by instrument group for item A6 χ2(2, N = 81)
9.209, p = .010. In Component 8, there was a significant difference by instrument group for
item A13 χ2(2, N = 81) 9.928, p = .007. Finally, in Component 9, there was a significant differ-
ence by instrument group for item A8 χ2(2, N = 81) 7.011, p = .030.
There was very little difference between the piano or harpsichord and violin or viola groups
and a significant difference between these two groups was found in only one item (A8). All
other significant differences occurred between one of these two and the composite woodwinds
and brass group. The differences may thus stem from the haphazard conflation of instruments
in the third group, rather than from any definable inter-group differences. The differences
occurred in five out of nine components, and affected only six out of the 29 items that describe
the most variation in the questionnaire. Interestingly, significant differences occurred not only
in those items where a physical difference between instruments would suppose it, such as in
V11, but also in those where physical differences between instruments would seem to make no
difference, such as V8, V13, or A13.

Higher order components.  Four higher order components (see Table 4 for contributing compo-
nents and loadings for each of the higher order components), accounting for 73.20% of
Odendaal 361

Table 4.  The contributing components and loadings of the four higher order components.

1 2 3 4  
1. Movement .76 2. Research .83 5. Getting to .90 3. Simplification .73
imagery know
4. Reflection apart .68 8. Metronome −.78 6. Musicality −.79 9. Vocalization −.75
from playing use
7. Distraction −.70  

variance between the lower-order components, were extracted (Gorsuch, 1983, p. 240). In the
first higher-order component, the contributing Components 1 and 4 both involved strategies
that take place when not playing, while Component 7 dealt with distraction from playing. This
higher-order component thus seemed to emphasize the tension between work done at the instru-
ment and work done away from the instrument, regardless of the modality of that work. The
second higher-order component described the tension between a more intellectual approach
involving research of various kinds, with a more physical and repetitive approach. Both the con-
tributing components to the third higher-order component seemed to have the early stages of
getting to know a work as a focus, but a slight difference was that Component 5 seemed to
approach this from a sonically external (use of recordings) and technical perspective, while
Component 6 approached it from a more internal sonic and non-technical perspective. The
fourth higher-order component had several vocalization items in both of the contributing com-
ponents. In Component 3, both the items were rhythmical vocalizations, whereas in Component
9, two of the items were melodic vocalizations.

Cluster analysis
Agglomerative hierarchical clustering with average linkage between groups (also known as
UPGMA) based on the Chi-square measure was used in this analysis (Norušis, 2011). No stand-
ardizing measure was used, since the variables were ordinal. The analysis made use of sets of
items that shared a common stem, but that were differentiated by modality focus. For instance,
V14, A16, and K1:

When I am struggling with a passage it helps when I

•• sit away from my instrument and study the passage carefully.

•• listen through the passage in my mind by imagining the sounds.
•• make the movements of playing the passage without making a sound.

Other groups of items used in the analysis were: V1, V11, A2, K3, and K4; V3, A1, and K2; V2,
A12, and K12. The dendrograms that were plotted indicated large variability between individu-
als, rather than clear groupings according to the divisions of PLS. Although some of the rela-
tionships between individuals were close, the analyses suggested that there was a core of
respondents who answered similarly, and who could be contrasted with the rest of the respond-
ents who did not answer similarly to the core nor to each other.
To investigate if the groupings of individuals in the dendrograms stayed similar across the
different groups of items used in the analysis, three individuals were selected completely at
random (S012, S068, and S081), and the clusters they occurred in in four dendrograms were
362 Psychology of Music 44(3)

compared. Cluster size was determined through membership of a common clade, or failing this
possibility, of a similar level of division from the main tree structure. A tabulation of this data
reveals that the number of individuals who occur together in more than one cluster is very low
(see Table 5). They are highest for S081, but even then they occur together only slightly more
than chance distribution. The almost non-existent triple and quadruple repeaters are very tell-
ing, indicating that very few people answered in ways that could be considered a perceptual
pattern across the different groups of items in the questionnaire.

