Content uploaded by Amy Fancourt
Author content
All content in this area was uploaded by Amy Fancourt
Content may be subject to copyright.
Pitch-Change Detection and Pitch-Direction Discrimination in Children
Amy Fancourt, Frederic Dick, and Lauren Stewart
University of London
The present study investigated developmental changes in the ability to detect a change in pitch and to
discriminate the direction of a pitch change using pitch glides. Adaptive-tracking procedures established
separate thresholds for both of these abilities in musically untrained participants across nine age-groups
(5-, 6-, 7-, 8-, 9-, 10-, 11-, 13-year-olds, and adults). The use of an odd-one-out paradigm avoided the
need for participants to use semantic labels when determining the direction of a pitch change, and
screening of the adaptive-staircase profile plots permitted the exclusion of inattentive performers.
Although adults achieved equivalent thresholds for pitch-change detection and pitch-direction discrim-
ination, there were age-related improvements for pitch-direction discrimination but not pitch-change
detection in children between the ages of 6 and 11 years. The findings may indicate that the capacities
to detect a change in pitch versus to discriminate the direction of a pitch change follow different
developmental trajectories.
Keywords: pitch processing, pitch-direction discrimination, pitch-change detection
Supplemental materials: http://dx.doi.org/10.1037/a0033301.supp
Pitch is “that property of a sound that enables it to be ordered on
scale going from low to high” (Acoustical Society of America
Standard Acoustical Terminology, cf. Randel, 1978). It is a central
organizing feature of the majority of the world’s musical systems
(McDermott & Oxenham, 2008), and the capacity to distinguish
variations in pitch is critical to the perception and appreciation of
music (Patel, 2008). The successful perception of Western music
requires the capacity to detect a change in pitch of one semitone at
a minimum. Equally important is the ability to encode the direction
of a change in pitch, as this can be viewed as a building block of
musical “contour.” Contour is the structural property that describes
a melody’s shape, and as such is one of the key features along
which melodies can be defined, classified, constructed, and re-
membered (Dowling & Fujitani, 1971; Dowling & Bartlett, 1981;
Dowling & Harwood, 1986; Dowling, Kwak, & Andrews, 1995).
For example, when asked to detect a change between two novel
melodies, accuracy is much poorer when the second melody is
transposed in pitch but the contour is unchanged (Bartlett &
Dowling, 1980).
Several strands of evidence suggest that the capacities to detect
a change in pitch (“pitch-change detection”) and to discriminate
the direction of a change in pitch (“pitch-direction discrimination”)
are functionally and anatomically separable in the adult (Stewart,
Kriegstein, Warren, & Griffiths, 2006). Impairments in pitch-
change detection are associated with damage to subcortical struc-
tures, ascending auditory pathways, and primary auditory cortex
(PAC) in medial Heschl’s Gyrus (HG) (Habib et al., 1995; Hat-
tiangadi et al., 2005; Terao et al., 2006; Tramo, Shah, & Braida,
2002), whereas impairments in pitch-direction discrimination have
been associated with lesions to lateral HG in the right hemisphere
(Johnsrude, Penhune, & Zatorre, 2000; Terao et al., 2006; Tramo
et al., 2002). Patients with temporal lobe excisions that encroach
on anterolateral portions of HG in the right hemisphere demon-
strate impaired pitch-direction discrimination and yet retain the
capacity to detect equivalently small changes in pitch (Johnsrude
et al., 2000; Tramo et al., 2002). A similar pattern of performance
is observed in individuals with congenital amusia, where thresh-
olds for pitch-direction discrimination are often markedly higher
than those for pitch-change detection (Foxton, Dean, Gee, Peretz,
& Griffiths, 2004; Liu, Patel, Fourcin, & Stewart, 2010; William-
son & Stewart, 2010).
Dissociations in pitch-change detection and pitch-direction dis-
crimination have been observed in some adult listeners (Mathias,
Micheyl, & Bailey, 2010; Semal & Demany, 2006). Semal and
Demany (2006) measured discrimination thresholds for pitch-
change detection and pitch-direction discrimination in nine adults
without any obvious audiological or neurological problems. They
found that three of the adults showed significantly higher thresh-
olds for pitch-direction discrimination as compared with pitch-
change detection. It should be noted here that although all partic-
This article was published Online First June 17, 2013.
