ArticlePDF Available

It's Not How Much; It's How: Characteristics of Practice Behavior and Retention of Performance Skills



We observed 17 graduate and advanced-undergraduate piano majors practicing a difficult, three-measure keyboard passage from a Shostakovich concerto. Participants' instructions were to practice until they were confident they could play the passage accurately at a prescribed tempo in a retention test session the following day. We analyzed the practice behaviors of each pianist in terms of numeric and nonnumeric descriptors and ranked the pianists according to the overall performance quality of their retention tests. Results indicated no significant relationship between the rankings of pianists' retention test performances and any of the following variables: practice time, number of total practice trials, and number of complete practice trials. There were significant relationships between retention test rankings and the percentage of all performance trials that were performed correctly, r = —.51, the percentage of complete performance trials that were performed correctly, r = —.71, and the number of trials performed incorrectly during practice, r = .48. The results showed that the strategies employed during practice were more determinative of performance quality at retention than was how much or how long the pianists practiced, a finding consistent with the results of related research.
This research was funded by the Marlene and Morton Meyerson Centennial Professorship in Music and
Human Learning at The University of Texas at Austin. Please address correspondence to Robert A. Duke,
Center for Music Learning, The University of Texas at Austin, 1 University Station E3100, Austin, TX
78712-1208; e-mail:
Journal of Research in
Music Education
Volume 56 Number 4
January 2009 310-321
© 2009 MENC: The National
Association for Music Education
It’s Not How Much; It’s How
Characteristics of Practice Behavior
and Retention of Performance Skills
Robert A. Duke
The University of Texas at Austin
Amy L. Simmons
The University of Texas at San Antonio
Carla Davis Cash
Texas Tech University, Lubbock
We ob s erve d 17 gr a dua t e and ad van c ed- u nde r gra d uat e p ian o m ajo r s prac t ici n g a dif fi c ult,
three-measure keyboard passage from a Shostakovich concerto. Participants’instructions
were to practice until they were confident they could play the passage accurately at a
prescribed tempo in a retention test session the following day. We analyzed the practice
behaviors of each pianist in terms of numeric and nonnumeric descriptors and ranked the
pianists according to the overall performance quality of their retention tests. Results
indicated no significant relationship between the rankings of pianists’ retention test
performances and any of the following variables: practice time, number of total practice
trials, and number of complete practice trials. There were significant relationships
between retention test rankings and the percentage of all performance trials that were
performed correctly, r=–.51,thepercentageofcompleteperformancetrialsthatwere
performed correctly, r=–.71,andthenumberoftrialsperformedincorrectlyduring
practice, r=.48.Theresultsshowedthatthestrategiesemployedduringpracticewere
more determinative of performance quality at retention than was how much or how long
the pianists practiced, a finding consistent with the results of related research.
Keywords: motor skill learning; music learning; practice
Tens of thousands of hours in musicians’professional lives are devoted to individual
practice, the mechanism through which music skills are learned, refined, and main-
tained (Davidson, Howe, & Sloboda, 1997; Ericsson, 1997; Ericsson, Krampe, &
Tesch-Römer, 1993; Howe, Davidson, & Sloboda, 1998; Madsen, 2004). Although
this private aspect of musicianship is invisible to most nonmusicians, who typically
hear only public performances, musicians are well aware of the centrality of practice
in their life’s work.
by Robert Duke on October 31, 2009 http://jrm.sagepub.comDownloaded from
Duke et al. / Characteristics of Practice 311
Much of the extant research about music practice comprises comparisons among
practice interventions or imposed strategies, testing the relative efficacy of modeling
(Henley, 2001; Hewitt, 2001; Rosenthal, 1984; Rosenthal, Wilson, Evans, & Greenwalt,
1988), mental practice (Coffman, 1990; Lim & Lippman, 1991; Ross, 1985; Rubin-
Rabson, 1941a, 1941b, 1941c; Theiler & Lippman, 1995), practice reports (Wagner,
1975), and distraction indexes (Madsen & Geringer, 1981).
Only more recently have scholars begun to study in context and over time the
content of expert musicians’ practice (Chaffin & Imreh, 1997, 2001, 2002; Chaffin,
Imreh, Lemieux, & Chen, 2003; Gruson, 1988; Maynard, 2006; Williamon,
Valentine, & Valentine, 2002) and the practice of novices (McPherson, 2005;
Rohwer & Polk, 2006). Williamon and Valentine (2000), for example, observed
practice among pianists at four different skill levels and found that quality, not quan-
tity, of practice predicted performance quality at all levels of skill. In her study of
pianists’ practice, Gruson (1988) reported that the single best predictor of skill level
was the extent to which players repeated larger sections of music, rather than indi-
vidual notes. McPherson (2005), whose 3-year investigation of 157 beginning
instrumentalists is one of the most substantive to date, also noted that the strategies
employed in young musicians’ practice, not the amount of time devoted to practice,
were the best predictors of achievement.
Irrespective of the pedagogical implications of the more recent studies of practice
behavior, making practice assignments in terms of time practiced instead of goals
accomplished remains one of the most curious and stubbornly persistent traditions in
music pedagogy (Kostka, 2002). Music teachers more often ask students to record
practice time than they ask them to record the achievement of practice goals (Barry
prescribed number of minutes each day, regardless of how long it takes individuals to
accomplish what they set out to do (Duke, Flowers, & Wolfe, 1997). In fact, informal
reviews of private teachers’ instructions for practice reveal that teachers commonly
assign only what to practice and how long to practice, with little attention given to
specific proximal goals to be accomplished each day. This is in stark contrast to
assignments in many academic disciplines in school, where students are given sets of
problems to solve, chapters to read, or essays to write, and the time devoted to home-
work is determined by the time required to complete the problems, read the chapters,
or compose the essays. It seems readily accepted in other disciplines by teachers and
students alike that all students will not devote the same amount of time to assign-
ments, because individual learners work at different rates and different learners will
not require the same amount of time to complete each assignment. How long one
works depends on how long it takes to accomplish the assigned goals.