Discussion
Neither of the analyses show an interpretable relationship between the data gathered and PLS.
The principal components analysis identified underlying variables that were not aligned with
perceptual modalities, but rather with musical and strategic issues. The weakness of the princi-
pal component structure – with generally low eigenvalues, few variables contributing to each
component, and the low percentage of variance explained by the model – points to less varia-
tion in the responses than the theory assumes. This is confirmed by the non-normal distribu-
tion of variables, which cannot be attributed to outliers and indicates a largely similar set of
responses to the questions posed. Of course, this data draws on self-reports and the respondents
may do very different things in their studios to what they say they do (see, for example, Lisboa,
Chaffin, & Logan, 2011, on this matter). For this reason observational studies investigating the
same topic from a different angle have also been conducted (Odendaal, 2013).
The higher-order components identified tensions inherent in the work of an instrumentalist,
poles between which to navigate the choices made in the practicing room every day. These pat-
terns suggest that the practicing of the respondents is not primarily characterized by choices
with regards to perceptual modality, but rather by issues such as whether to work more techni-
cally or more expressively, or whether to analyse more or to play more. Alternatively, perhaps
these patterns reveal choices in approach already made, and more fixed distinctions between
different musicians. This second interpretation could show the influence of aspects of personal-
ity or of habitual approach to practicing, both of which have been shown to influence strategy
selection (Hallam, 2001b; Kemp, 1997; Miksza, 2006).
There was some influence of instrument grouping on the total variation in the question-
naire, but this may be an artefact of the division of the groups used for the analysis rather than
a reflection on instrumental difference. However, some of these differences were distinct from
any physical properties inherent in the instruments, and the affordances or limitations that
they present. These may point to differences in instructional culture in the different groupings.
Gaunt and Hallam (2009) have suggested that interactions between personality, practicing and
performance may be affected by instrument type in professional musicians (see also Kemp,
1997) and this may be another factor in the variation observed in the questionnaire data with
regards to instrument type.
The cluster analysis further confirmed the conclusions of the principal components analysis
by showing that individuals did not readily group into any kind of distinctive modality prefer-
ence clusters, and neither did those who clustered together for a specific set of variables do the
same with other sets of variables that ostensibly measure the same modality differences.
Based on these analyses, the claim that PLS plays a role in the ways that students from the
researched institution describe their strategy selection must be rejected. The descriptions of
practicing in this study are not modality specific, nor do they suggest any kind of pattern that
can be understood to be modality specific. This finding occurred despite great care taken to use
the formulations and categories given by theorists of PLS in music, and thus points to weak-
nesses in the formulation of the theory as applied to musicians.
 Odendaal

Table 5.  Tabulation of the number of individuals that occur together across the four dendrograms.

Respondent Total respondents Respondents Respondents Respondents who Respondents who Percentage of respondents
in shared clusters who occur in one who occur in two occur in three occur in four clusters who occur together in more
across the four cluster across the clusters across the clusters across the across the four than one cluster across the
dendrograms four dendrograms four dendrograms four dendrograms dendrograms four dendrograms
S012 111 72 (65%) 31 (28%) 8 (7%) 0 (0%) 35%
S068  59 47 (80%) 12 (20%) 0 (0%) 0 (0%) 50%
S081 117 56 (48%) 59 (50%) 2 (2%) 0 (0%) 52%
363
364 Psychology of Music 44(3)