Amy Fancourt, Department of Psychology, Goldsmiths, University of
London, London, United Kingdom; Frederic Dick, Department of Psycho-
logical Sciences, Birkbeck, University of London; Lauren Stewart, Depart-
ment of Psychology, Goldsmiths, University of London.
We thank Sukhbinder Kumar for providing the scripts to run the adap-
tive tracking procedure. We are very grateful to The Coombes and
Crowthorne Church of England Primary Schools for their help in identi-
fying participants and to the children and their parents for their participa-
tion. In addition, we are grateful to the anonymous reviewers for their
helpful comments on earlier drafts.
A photo and brief biography of the corresponding author is available at
http://dx.doi.org/10.1037/a0033301.supp
Correspondence concerning this article should be addressed to Amy
Fancourt, Department of Psychology, Goldsmiths, University of London,
New Cross, London SE14 6NW, United Kingdom. E-mail: a.fancourt@
gold.ac.uk
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Psychomusicology: Music, Mind, and Brain © 2013 American Psychological Association
2013, Vol. 23, No. 2, 73–81 0275-3987/13/$12.00 DOI: 10.1037/a0033301
73
ipants in the Semal and Demany (2006) study were classified as
“typical,” there was no formal screening for congenital amusia,
and it is possible that these individuals may have fulfilled the
criteria for congenital amusia, had they been formally tested.
The findings to date seem to indicate that pitch-change detection
and pitch-direction discrimination are separable processes in the
adult. Neuroimaging work has shown that the capacity to detect a
change in pitch may be reliant on sensory processing mechanisms
at a subcortical level and in the PAC, whereas the capacity to
discriminate the direction of a pitch change appears to require
higher order perceptual processing taking place in cortical regions
beyond the PAC (Griffiths, 2003; Johnsrude et al., 2000; Patterson,
Uppenkamp, Johnsrude, & Griffiths, 2002).
The observation that simple sensory discrimination and higher
order pitch perception may rely on differentially weighted neural
substrates could mean that these two capacities also have different
developmental trajectories. Several studies have separately inves-
tigated developmental changes in pitch-change detection (Jensen
& Neff, 1993; Maxon & Hochburg, 1982; Olsho, 1984; Olsho,
Koch, & Halpern, 1987; Olsho, Schoon, Sakai, Turpin, & Sper-
duto, 1982; Soderquist & Moore, 1970; Thompson, Cranford, &
Hoyer, 1999; Trehub, Cohen, Thorpe, & Morrongiello, 1986) and
pitch-direction discrimination (Stalinski, Schellenberg, & Trehub,
2008; Van Zee, 1976; White, Dale, & Carlsen, 1990). Jensen and
Neff (1993) investigated the development of a range of auditory
abilities in preschool children and compared performance on these
measures with that of a group of adult controls. They found a
sequential pattern of development for intensity, frequency, and
duration discrimination, with early maturation of intensity discrim-
ination followed by frequency discrimination and finally duration
discrimination. Jensen and Neff (1993) found that frequency dis-
crimination continues to improve up to the age of around 6 years
at which point adult-like competence is shown. Thompson et al.
(1999) observed a slightly longer maturational course for brief-
tone discrimination and found that performance became adult-like
when children reached 7 years. Taken together, both of these
studies indicate that frequency discrimination is fully mature by
the age of around 6 to 7 years.
Developmental investigations of pitch-direction discrimination
have demonstrated that infants and young children are able to
discriminate the direction of a change in pitch (Trehub, Bull, &
Thorpe, 1984; Trehub, Thorpe, & Morrongiello, 1985) and that the
sensitivity of such discriminations continues to show improvement
over age (Stalinski et al., 2008). White et al. (1990) measured
pitch-direction discrimination in preschool children aged between
3 and 5 years of age and found that the majority of 5-year-olds
were able to make discriminations of pitch on the basis of direc-
tion. Stalinski et al. (2008) investigated pitch-direction discrimi-
nation in musically untrained participants aged 5 years to 11 years
and found that while performance improved with age, children as
young as 5 years were able to make directional discriminations of
changes in pitch. These authors found that the majority of 6-year-
old participants performed above chance when discriminating the
direction of pitch changes of 0.5 semitones (ST), whereas less than
half of the 6-year-olds performed above chance for discriminations
of 0.3 ST, and a minority showed above chance performance for
discriminations of 0.1 ST. For the 8-year-old and 11-year-old
groups, the majority of participants performed above chance for
directional discriminations of 0.3 ST, with approximately 33%
able to make discriminations for changes in pitch of 0.1 ST.