Although similar individual variations exist in the time required to accomplish
performance goals in music, setting daily goals according to time spent seems to
have become an accepted convention in planning music practice, most noticeably
by Robert Duke on October 31, 2009 http://jrm.sagepub.comDownloaded from
312 Journal of Research in Music Education
among developing musicians. This is perhaps because of the notion that setting a
routine for practice develops habits of consistent effort or because of the belief that
daily repetition will inevitably lead advantageously to automaticity of motor skills.
But, if the efforts made in practice are generally ineffective in improving performance,
it is understandable that some learners conclude that their limited accomplishment is
not worth the time invested.
There have been few studies to date in which skilled performers’ practice behav-
iors have been observed in detail, the work of Chaffin and colleagues being the most
obvious exception. The reasons for this are understandable, because the challenges
of assessing practice behavior in a way that is systematic yet informative (beyond the
measurement of discrete variables, such as time or duration of performance episodes)
are daunting, to say the least.
The purpose of this study was to test the extent to which the quality of advanced
pianists’ performances of a difficult passage approximately 24 hr after it was introduced
could be predicted based on what the pianists did during practice on the passage. We also
set out to describe practice behaviors of the most effective learners in our sample.
Participants were 17 graduate and advanced-undergraduate piano majors enrolled
in piano performance and piano pedagogy degree programs in the Sarah and Ernest
Butler School of Music at The University of Texas at Austin. We were able to iden-
tify 17 pianists who were willing to complete the research sessions.
Participants learned to play a three-measure passage from Dmitri Shostakovich’s
Concerto No. 1 for Piano, Trumpet and String Orchestra, Op. 35 on a Yamaha
Disklavier acoustic piano (see Figure 1). We chose the excerpt on the basis of its dif-
ficulty and accessibility—although it presents a number of technical challenges
to the pianist and is quite difficult to sightread at tempo, it can be learned within a
single practice session.
Upon arrival at the test location, participants were given approximately 2 min to
warm up in whatever manner they wished. At the conclusion of the warm-up period,
they were given a printed copy of the test excerpt (marked with their participant
number), an electronic metronome, and a pencil. The following instructions were
read aloud by the proctor, who remained in the room throughout all practice and test
Practice this excerpt until you feel that you have learned it well and can play it confidently
at the target tempo (120 bpm) without the metronome. Take as much time as you need.
When you return tomorrow, you will play this excerpt again. The purpose of this project
is to describe the changes that occur in your playing of the excerpt between today and
by Robert Duke on October 31, 2009 http://jrm.sagepub.comDownloaded from
Duke et al. / Characteristics of Practice 313
We permitted participants to practice for as long as they wished. There were no
explicit instructions concerning use of the metronome and pencil given to partici-
pants at the start of the session, and no apparent patterns were observed in the use of
these items. Some participants used the metronome intermittently throughout prac-
tice, whereas others used it only as they neared the end of practice. The majority of
participants wrote several fingerings in their scores (mostly for left-hand passages),
and only 3 of the 17 pianists made additional markings in their music (e.g., circling
notes, fingerings, or challenging transition points). When participants indicated that
they were confident they had learned the excerpt and that they could play the excerpt
at the target tempo on the following day, the proctor collected the music and
instructed the participants not to practice the excerpt (even from memory) before
returning 24 hr later. (The participants reported that they complied with our request
not to practice the passage.)
When participants returned the following day, they were read the following
You have approximately 2 minutes to warm up as you wish. Please do not play any part
of the excerpt you learned yesterday during the warm-up.
At the end of the warm-up period, participants were given the same copy of the
music they had used during their practice session on the previous day. The following
instructions were read:
Play straight through this excerpt at the target tempo 15 times. Please do not stop during
any of your performance trials.
The proctor then played the metronome at the target tempo until the participant
began his or her first trial, at which time the metronome was turned off. The 15
Figure 1
Test Excerpt From Dmitri Shostakovich’s Concerto No. 1
for Piano, Trumpet and String Orchestra, Op. 35
by Robert Duke on October 31, 2009 http://jrm.sagepub.comDownloaded from
performance trials were played in succession, separated by brief pauses whose dura-
tion were determined by the participants. We included 15 trials in the retention ses-
sion to provide us with enough material to make reliable discriminations among the
We rec o r ded al l p r actic e a n d test se s sions o n d i gital v i d eota p e a nd reco r d ed all M I D I
data from the keyboard for subsequent analysis. In addition, we converted the MIDI signal
from the retention test performances to QuickTime audio files for evaluation.
We observed all video recordings of the 17 participants’ practice sessions and
compiled the following numerical data from each session: total practice time,
number of performance trials (the number of times the pianist began playing), number of
complete performance trials (the number of hands-apart or hands-together performances
of the entire excerpt), number of correct performance trials (complete performances
of the entire excerpt at any tempo without error or hesitation), number of near-
correct performance trials (complete performances of the entire excerpt at any tempo with
only one or two minor errors or hesitations), the sum of correct and near-correct
performance trials, number of incorrect performance trials (performances of the
entire excerpt that contained errors), the percentage of complete trials that were cor-
rect (the proportion of performances of the entire excerpt that were without error or
hesitation), the percentage of complete trials that were correct and near-correct (the
proportion of performances of the entire excerpt that were without error or hesita-
tion plus those with only one or two minor errors or hesitations), and the percentage
of all trials (including incomplete trials) that were correct. From the retention tests,
we recorded the number of trials out of 15 attempts that were correct, the number of
trials that were near-correct, and the sum of the correct and near-correct trials. Note
that the correct and near-correct trials in these analyses were defined only in terms
of pitch and rhythm accuracy. Interjudge reliability for our assessments of correct
and near-correct trials was .96.