A possible objection to this finding is that some of the items in the questionnaire are vague
with regards to their principal perceptual modality. For instance: ‘I write out sections of the
piece, either in full or in reduced notation.’ Does this represent a primarily visual or kinesthetic
experience? Others are vague in the sense that they might be about something other than gath-
ering and processing information through a modality. For instance: ‘I use a metronome while I
am playing.’ Is this about hearing the regular beat or about external regulation of behavior?
The difficulty in designing questionnaire items that address a single modality reflects the gen-
eral problem with adapting PLS to musical learning, which is that modalities cannot be easily
or clearly separated in musical activities. Visual, auditory and kinesthetic sensory information
are integrally linked in musical performance, and to separate them involves a fundamental
reduction of the experience of playing (for discussion of this issue in general learning, see
Bertelson & De Gelder, 2004; Massaro, 2004).
Why is there a discrepancy between the findings presented here and the literature cited in
support of the theory in the introduction? Firstly, several of the studies cited investigated young
children, and there is surely a developmental aspect to perceptual use. This is certainly the case
with strategy selection (Hallam, 2001a; Nielsen, 1999), and makes sense, given the accumu-
lated hours of work developing aspects such as inner hearing or fine motor control typical of an
instrumentalist in higher music education (see, Ericsson, 1997; Ericsson & Charness, 1994).
This must challenge the assumption that PLS is stable over time (see Coffield, Moseley, Hall, &
Ecclestone, 2004, p. 9; although, see also Fleming, 2006, p. 56). And if PLS is not stable over
time, would a more appropriate path not be to train aspiring musicians to develop weaker
aspects, rather than to pander to strengths?
Secondly, the experimental studies that supported PLS all investigated a single aspect such as
listening or learning a rhythm. In each case the contact period was relatively short, and the
tasks focused. Learning to perform a complex work requires much more than such focused
tasks from the musician. Emotional regulation, motivational regulation, meaning making,
coordination, long-term memory encoding, and managing stress and energy levels, are only
some of the things that musicians deal with as they are learning to play a new work (c.f.
Gabrielsson, 1999, 2003). Arguably, many of these have less impact on a short experimental
setting. It is possible that these layers in the learning process mask any effects that PLS might
have had, had it survived the years of training to develop each of the modalities that musicians
typically engage in.

Future research
Although the questionnaire was based on claims by PLS theorists, it should not be used as a tool
for perceptual modality preference identification because it seems that the theory on which it is
based has been poorly formulated. The questionnaire is furthermore not complete enough to
aid reflection on practicing, and inclusion of other items drawn from the practicing literature
that do not specifically address only perceptual issues but also time management, motivation,
organization, and other aspects need to be included if this was the intention. Future research
into the variation between individual musicians would need to consider social and environ-
mental factors, and historical, motivational, cognitive, and personality factors (see, for exam-
ple, Gabrielsson, 1999, 2003; Gaunt & Hallam, 2009).

Funding
The research was funded by a grant which I received for my doctoral studies from the “Finnish Doctoral
Programme for Music Research.” It is administered by the Sibelius Academy of the University of the Arts
Helsinki.
Odendaal 365

Notes
1. Several other names are used to denote the theory: most popular are VAK (Visual, Auditory,
Kinesthetic), VAKT (Visual, Auditory, Kinesthetic, Tactile) and VARK (Visual, Auditory, Read/Write,
Kinesthetic).
2. See the discussion on this topic at www.talkstats.com/showthread.php/17326-Normality-
assumption-for-PCA