Stalinski et al. (2008) found that a greater number of 11-year-olds
were able to accurately discriminate the direction of fine-grained
changes in pitch of around 0.3 ST than 8-year-olds. This finding
indicates that the sensitivity with which fine-grained discrimina-
tions of pitch direction are made may continue to show improve-
ment up until the age of around 11 years.
Despite this body of research, to our knowledge, only a single
study has compared the developmental trajectories of pitch-change
detection and pitch-direction discrimination using a within-
subjects design. Cooper (1994) measured the capacity to detect and
discriminate changes in pitch smaller than a semitone in children
aged between 6 and 11 years. Participants were presented with a
5-tone sequence, in which a target tone differed from the standard
tone by a variable interval (between 0.06 and 1.15 ST). Partici-
pants were required to indicate via pencil and paper which of five
dots corresponding to the five tones in the sequence differed in
pitch and then to determine whether the tone that differed was
higher or lower. Accuracy was higher for all age-groups for
pitch-change detection as compared with pitch-direction discrim-
ination, and age-related improvements were seen for both tasks.
Although informative, the conclusions that can be drawn from the
Cooper (1994) study, concerning the extent to which pitch-change
detection versus pitch-direction discrimination may follow differ-
ent developmental trajectories, are somewhat limited by certain
features of the design. In particular, while the 30 test items in-
cluded a range of pitch intervals, between 0.06 and 1.15 ST, the
total score represents an average across all pitch interval levels.
Such a method makes it impossible to ascertain whether age-group
differences in performance accuracy varied according to the size of
the interval; this limits the sensitivity with which any developmen-
tal changes in the capacity to discriminate changes in pitch can be
demonstrated. In addition, the task required the participants to
associate a relational concept (high or low) to a change in pitch
direction. Young children have been shown to have difficulty in
understanding concepts such a high and low in relation to pitch
(Andrews & Madeira, 1977); hence, a task which requires the use
of such labels might underestimate pitch-direction discrimination
abilities in younger children.
To address these afore-mentioned limitations, the present study
used an “odd-one-out” procedure and psychophysical paradigm to
measure separate thresholds for pitch-change detection and pitch-
direction discrimination. The measurement of thresholds provides
an index of perceptual ability while the use of an odd-one-out
procedure enables the discrimination of a change in pitch without
the need to explicitly name the spatial relationships between tones
(e.g., “higher” and “lower”). In the pitch-change detection task,
participants were presented with three tones, two steady-state tones
and one pitch-glide, and were instructed to identify the “odd-one-
out.” Similarly, in the pitch-direction discrimination task, partici-
pants were presented with three pitch glides, two of them going in
one direction (e.g., ascending) and one of them going in the
opposite direction (e.g., descending) and again were instructed to
identify the “odd-one-out.” Previous studies have indicated that in
typical adults the use of continuous as opposed to discrete pitch
changes does not influence detection thresholds (Demany, Cary-
lon, & Semel, 2009; Foxton et al., 2004; Williamson & Stewart,
2010), and gliding stimuli have been successfully used to measure
pitch-direction discrimination in adults and in children with lan-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
74
FANCOURT, DICK, AND STEWART
guage difficulties (Bishop, Adams, Nation, & Rosen, 2005; Wil-
liamson, Liu, Peryer, Grierson, & Stewart, 2012). Pitch glides
were used in the present study for both pitch-change detection and
pitch-direction discrimination to minimize the memory demands
and reduce the influence of nonauditory factors on task perfor-
mance.
Method
Participants
A total of 130 children were recruited from two primary schools
in the south-east of England. The children were categorised into
the following age-groups: 13 five-year-old children (ages in years;
months—mean: 4.11, range: 4.7–5.5), 20 six-year-old children
(mean: 5.11, range: 5.7– 6.5), 19 seven-year-old children (mean:
7.0, range: 6.7–7.5), 23 eight-year-old children (mean: 8.0, range:
7.7– 8.5), 18 nine-year-old children (mean 9.2, range: 8.9 –9.6), 20
ten-year-old children (mean: 10.2, range: 9.8–10.5), 13 eleven-
year-old children (mean: 11.1, range: 10.7–11.6), and 4 thirteen-
year-old children (mean 13.0, range 13.0 –13.1). Thirteen adults
also participated (mean age: 30, range: 18 –41).