After recording the numerical data, we observed the video recordings again in an
effort to characterize further the features of each participant’s practice procedures. We
wrote explicit descriptions of the practice behaviors that appeared rather consistently
throughout each participant’s practice session. When there were discrepancies among
our descriptions, we arrived at consensus after viewing the tapes together to clarify one
another’s observations. Thus, the descriptors of practice presented in Table 1 represent
only those characteristics that were agreed upon by all three authors.
We independently ranked the audio recordings of the retention test performances,
taking into account the tone, character, and expressiveness of the performances. This
ranking procedure permitted a more encompassing evaluation of the retention test
performances, beyond the simple counting of correct and near-correct trials. The 17
QuickTime audio files were assigned random ID numbers that differed from the
sequential numbering of the practice videotapes and were placed in an otherwise-
empty folder on an Apple Macintosh computer screen. Clicking on a given icon ini-
tiated playback of the participant’s 15 retention test performances. We listened
314 Journal of Research in Music Education
by Robert Duke on October 31, 2009 http://jrm.sagepub.comDownloaded from
Duke et al. / Characteristics of Practice 315
individually to the recordings over high-quality loudspeakers. Our task was to rank
the participants from best to worst in terms of overall performance quality across the
15 trials in the retention test. This procedure for evaluating multiple music perfor-
mances proved quite advantageous, as it combined unlimited opportunities for
rehearing, ease in hearing performances in juxtaposition, and the ability to order on
the computer screen the icons representing the performers’ retention tests.
Agreement across our rankings was moderately high and certainly acceptable for the
purposes of our investigation, Kendall’s W= 0.83, p< .001.
We report the characteristics of each participant’s practice session in Table 1.
Participants’ data are ordered based on the mean of our rankings of the retention test
performances, which are positively correlated, but not perfectly correlated, with the
total number of correct and near-correct repetitions in the 15 retention test trials, r=
–.79, p< 001. The size of this correlation (64% shared variance between rankings of
overall performance quality and counts of correct and near-correct trials) provides
some indication of the extent to which performance quality variables other than
correct notes and rhythmic precision affected our ranking judgments.
We per f o rmed b i va r i ate cor r e latio n s betwe e n e ach of th e varia b les in th e t a ble and
the participants’ rankings. These results are reported in the bottom row of Table 1. In
addition to the understandably high correlation between the mean judge ranking and
the sum of correct and near-correct trials in the retention test, there were significant
correlations between the pianists’ rankings and the following: the number of complete,
incorrect performance trials, r=.48,p=.05;thepercentageofallcompletetrialsthat
were correct, r=–.71,p=.001;thepercentageofcompletetrialsthatwerecorrectand
near-correct, r=–.64,p=.006;andthepercentageoftotalperformancetrials(includ-
ing incomplete trials) during practice that were correct, r=–.51,p<.04.
It seems equally important to point out the variables that were not related signif-
icantly to participants’ retention test ranks: the total time practiced, r= .18, p> .49;
the total number of performance trials, r= .12, p> .65; the total number of complete
trials, r= .02, p> .93; and the total numbers of correct and near-correct trials, r=
–.15, p> .56. This seems an indication that the nature of the practice defined in our
observations was more determinative of retention test performance than was the
amount of practice. We found that the three participants whose retention tests earned
the highest ranks were clearly superior to the next-highest-ranked participants (data
for those three participants appear above the line in the table). The retention test per-
formances by these three pianists were distinguished from the performances of the
other participants by a more consistently even tone, greater rhythmic precision,
greater musical character (purposeful dynamic and rhythmic inflection), and a more
fluid execution.
by Robert Duke on October 31, 2009 http://jrm.sagepub.comDownloaded from
Table 1
Descriptive Data for Individual Participants, Rank-Ordered by Retention Test Performance Quality
Practice Session Retention Test (15 trials)
% of
Time Correct +% of Completed % of Correct +
Practice (in Near- Near- Completed Correct +Total Near- Near-
Characteristics Degree Rank seconds) Trials Complete Correct Correct Correct Incorrect Correct Near-Correct Correct Correct Correct Correct
A B C D E F G H I J K DMA 1 1400 288 81 59 16 75 6 .73 .93 .20 3 12 15
A B C D E F G H I J K MM 2 1620 204 48 45 2 47 1 .94 .98 .22 7 5 12
ABCDEFGHIJ K DMA 3 711 116 48 28 11 39 9 .58 .81 .24 11 2 13
ABCDEFGHIJKMM 4 514 114 38 21 13 34 3 .55 .89 .18 5 6 11
ABCDEFGHIJKDMA 5 530 113 27 14 12 26 1 .52 .96 .12 4 6 10
ABCDEFGHIJ K MM 6 556 110 34 28 6 34 0 .82 1.00 .25 4 7 11
ABCDEFGHIJ K MM 7 598 124 47 38 8 46 1 .81 .98 .31 3 7 10
ABCDEFGHIJK BM (jr) 8 1146 339 27 15 9 24 3 .56 .89 .04 0 9 9
ABCDEFGHIJK BM (jr) 9 2998 739 79 57 10 67 12 .72 .85 .08 0 9 9
ABCDEFGHIJK DMA 10 535 112 43 14 21 35 16 .33 .81 .13 10 3 13
ABCDEFGHIJK DMA 11 678 111 24 3 18 21 3 .13 .88 .03 0 12 12
ABCDEFGHIJK MM 12 1110 209 74 37 23 60 14 .50 .81 .18 3 7 10
ABCDEFGHIJK BM (sr) 13 3410 689 114 2 80 82 32 .02 .72 .00 3 8 11
ABCDEFGHIJK BM (sr) 14 1520 225 100 63 32 95 5 .63 .95 .28 0 7 7
ABCDEFGHIJK BM (jr) 15 538 82 31 15 11 26 5 .48 .84 .18 2 3 5
ABCDEFGHIJK DMA 16 1590 226 40 0 11 11 29 .00 .28 .00 0 4 4
ABCDEFGHIJK MM 17 887 160 13 0 7 7 6 .00 .54 .00 0 0 0
Correlation (r) with 1.00 0.18 0.12 0.02 –0.44 0.31 –0.15 .48* –.71** –.64** –.51* –0.79**
retention test rank
Note: Reliability for judgments of ranking (three judges), W= .83; reliability for judgments of correct and near-correct trials (two judges), .96 agreements.