References
Barbe, W. B., & Swassing, R. H. (1979). Teaching through modality strengths: Concepts and practices.
Columbus, OH: Zaner-Bloser.
Beheshti, S. (2009). Improving studio music teaching through understanding learning styles. International
Journal of Music Education, 27(2), 107–115.
Bertelson, P., & De Gelder, B. (2004). The psychology of multimodal perception. In C. Spence & J. Driver
(Eds.), Crossmodal space and crossmodal attention (pp. 141–177). Oxford, UK: Oxford University Press.
Bruner, J. S. (1964). The course of cognitive growth. American Psychologist, 19, 1–15.
Bruner, J. S. (1966). Toward a theory of instruction. New York, NY: W.W. Norton.
Calissendorff, M. (2006). Understanding the learning style of pre-school children learning the violin.
Music Education Research, 8(1), 83–96.
Campbell, P. S., Scott-Kassner, C., & Kassner, K. (2006). Music in childhood: From preschool through the
elementary grades. Belmont, CA: Thomson Schirmer.
Coffield, F., Moseley, D., Hall, E., & Ecclestone, K. (2004). Learning styles and pedagogy in post-16 learning: A
systematic and critical review. London, UK: Learning and Skills Research Centre.
Dunn, R., & Griggs, S. A. (2003). Synthesis of the Dunn and Dunn Learning-Styles Model Research: Who,
what, where, when, why and so what? Jamaica, NY: St. John’s University.
Dunn, R. E. (2008). The effect of auditory, visual or kinesthetic perceptual strengths on music listening.
Contributions to Music Education, 35, 47–78.
Ericsson, K. A. (1997). Deliberate practice and the acquisition of expert performance: An overview. In H.
Jørgensen & A. C. Lehmann (Eds.), Does Practice Make Perfect? Current theory and research on instru-
mental music practice (pp. 9–52). Oslo, Norway: Norges Musikkhøgskole.
Ericsson, K. A., & Charness, N. (1994). Expert performance: Its structure and acquisition. American
Psychologist, 49(8), 725–747.
Everett, Y. U. (1997). Modality-based aural skills pedagogy: Ear-training strategies for post-tonal and
non-western musics. Journal of Music Theory Pedagogy, 11, 27–58.
Fleming, N. D. (2006). Teaching and learning styles: VARK strategies (2nd ed.). Christchurch, New Zealand:
Neil D. Fleming.
Gabrielsson, A. (1999). The performance of music. In D. Deutsch (Ed.), The Psychology of Music (2nd ed.,
pp. 501–602). San Diego, CA: Academic Press.
Gabrielsson, A. (2003). Music performance research at the millennium. Psychology of Music, 31(3), 221–
272.
Gadermann, A. M., Guhn, M., & Zumbo, B. D. (2012). Estimating ordinal reliability for Likert-type and
ordinal item response data: A conceptual, empirical, and practical guide. Practical Assessment,
Research & Evaluation, 17(3), 1–13.
Garcia, S. (2002). Learning styles and piano teaching. Piano Pedagogy Forum, 5(1). Retrieved from http://
www.music.sc.edu/ea/keyboard/PPF/5.1/5.1.PPFpp.html
Gardner, H. (1985). Frames of mind: The theory of multiple intelligences. London, UK: Paladin.
Gault, B. (2005). Music learning through all the channels: Combining aural, visual, and kinesthetic strat-
egies to develop musical understanding. General Music Today, 19(1), 7–9.
Gaunt, H., & Hallam, S. (2009). Individuality in the learning of musical skills. In S. Hallam, I. Cross
& M. H. Thaut (Eds.), The Oxford handbook of music psychology (pp. 274–284). Oxford, UK: Oxford
University Press.
Gorsuch, R. L. (1983). Factor analysis (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.
366 Psychology of Music 44(3)