All participants spoke English as their primary language, had
not received any prior musical training and had received no prior
diagnoses of audiological, neurological, or educational problems.
Informed consent was obtained from the parents of all children
who were under 18 years and from the adult participants. Ethical
approval was obtained from the local research ethics committee at
Goldsmiths, University of London.
Design and Procedure
Pitch-change detection and pitch-direction discrimination.
Thresholds were measured for the detection of a change in pitch
(pitch-change detection task) and the discrimination of the direc-
tion of change in pitch (pitch-direction discrimination task) using
an AXB paradigm—a three-interval two-alternative forced-choice
procedure in which the listener is presented with three intervals
and has to determine whether the first or third interval is the same
as the second. With this type of paradigm, there is no need to label
the stimuli or remember associations and the task can be done by
listening for and rejecting the “odd-one-out.” The stimuli were
modeled after those used to measure pitch discrimination in adult
congenital amusics (Foxton et al., 2004; Williamson & Stewart,
2010). Each trial consisted of three sounds that were 600 ms in
duration, with a 600 ms interstimulus interval (ISI). For the pitch-
change detection task, the target was a frequency glide (250 ms
steady-state onset, 100 ms glide, and 250 ms steady-state offset),
logarithmically centered on 500 Hz (log2(f/500) ⫽ (dE/(12 ⫻ T))
⫻ (t ⫺ T/2)), while nontargets were steady-state tones of 500 Hz
of the same length. A 10-ms window was used to smooth both the
onset and offset of the stimuli. The window was cosine in shape.
For the pitch-direction discrimination task, both targets and non-
targets were frequency glides (250 ms steady-state onset, 100 ms
glide, and 250 ms steady-state offset), again logarithmically cen-
tered on 500 Hz. In the pitch-direction discrimination task, the
target was the glide with a frequency excursion in the opposite
direction to the nontarget glides. In both the pitch-change detection
and pitch-direction discrimination tasks, targets randomly occurred
with equal probability in the first or last position. Participants were
required to indicate via verbal response whether the “first” or
“last” interval was the odd-one-out and the experimenter made the
appropriate response on the computer keyboard.
Both tasks incorporated 18 levels of difficulty, where each level
corresponded to a particular frequency excursion of the target
stimulus. The initial starting level was a 10-semitone glide, allow-
ing participants to familiarize themselves with the task. After two
consecutive correct responses at this level, participants moved to
the next level of the task. Following a single incorrect response,
participants reverted to the previous level of the task (known as a
“reversal”), and so on. The target stimulus appeared in either the
first or last position at random. In addition, trials in which the
target was either positive or negative in its frequency glide were
randomly interleaved across single block, so that separate thresh-
olds could be established for upward and downward targets. The
procedure terminated following 10 reversals for each target type,
and separate thresholds were calculated for both. Thresholds cor-
responded to the median value of the target stimulus for the last
five reversals. To check the reliability of the estimated thresholds,
we also calculated the mean value of the target stimulus for the last
six reversals and found that this made little difference to the
threshold estimates. The frequency glide of the target at each of the
18 levels were (in ST) as follows: 10, 8, 6, 4, 2, 1, 0.9, 0.8, 0.7, 0.6,
0.5, 0.4, 0.3, 0.2, 0.1, 0.075, 0.05, 0.025, 0.
1
The pitch-change detection and pitch-direction discrimination
tasks were administered to all participants. The pitch-change de-
tection task was always conducted first, followed by the pitch-
direction discrimination task. While counterbalancing of tasks
would have been a priority in an adult-only study, the priority for
the present study was to ensure that as many children as possible
comprehended the task requirements and so the pitch-change de-
tection task (predicted to be the easier of the tasks to grasp) was
always run first.
Testing was conducted in a quiet room within the children’s
school environment, using a laptop computer. Sound delivery was
via an external sound card (Edirol UA-4FX), and Sennheiser
HD265 headphones and volume was set at a comfortable listening
level. Programs for stimulus presentation and the collection of data
were written in Matlab
®
. Each of the tasks took 10 minutes to
complete. A 5-min break between tasks was included to prevent
fatigue, and participants were given stickers to maintain their
interest.