*p< .05. **p< .01.
by Robert Duke on October 31, 2009 http://jrm.sagepub.comDownloaded from
Duke et al. / Characteristics of Practice 317
This finding led us to begin our practice session observations with the sessions of
these 3 pianists in an effort to identify the elements that best characterized their
work. We reached consensus on the following eight elements, all of which except
Item D below appeared in the 3 top-ranked pianists’ practice sessions; Item D was
in evidence in 2 of the top 3 pianists’ sessions. Letter designations below correspond
to those in Table 1. The combination of practice strategies that characterized the
practice sessions of the top-ranked pianists was clearly absent in the sessions of
the other pianists, although many of the 14 lower-ranked pianists included some of the
strategies used by the top 3.
A. Playing was hands-together early in practice.
B. Practice was with inflection early on; the initial conceptualization of the music was
with inflection.
C. Practice was thoughtful, as evidenced by silent pauses while looking at the music,
singing/humming, making notes on the page, or expressing verbal “ah-ha”s.
D. Errors were preempted by stopping in anticipation of mistakes.
E. Errors were addressed immediately when they appeared.
F. The precise location and source of each error was identified accurately, rehearsed,
and corrected.
G. Tempo of individual performance trials was varied systematically; logically under-
standable changes in tempo occurred between trials (slowed down enough; didn’t
speed up too much).
H. Target passages were repeated until the error was corrected and the passage was
stabilized, as evidenced by the error’s absence in subsequent trials.
The following three observations also were made based on the three top-ranked
pianists’ practice sessions. We list them separately here because they are not practice
strategies but are nevertheless descriptive of the sessions we observed.
I. When tempo was changed, the first trial at the new tempo was nearly always accu-
J. After the initial learning phase, errors were only intermittent; there were no per-
sistent errors.
K. At least 20% of all starts were complete, correct performances, although not nec-
essarily at the target tempo of 120 bpm.
Our data describe the practice behaviors of multiple, advanced-level performers
learning the same excerpt. This is one of the few reported examples of research that
defines the characteristics of effective practice based on the observed behaviors of
multiple advanced performers with varied levels of practice skill. Our findings illu-
minate some of the important aspects of practice that differentiate more and less able
practicers, as determined by their performances 24 hr after practice.
by Robert Duke on October 31, 2009 http://jrm.sagepub.comDownloaded from
318 Journal of Research in Music Education
Our results show that, among our sample of performers, the strategies employed
during practice were more determinative of their retention test performances than
was how much or how long they practiced. It seems particularly notable that total
time and total number of performance trials were unrelated to the quality of the
retention test performances and that the best-performing pianists took no less time
to learn the passage than did the other pianists. This seems to contravene the notion
that the pianists who performed best on the retention test were able to learn the pas-
sage more quickly and more easily than the others. A more accurate summary of
what took place is that the top-ranked pianists learned the excerpt differently from
the other pianists.
The most notable differences between the practice sessions of the top-ranked
pianists and the remaining participants are related to their handling of errors. The
observations labeled F, G, and H in Table 1 were present in the sessions of all three
of the top-ranked pianists, but they appeared in few of the other pianists’ practice
sessions (none of the other pianists demonstrated all three of these characteristics).
The three characteristics are as follows:
F. The precise location and source of each error was identified accurately, rehearsed,
and corrected.
G. Tempo of individual performance trials was varied systematically; logically under-
standable changes in tempo occurred between trials (slowed down enough; didn’t
speed up too much).
H. Target passages were repeated until the error was corrected and the passage was
stabilized, as evidenced by the error’s absence in subsequent trials.
Thus, it seems that the actions taken subsequent to the discovery of errors were
major determinants of the effectiveness of practice. It was not the case that the top-
ranked pianists made fewer errors at the beginning of their practice sessions than did
the other pianists. But, when errors occurred, the top-ranked pianists seemed much
better able to correct them in ways that precluded their recurrence. This is an
extremely important point—that the effective handling of error correction led to a
higher proportion of correct, complete performance trials during practice.
The most effective way that the participants corrected errors was by making judi-
cious changes in performance speed that facilitated the maintenance of accuracy fol-
lowing the correction of a given error. Of course, there were other methods of
decontextualization, in addition to tempo change, that appeared among the strategies
employed by these performers (e.g., playing shorter passages, playing hands sepa-
rately), but the method of varying tempos was a distinctive feature of the top-ranked
pianists’ approaches. In fact, two of the top-ranked pianists made alterations in their
performance tempos that preempted errors before they occurred (labeled Item D in
Table 1). In other words, after a given error was discovered, these two pianists tended
to hesitate in subsequent trials as they approached the location of the error, often
slowing the tempo (without stopping) to a point at which the playing could continue
accurately past the location where the error had occurred in a preceding trial.
by Robert Duke on October 31, 2009 http://jrm.sagepub.comDownloaded from
These results point to the importance of developing in young musicians effective
approaches to correcting errors—procedures that preclude errors’ persistence.Yet, it
is rare in published methods to see examples of systematic instruction in problem
solving and error correction, even though devising solutions to problems is one of the
central features of learning. It is generally not the case that experts (in any discipline)
simply avoid making mistakes when they are learning something new, but experts
correct their mistakes efficiently and effectively. Thus, it seems that error correction
should be a prominent part of novices’ instruction and that the most appropriate goal
for young learners is not that they play their instruments for 30 minutes a day but
that they skillfully identify and systematically address the mistakes that are an
inevitable part of learning.