Grinder, J., & Bandler, R. (1976). The structure of magic II. Palo Alto, CA: Science and Behaviour Books.
Hallam, S. (2001a). The development of expertise in young musicians: Strategy use, knowledge acquisi-
tion and individual diversity. Music Education Research, 3(1), 7–23.
Hallam, S. (2001b). The development of metacognition in musicians: Implications for education. British
Journal of Music Education, 18(1), 27–39.
Hannaford, C. (1995). Smart moves: Why learning is not all in your head. Arlington, VA: Great Ocean
Publishers.
Hatcher, L. (1994). A step-by-step approach to using the SAS system for factor analysis and structural equation
modeling. Cary, NC: SAS Institute.
Hill, T., & Lewicki, P. (2007). Statistics: Methods and applications. Tulsa, OK: StatSoft.
Holmes, P. (2005). Imagination in practice: A study of the integrated roles of interpretation, imagery and
technique in the learning and memorisation processes of two experienced solo performers. British
Journal of Music Education, 22(3), 217–235.
Hultberg, C. (2008). Instrumental students’ strategies for finding interpretations: Complexity and indi-
vidual variety. Psychology of Music, 36(1), 7–23.
Jacobson, J. (2006). Professional piano teaching: A comprehensive piano pedagogy textbook for teaching elemen-
tary-level students. Van Nuys, CA: Alfred.
Kemp, A. E. (1997). Individual differences in musical behaviour. In D. J. Hargreaves & A. C. North (Eds.),
The Social Psychology of Music (pp. 25–45). Oxford, UK: Oxford University Press.
Korenman, L., & Peynircioglu, Z. (2007). Individual differences in learning and remembering music:
Auditory versus visual presentation. Journal of Research in Music Education, 55(1), 48–64.
Krätzig, G. P., & Arbuthnott, K. D. (2006). Perceptual learning style and learning proficiency: A test of the
hypothesis. Journal of Educational Psychology, 98(1), 238–246.
Lisboa, T., Chaffin, R., & Logan, T. (2011). A self-study of practice: Words versus action in music problem
solving. In A. Williamon, D. Edwards & L. Bartel (Eds.), Proceedings of the International Symposium
on Performance Science 2011 (pp. 517–522). Toronto, ON: European Association of Conservatoires.
MacCallum, R. C., Widaman, K. F., Preacher, K. J., & Hong, S. (2001). Sample size in factor analysis: The
role of model error. Multivariate Behavioural Research, 36(4), 611–637.
MacCallum, R. C., Widaman, K. F., Zhang, S., & Hong, S. (1999). Sample size in factor analysis.
Psychological Methods, 4(1), 84–99.
Massaro, D. W. (2004). From multisensory integration to talking heads and language learning. In G.
A. Calvert, C. Spence & B. E. Stein (Eds.), The Handbook of Multisensory Processes (pp. 153–176).
Cambridge, MA: MIT Press.
Miklaszewski, K. (1995). Individual differences in preparing a musical composition for public perfor-
mance. In M. Manturzewska, K. Miklaszewski & A. Bialkowski (Eds.), Psychology of Music Today:
Proceedings of the international seminar of researchers and lecturers in the psychology of music (pp. 137–
148). Warsaw, Poland: Fryderyk Chopin Academy of Music.
Miksza, P. (2006). Relationships among impulsiveness, locus of control, sex, and music practice. Journal
of Research in Music Education, 54(4), 308–323.
Miller, B. A. (2002). Touch the music! Learning modalities in elementary music class. General Music
Today, 15(2), 4–13.
Mishra, J. (2007). Correlating musical memorization styles and perceptual learning modalities. Visions of
Research in Music Education, 9/10, 1–19.
Mixon, K. (2004). Band: Three learning styles … four steps to reach them. Teaching Music, 11(4), 48–51.
Nielsen, S. G. (1999). Learning strategies in instrumental music practice. British Journal of Music Education,
16(3), 275–291.
Nielsen, S. G. (2008). Instrumental practicing and self-regluation: A social cognitive perspective. Finnish
Journal of Music Education, 11(1–2), 19–25.
Norušis, M. J. (2011). The IBM SPSS Statistics 19 Statistical Procedures Companion. Upper Saddle River, NJ:
Prentice Hall.
Odendaal, A. (2011, May). Practicing habits shaped by perception: Results of two pilot studies. Poster pre-
sented at the Performa’11 Conference for Performance Studies, Aviero, Portugal.
Odendaal 367

Odendaal, A. (2013). Perceptual learning style as an influence on the practising of instrument students in higher
music education. Studia Musica 56. Helsinki: Sibelius Academy.
Odendaal, A. (2014, July). Solo practising? Challenging assumptions about the role of social mediation
in the practising of musicians. In D. Forrest (Ed.), Proceedings of the International Society for Music
Education 31st World Conference on Music Education, Porto Alegre, Brazil (pp. 222–227). Victoria,
Australia: International Society for Music Education.
Pashler, H., McDaniel, M., Rohrer, D., & Bjork, R. (2009). Learning styles: Concepts and evidence.
Psychological Science in the Public Interest, 9(3), 105–119.
Persellin, D. C., & Pierce, C. (1988). Association of preference for modality to learning of rhythm patterns
in music. Perceptual and Motor Skills, 67, 825–826.
Sharpley, C. F. (1984). Predicate matching in NLP: A review of research on the preferred representational
system. Journal of Counseling Psychology, 31(2), 238–248.
Sharpley, C. F. (1987). Research findings on neurolinguistic programming: Nonsupportive data or an
untestable theory? Journal of Counseling Psychology, 34(1), 103–107.
Sullivan, Y., & Cantwell, R. H. (1999). The planning behaviors of musicians engaging traditional and
non-traditional scores. Psychology of Music, 27(2), 245–266.
Swanson, C. D. (2005). The private studio: Learning styles and the voice student. Journal of Singing, 62(2),
203–208.
Zikmund, A. B., & Nierman, G. E. (1992). The effect of perceptual mode preferences and other selected var-
iables on upper elementary school students’ responses to conservation-type rhythmic and melodic
tasks. Psychology of Music, 20(1), 57–69.