It is understandable that many students of music are of the mind that their primary
goal is to avoid making mistakes, and most young players are not privy to what goes
on in experts’ individual practice sessions, including their own teachers’ practice ses-
sions. These same students may come to believe that a major difference between them
and their teachers is that their teachers seldom make mistakes at all, when in fact all
learners, including experts, make mistakes as they take on new challenges, learn new
repertoire, encounter new problems, and teach new students. Experienced musicians’
expertise is characterized by their ability to deal with mistakes and solve knotty prob-
lems in ways that maximize efficiency and lead to lasting solutions.
There is no doubt that most students have heard their teachers demonstrate good
playing, but it is probably also true that few have observed their teachers encounter-
ing performance problems and advantageously addressing them. If there is broad
agreement that providing good models is an effective strategy for learning, then why
are there so few available models of effective practice? It is clear that this question
deserves considerable attention in the future.
Barry, N. H., & McArthur, V. (1994). Teaching practice strategies in the music studio: A survey of applied
music teachers. Psychology of Music,22, 44–55.
Chaffin, R., & Imreh, G. (1997). “Pulling teeth and torture”: Musical memory and problem solving.
Thinking & Reasoning,3(4), 315–336.
Chaffin, R., & Imreh, G. (2001). A comparison of practice and self-report as sources of information about
the goals of expert practice. Psychology of Music,29, 39–69.
Chaffin, R., & Imreh, G. (2002). Practicing perfection: Piano performance as expert memory.
Psychological Science,13, 342–349.
Chaffin, R., Imreh, G., Lemieux, A. F., & Chen, C. (2003). “Seeing the big picture”: Piano practice as
expert problem solving. Music Perception,20, 465–490.
Coffman, D. D. (1990). Effects of mental practice, physical practice, and knowledge of results on piano
performance. Journal of Research in Music Education,38, 187–196.
Davidson, J. W., Howe, M. J. A., & Sloboda, J. A. (Eds.). (1997). Environmental factors in the develop-
ment of musical performance skill over the life span.NewYork:OxfordUniversityPress.
Duke et al. / Characteristics of Practice 319
by Robert Duke on October 31, 2009 http://jrm.sagepub.comDownloaded from
Duke, R. A., Flowers, P. J., & Wolfe, D. E. (1997). Children who study piano with excellent teachers.
Bulletin of the Council for Research in Music Education,132, 51–84.
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
instrumental music practice, pp. 9-51. Oslo, Norway: Norges Musikkhøgskole.
Ericsson, K. A., Krampe, R. T., & Tesch-Römer, C. (1993). The role of deliberate practice in the acquisi-
tion of expert performance. Psychological Review,100, 363–406.
Gruson, L. M. (Ed.). (1988). Rehearsal skill and musical competence: Does practice make perfect? In
J. A. Sloboda (Ed.), Generative processes in music: The psychology of performance, improvisation
and composition (pp. 91-112). Oxford, UK: Clarendon.
Henley, P. T. (2001). Effects of modeling and tempo patterns as practice techniques on the performance
of high school instrumentalists. Journal of Research in Music Education,49, 169–180.
Hewitt, M. P. (2001). The effects of modeling, self-evaluation, and self-listening on junior high instrumentalists’
music performance and practice attitude. Journal of Research in Mu sic Education,49,307322.
Howe, M. J. A., Davidson, J. W., & Sloboda, J. A. (1998). Innate talents: Reality or myth? Behavioral &
Brain Sciences,21, 399–442.
Kostka, M. J. (2002). Practice expectations and attitudes: A survey of college-level music teachers and
students. Journal of Research in Music Education,50, 145–154.
Lim, S., & Lippman, L. G. (1991). Mental practice and memorization of piano music. Journal of General
Psychology,118(1), 21–30.
Madsen, C. K. (2004). A 30-year follow-up study of actual applied music practice versus estimated practice.
Journal of Research in Mu sic Education,52,7788.
Madsen, C. K., & Geringer, J. M. (1981). The effect of a distraction index on improving practice attentiveness
and musical performance. Bulletin of the Council for Research in Music Education,66-67,4652.
Maynard, L. (2006). The role of repetition in the practice sessions of artist-teachers and their students.
Bulletin of the Council for Research in Music Education,167, 61–72.
McPherson, G. E. (2005). From child to musician: Skill development during the beginning stages of
learning an instrument. Psychology of Music,33(1), 5–35.
Rohwer, D., & Polk, J. (2006). Practice behaviors of eighth-grade instrumental musicians. Journal of
Research in Music Education,54, 350–362.
Rosenthal, R. K. (1984). The relative effects of guided model, model only, guide only, and practice only
treatments on the accuracy of advanced instrumentalists’ musical performance. Journal of Research
in Music Education,32, 265–273.
Rosenthal, R. K., Wilson, M., Evans, M., & Greenwalt, L. (1988). Effects of different practice conditions
on advanced instrumentalists’ performance accuracy. Journal of Research in Music Education,36,
Ross, S. L. (1985). The effectiveness of mental practice in improving the performance of college trom-
bonists. Journal of Research in Music Education,33, 221–230.
Rubin-Rabson, G. (1941a). Mental and keyboard overlearning in memorizing piano music. Journal of
Musicology,3, 33–40.
Rubin-Rabson, G. (1941b). Studies in the psychology of memorizing piano music: V. A comparison of
pre-study periods of varied length. Journal of Educational Psychology,32, 101–112.
Rubin-Rabson, G. (1941c). Studies in the psychology of memorizing piano music: VI. A comparison of
two forms of mental rehearsal and keyboard overlearning. Journal of Educational Psychology,32,
Theiler, A. M., & Lippman, L. G. (1995). Effects of mental practice and modeling on guitar and vocal
performance. Journal of General Psychology,122, 329–343.
Wagner, M. (1975). The effect of a practice report of practice time on musical performance. In C. K.
Madsen, R. D. Greer, & C. H. Madsen, Jr. (Eds.), Research in music behavior: Modifying music
behavior in the classroom, pp. 125-130. New York: Teachers College Press.
320 Journal of Research in Music Education
by Robert Duke on October 31, 2009 http://jrm.sagepub.comDownloaded from
Williamon, A., & Valentine, E. (2000). Quantity and quality of musical practice as predictors of perfor-
mance quality. British Journal of Psychology,91, 353–376.
Williamon, A., Valentine, E., & Valentine, J. (2002). Shifting the focus of attention between levels of
musical structure. European Journal of Cognitive Psychology,14, 493–520.
Robert A. Duke is the Marlene and Morton Meyerson Centennial Professor in Music and Human
Learning and director of the Center for Music Learning in the Sarah and Ernest Butler School of Music
at The University of Texas at Austin. His research interests include motor learning, procedural memory
consolidation, and the development of expertise.
Amy L. Simmons is assistant professor of music education and a member of the Institute for Music
Research at The University of Texas at San Antonio. Her research interests include procedural memory
consolidation, motor skill development, and preservice teacher preparation.
Carla Davis Cash is assistant professor of piano and piano pedagogy in the School of Music at Texas
Tech University. Her research interests include motor skill development and procedural memory consol-
idation in musicians.
Submitted December 10, 2007, accepted October 28, 2008.
Duke et al. / Characteristics of Practice 321
by Robert Duke on October 31, 2009 http://jrm.sagepub.comDownloaded from
... Given the nature of choir singing, it is also possible that, in the present study, those with a longer experience were not more skilled singers. Consistent with these notions, Duke, Simmons, and Cash (2009) showed that, among a sample of 17 adult pianists, the strategy employed during practice was more predictive of their retention test performances than was how much or how long they practiced. In sum, the present finding suggests that, to reach their optimal transfer potential, singing activities would not have to last for decades. ...
Full-text available
The notion that lifestyle factors, such as music-making activities, can affect cognitive functioning and reduce cognitive decline in aging is often referred to as the mental exercise hypothesis. One ubiquitous musical activity is choir singing. Like other musical activities, singing is hypothesized to impact cognitive and especially executive functions. Despite the commonness of choir singing, little is known about the extent to which singing can affect cognition in adulthood. In this cross-sectional group study, we examined the relationship between age and four auditory executive functions to test hypotheses about the relationship between the level of mental activity and cognitive functioning. We also examined pitch discrimination capabilities. A non-probabilistic sample of 147 cognitively healthy adults was recruited, which included 75 non-singers (mean age 52.5 ± 20.3; 20–98 years) and 72 singers (mean age 55.5 ± 19.2; 21–87 years). Tests of selective attention, processing speed, inhibitory control, and working memory were administered to all participants. Our main hypothesis was that executive functions and age would be negatively correlated, and that this relationship would be stronger in non-singers than singers, consistent with the differential preservation hypothesis. The alternative hypothesis - preserved differentiation – predicts that the difference between singers and non-singers in executive functions is unaffected by age. Our results reveal a detrimental effect of age on processing speed, selective attention, inhibitory control and working memory. The effect of singing was comparatively more limited, being positively associated only with frequency discrimination, processing speed, and, to some extent, inhibitory control. Evidence of differential preservation was limited to processing speed. We also found a circumscribed positive impact of age of onset and a negative impact of singing experience on cognitive functioning in singers. Together, these findings were interpreted as reflecting an age-related decline in executive function in cognitively healthy adults, with specific and limited positive impacts of singing, consistent with the preserved differentiation hypothesis, but not with the differential preservation hypothesis.
... The improvement in the performance of the quaver beat by the experimental group, after applying the analytic strategy, may support studies that show that conscious awareness and attention are required for sensorimotor synchronization in time changes (Repp and Keller 2004). Therefore, improving rehearsal strategy is a higher priority than spending more time on rehearsals, as shown by the results of studies that found that the strategies used during piano practice have a greater effect on the quality of the performance than the time pianists spend practicing (Duke, Simmons, and Cash 2009). Consequently, teachers must invest in their teacher training and the strategies they use in their teaching practice since, just as musicians must learn to play, teachers must learn to teach (Robinson 2012) using the analytic strategy and rhythmic education. ...
Full-text available
Flamenco is widely practiced around the world, both in its musical and dance forms. It has been taught and studied through formal, non-formal, and informal teaching processes. Traditionally, it is taught through imitation and repetition and learned by ear. Zapateado is the percussive element of the dance. This study aimed to investigate whether a change in teaching strategy and resources could improve motivation and rhythmic precision when performing zapateado. The research used a quasi-experimental approach. Twelve professional flamenco dancers participated. The most significant results show that a change in methodology can encourage motivation and improve the rhythmic precision of zapateado.
... Intuitively, it would seem that practice would be highly linked with success and comfort in both lessons and practice. However, using practice quantity as indicative of practice success may be a flawed premise (Duke et al., 2009;McPherson & Renwick, 2001;Williamon & Valentine, 2000). One wonders whether a measure of practice quality, which might also be associated with the self-perception of competence, might yield more information on this front. ...
Full-text available
Solo performance is a common experience for children learning to play an instrument, yet the research literature on these experiences is limited, with a focus on older children and adolescents. The purpose of this study was to examine younger children’s feelings about performance over the course of a year of study. Forty-one children were interviewed about their piano lessons and performance experiences at the end of two consecutive semesters of study. They also responded to a pictorial scale on their feelings about performance at each interview and again at two piano recitals. Results indicate that children are remarkably consistent in their feelings about performing in piano recitals, with few significant changes over time and context. Correlation analyses indicate changes in the relationships between feelings about performance and certain study variables over time—in particular age, liking of lessons, liking of performing, practice time, and perception of being good at piano. In the fall term, gender and age are significant predictors of feelings about performance, with younger children and boys feeling most positive. In the spring, the findings shift and the only significant predictor is children’s liking of piano lessons. Implications and directions for further research are discussed.