Appendix

English version of the questionnaire used in the study, arranged by variable

number
Respond to the following statements by imagining that you are about to start learning to play a large-
scale work (like a sonata or a concerto) from the classical or romantic era on your primary instrument.
(Never, very rarely, rarely, occasionally, frequently, always)
V1 One of my aims is to be able to visualize the score in my mind’s eye.
V2 I find it hard to concentrate on my playing when I see movement in the room.
V3 If I feel that I will lose my place while I am practising, I look at the notation as a reminder.
V4 I close my eyes in order to see my instrument and fingers/hands in my mind’s eye.
V5 I figure out how the piece is structured by looking carefully through the score away from my
instrument.
V6 I figure out how the piece is structured by playing and not by analysis.
V7 I make notes to myself on the sheet music.
V8 I use colours to mark my sheet music.
V9 I consult several editions of the piece and compare similarities and differences.
V10 I look at video recordings of other performers playing the same piece I am learning.
V11 One of my aims is to remember what my fingers/hands look like when playing the piece.
V12 I write out sections of the piece, either in full or in reduced notation.
V13 I draw graphic representations of the piece to make the structure clear.
V14 When I am struggling with a passage it helps when I sit away from my instrument and
study the passage carefully.
V15 I read a lot of information about the piece I am playing.
V16 It is important to me to recognize patterns in the music that I am learning.
368 Psychology of Music 44(3)

A1 If I feel that I will lose my place while I am practising, I sing or hum along (aloud or inter-
nally) to help me concentrate.
A2 One of my aims is to be able to hear every note of the piece in my mind’s ear.
A3 I play sections of the piece and then listen to those passages in my head without playing.
A4 I listen to audio recordings of other performers playing the same piece I am learning.
A5 I record myself and listen to my playing.
A6 I speak the rhythms of sections that I find difficult.
A7 I sing melodic material from the piece to myself when I am not playing.
A8 I hum melodies and/or rhythms to myself while playing.
A9 I talk to myself about details while I am working.
A10 I try to pay attention to phrasing and dynamics from the first reading.
A11 I talk to someone who has played the piece before.
A12 I find it hard to concentrate on my playing when I can hear other music playing
A13 I use a metronome while I am playing.
A14 I use a metronome to find the tempos, but not while I play.
A15 I count out loud to help me figure out difficult rhythms.
A16 When I am struggling with a passage it helps when I listen through the passage in my
mind by imagining the sounds.
K1 When I am struggling with a passage it helps when I make the movements of playing the
passage without making a sound.
K2 If I feel that I will lose my place while I am practising, I try to let my fingers play automatically.
K3 One of my aims is to be able to play the music automatically, without too much thought.
K4 One of my aims is to be able to think through the movements I will make while playing.
K5 I go through the motions of playing sections of the piece without making a sound.
K6 I stop playing and think through the movements I will make while playing.
K7 I make technical exercises from sections of the piece.
K8 I use conducting gestures or other bodily movements to help me find the right phrasing.
K9 I focus on technical demands in the early stages of practising, and work on musical shaping
at a later stage.
K10 I try to find the parts that are technically difficult in order to focus on them.
K11 I try to play slowly in the early practice sessions.
K12 I find it hard to concentrate on my playing when the physical environment (seating, tem-
perature, lighting) is not comfortable.
K13 I simplify difficult passages by taking away some of the notes.
K14 I tap my feet (or other body parts) while I play to help with rhythmical difficulties.
K15 I play passages from the piece using different rhythms in order to ‘get it in my fingers.’