The purpose of this study was to examine graduate instrumentalists’ practice habits, motivation orientations, and the impact of COVID-19 pandemic on their instrumental practice. Data were collected for this qualitative descriptive study through four participants’ videotaped practice recordings ( N = 4), video-stimulated recall ( n = 4), semi-structured interviews ( n = 4), and a questionnaire ( n = 4) about practice strategies adopted and motivation orientations. Four themes were emergent including (a) task-oriented practice routines, (b) solution-oriented approaches, (c) mixed motivations, and (d) challenges faced during COVID-19 pandemic. Results indicated that the COVID-19 pandemic negatively impacted graduate instrumentalists’ practice progress and psychological state. Their experience could be a reference for other instrumentalists in applying effective practice strategies and achieving strong and independent musicianship. The findings of this study also suggested that instrumental instructors may consider encouraging their students to flexibly plan practice routines and give priority to short-term goals.
Who is in charge of rehearsing music? Traditionally, it has been teachers who guide the music learning, but the National Core Arts Standards ask that students be the ones who identify and select rehearsal strategies to refine musical performances. However, these standards do not define what rehearsal strategies are, so this article digs into research on practicing and rehearsing to create a set of rehearsal strategies applicable to a general music classroom.
Throughout the learning process, it is common for students to transition to a new teacher, whether they relocate, decide to change teachers, or study at a summer program. During this transition, students must adapt to a potentially different pedagogical philosophy, language, and performance expectations. Using a multiple case study research design, we explored the perceptions of four adolescent violin students who experienced this transitional process while studying with Mimi Zweig at the four-week Indiana University Summer String Academy. We analyzed the interview data, observations, and external data sources through the lens of attribution theory. Results indicated that the participants not only acquired greater self-efficacy and motivation after studying with a new teacher but also endured performance anxiety. These findings inform studio instrumental instructors of the challenges students experience when studying with a new teacher and the physiological and emotional challenges that often accompany change.
This chapter goes beyond the boundaries of the classroom and emphasizes musical learning that also occurs in non-formal and informal educational settings. Based on cases that illustrate the comparison between 3 cultures of instrumental learning: a) the academic sphere of classical music of Western European tradition, b) at the opposite pole, the learning of flamenco in communities of oral tradition, and c) the culture of jazz learning halfway between the two poles, where the jam-sessions and the musical groups show structures with different intentions and regulation of the apprentices. Different types of learning and cognitive, bodily and technological tools will be shown according to the forms of musical production characteristic of each learning culture.
Practice occupies an essential and significant place in every musician’s life, yet the practice habits of highly experienced, professional musicians, remain under-researched. As such, this study sought to contribute to the emerging body of knowledge into professional musicians’ practice. Semi-structured interviews were undertaken with six such musicians, and through the qualitative content analysis of the resulting transcripts, and an evaluation of the findings in relation to current literature, insights into professional musicians’ practice time, practice strategies, and motivation were uncovered. The analysis of participant transcripts revealed that, due to greater work and family responsibilities, as well as greater efficiency, the participants spend significantly less time practicing now as professional musicians than they did as university students. The practice strategies reported by participants to be the most helpful in their practice were listening to recordings, slow practice, and employing a mindset of focus and self-awareness. The analysis also revealed that each participant reported intrinsic motivation towards practicing, and used the strategies of taking breaks, and simply picking up their instruments, to increase their motivation. The findings from this study are useful to those who might seek to one day be a professional musician, as well as to professional musicians who might find it interesting to know how their own practice habits align with those of other professional musicians.
The purpose of this study was to investigate undergraduate choral conducting students’ conducting practice approaches, time allocation, and beliefs. A secondary purpose was to examine conducting instructors’ pedagogy and beliefs about their students’ practice approaches. Undergraduate students enrolled in an introductory choral conducting class ( N = 126) and their instructors ( N = 12) responded to separate questionnaires. Students and instructors agreed that technical skills came more naturally than expressive skills, and students reported practicing technical skills more often than expressive skills. Although instructors and students agreed to the frequent use of “singing or humming musical lines” as a practice strategy, they did not align regarding the importance of using video recordings within practice sessions. Based on our results, we recommend that instructors explicitly share their practice expectations with students and introduce an incremental approach to developing expressive conducting skills earlier in the curriculum.
The process of internationalizing a music appreciation class is discussed. The role of music appreciation in an internationalized curriculum is examined and a philosophy of music education is developed. Curriculum design, assessment practices, and teaching and learning activities are viewed through the framework of Fink's Taxonomy of Learning, including descriptions of specific practices and assignments and their relationship to the taxonomy of learning the philosophy of music education. Special consideration is given to the power of music, the arts, and aesthetic experiences to cultivate knowledge of the self and of the other.
Full-text available
A concert pianist the second author videotaped herself learning J.S. Bach's Italian Concerto Presto , and commented on the problems she encountered as she practised. Approximately two years later the pianist wrote out the first page of the score from memory. The pianist's verbal reports indicated that in the early sessions she identified and memorised the formal structure of the piece, and in the later sessions she practised using this organisation to retrieve the memory cues that controlled her playing. The practice and recall data supported this account. Both were organised by the formal structure of the music. Practice segments were more likely to start and stop at boundaries of the formal structure than at other locations, and recall was higher for the beginnings of sections than for later portions. Like other forms of expert memory, pianistic memory appears to be based on use of a highly practised retrieval scheme which permits rapid retrieval of information from long-term memory.
Full-text available
Experts in many fields approach a new problem by identifying the general principles involved before starting work on details. Do expert musicians similarly begin work on a new piece with the big picture, an artistic image of the piece, in mind? To find out, a concert pianist recorded her practice of the third movement, Presto, of J1. S. Bach's Italian Concerto, commenting as she did so about what she was doing. The behavioral record of where playing started, stopped, and slowed down indicated the musical dimensions affecting practice, while the comments indicated the main focus of the pianist's attention. An artistic image for the piece was already evident in the initial sight-read performance, guided work on technique in sessions 1-6, and was transformed into a plan for performance by practice of performance cues in sessions 7-8. Interpretive details were added in sessions 9-10 and remaining problems touched up in session 11-12. Despite its pervasive effects on practice, the pianist's artistic image was mentioned only indirectly in comments about technique in sessions 1-6 and about structure, memory, and interpretation in later sessions.
The theoretical framework presented in this article explains expert performance as the end result of individuals' prolonged efforts to improve performance while negotiating motivational and external constraints. In most domains of expertise, individuals begin in their childhood a regimen of effortful activities (deliberate practice) designed to optimize improvement. Individual differences, even among elite performers, are closely related to assessed amounts of deliberate practice. Many characteristics once believed to reflect innate talent are actually the result of intense practice extended for a minimum of 10 years. Analysis of expert performance provides unique evidence on the potential and limits of extreme environmental adaptation and learning.
While it is known that repetition is a common strategy employed in the practice sessions of all musicians, very little is known about the extent to which musicians of high proficiency levels use repetition in their practice. The purpose of this study was to describe the role of repetition in the practice activities of musicians representing four levels of high musical proficiency. Specifically, this study examined in detail the process used to isolate, remediate, and recontextualize music in the subjects' practice sessions, and examined the number of consecutive performances of passages selected for repeated practice. Nineteen subjects were videotaped on two separate occasions practicing for a minimum of one hour. One extended segment during which each subject practiced a single work was selected for detailed analysis. Data were recorded graphically. The total number and mean durations of practice frames and performance trials per frame were then calculated. Based on 14 hours 54 minutes of practice time, 800 practice frames, and 8,527 performance trials, the results illustrate the extensive use of repetition in advanced performers' practice. The mean number of performance trials devoted to each target passage was 10.7, the mean duration of all performance trials was 6.3 seconds, and the mean duration of each practice frame was 1 minute 7 seconds.
This study examined the relative effectiveness of mental practice in improving trombone performance. The experiment involved 30 trombonists from three colleges. They were randomly assigned to one of five experimental practice conditions: (1) all physical practice, (2) all mental practice, (3) a combination of physical and mental practice, (4) mental practice with simulated slide movement, and (5) no practice (control). A short étude served as both a pre- and posttest. A one-way analysis of covariance and subsequent Scheffé's Test found three pairs of groups significantly different from each other: (1) combined practice and no practice, (2) combined practice and mental practice, and (3) physical practice and no practice. The results of the study suggest that the use of mental with physical practice can accrue many benefits for the skilled trombonist.
This paper examines the practice and performance of musicians in two separate studies. In the first, 22 pianists, classified into four levels of skill, were asked to learn and memorise an assigned composition by J.S. Bach (different for each level). From cassette recordings of the pianists' practice, values for the mean and standard deviation of ''segment lengths'' were obtained, as well as a derived coefficient of variation. Following the learning process, the pianists performed their composition in a recital setting and were interviewed about their performance strategies. In the second study, two expert pianists were asked to learn and memorise an assigned composition by Bach and to provide verbal protocols of their performance of the piece under two separate conditions. Analyses of the data from Study 1 revealed that pianists at all levels of skill used longer practice segments as practice progressed. Moreover, they increasingly interspersed short and long segments across the practice process. Analyses of the coefficient of variation revealed that this occurred most often for performers at higher levels of ability, supporting the notion that shifting the focus of attention between levels of musical structure is a salient characteristic of the practice of skilled performers and develops gradually as skill is acquired. The post-performance interviews suggested that the attentional shift was employed during performance and was highly individual-specific with regard to the locations and reasons for shifting. Data from Study 2 provided additional evidence to support the proposed process and demonstrated the use of one possible hierarchical system within which shifts in attentional focus can occur—the music's formal structure. The data are discussed with regard to their implications for furthering an understanding of performance preparation in music and in other domains.
The effectiveness of various mental practice regimens in music performance was compared. Guitarists and vocalists who were college-level music majors learned short musical excerpts under four conditions: (a) continuous physical practice; (b) mental practice alternating with physical practice; (c) mental practice with a modeled recording of the music alternating with physical practice; and (d) a motivational control activity alternating with physical practice. Performances with the score and by memory were rated on various musical dimensions. Mental practice with a model resulted in superior performance on a number of dimensions for vocalists and on tonal quality and memory coding for guitarists. For guitarists, mental practice (without a model) and continuous physical practice also produced greater pitch accuracy than the control condition did. It was concluded that mental practice may facilitate cognitive coding and help to create optimal levels of attentional focus and arousal.
Two independent groups, one consisting of college-level studio music teachers (n = 127) and the other of music majors (n = 134), completed a written survey containing 10 questions about attitudes and expectations for practicing music. The questionnaire was designed to address four major areas of interest: (1) attitudes about specific music skills, (2) expectations concerning use of practice time, (3) expectations for routines and strategies for practicing, and (4) attitudes toward practice in general. Results indicated that teachers expected more weekly practice time to be taking place than was actually reported by students. Additionally, most teachers expected that students should follow a specific practice routine, but 55% of students indicated that they do not do so; and although nearly all teachers in the survey stated that they discussed practice strategies with students, 67% of students reported that practice strategies were not discussed in their studio lessons. Finally, teachers and students had widely differing views regarding their feelings about practice in general.
The purpose of this investigation was to examine the effects that modeling, self-evaluation, and self-listening have on junior high school instrumentalists' music performance and attitude about practice. The pretest/posttest 2 × 2 × 2 factorial design involved a total of 82 woodwind, brass, and percussion students in the seventh (n = 36), eighth (n = 31), or ninth grade (n = 15). Data indicated that participants who listened to a model during self-evaluation improved more than those not listening to a model in the areas of tone, melodic accuracy, rhythmic accuracy, interpretation, and overall performance, but not intonation, technique/articulation, or tempo. When self-evaluation was not undertaken, modeling groups were no different in any performance subarea. Also, the main effects for modeling revealed that groups that listened to a model improved their performance more than did students who did not listen to a model in the areas of tone, technique/articulation, rhythmic accuracy, tempo, interpretation, and overall performance but not intonation or melodic accuracy. No statistically significant findings for self-listening or practice attitude were identified.