To ensure that all participants understood the concepts involved
in the pitch-change detection and pitch-direction discrimination
tasks, the experimenter first engaged in a short interactive “game”
with all participants under 11 years old. First, to introduce the
concept of direction, the experimenter showed the participant
pictures relating to the concepts of “up” and “down” (e.g., smoke
rising; a child descending a slide in a playground). Once the
experimenter was confident that this concept was clear, the idea of
identifying an “odd-one-out” was introduced. The participant was
1
Unequal step sizes were used to enable participants to progress quickly
through the task to their near-threshold level. Although the use of unequal
step sizes may have resulted in some difficulty calculating exact thresholds
if reversals occurred around one of the unequal changes (e.g., 0.2, 0.1,
0.075), for the majority of participants, reversals did not occur around an
unequal change.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
75
PITCH DISCRIMINATION
shown three pictures in which two of the pictures conveyed the
same direction, and one conveyed the opposite direction. The
participant was asked to identify the “odd-one-out.” Following
successful completion of four examples (two upward, two down-
ward), the experimenter discussed the concept of direction as
applied to sounds. Both upward and downward glides, as well as
steady-state tones, were played to illustrate this. Following this
introductory session, the pitch-change detection and pitch-
direction discrimination tasks were run in fixed order. Prior to each
task, instructions were verbally provided, and four practice trials
were administered, during which the participants were given feed-
back on their performance.
One issue to consider with the use of psychophysical adaptive
staircase procedures with children is that children are prone to
distraction in such tasks, which can lead to unreliable thresholds
(Horváth, Czigler, Birkas, Winkler, & Gervais, 2009; Wetzel &
Schroeger, 2007). In tasks where participants are presented with a
series of trials on which to make perceptual judgments, the per-
formance of the participant across all trials can be visualized as a
“track.” In typical adult listeners, tracks will show a steady step-
wise reduction in stimulus level and a series of oscillations toward
the end of the track, when performance is approaching threshold.
However, in cases where participants are not focused on the task,
tracks will show a large degree of variability in the reversal points.
Although such patterns of inattentive performance can be occa-
sionally exhibited by adults, children are on the whole more prone
to distraction, and the impact of inattention on task performance is
the most marked for younger children (Moore, Ferguson, Halliday,
& Riley, 2008). Therefore, in the present study, the individual
performance tracks were plotted to screen for inattentive perfor-
mance.
Shapiro–Wilk normality tests and visual inspection indicated
significant deviations from normal distributions for the age-group
thresholds for both pitch-change detection and pitch-direction dis-
crimination tasks. Thresholds were log transformed to reduce the
impact of a positive skew on subsequent analyses (Field, 2005).
Paired sample t tests of the log-transformed data revealed that
there were no systematic differences in threshold according to
whether the direction of the target was upward or downward, for
any age-group on either task. Therefore, a mean threshold was
calculated across both upward and downward targets for the pitch-
change detection and pitch-direction discrimination tasks for each
subject. This mean threshold was used in all subsequent analyses.
Owing to inequality in the sample sizes and population vari-
ances (as indicated by Levene’s tests), the p values for within-
subject effects were Greenhouse–Geisser corrected (Geisser &
Greenhouse, 1958), and the between-subjects ANOVA results
were supplemented with Welch tests followed by Games–Howell
post hoc procedures. The Games–Howell procedure was used for
multiple post hoc comparisons because it is accurate when both the
sample size and population variances are unequal (Howell, 1997).
Results
Performance Profile Screening
All children completed the pitch-change detection task, while
four 5-year-olds, one 6-year-old and one 9-year-old failed to
complete the pitch-direction discrimination task.
To screen for inattentive performers, the visual tracks for each
task were closely inspected on an individual basis and were clas-
sified by two independent raters who were blind to the hypothesis.
Tracks from the adult participants indicated no difficulties with
maintaining attention, but the children’s performance profiles were
variable. Following Moore et al. (2008), three patterns of perfor-
mance could be observed. The most common was a “good per-
former” (Figure 1A), a pattern characterized by a succession of
correct responses, leading to a rapid approach to a near-threshold
level. A second pattern of performance, termed “compliant” (Fig-
ure 1B), was characterized by cyclical variations in performance
but with evidence of a steady progression toward threshold. The
third type of track observed was “noncompliant” (Figure 1C) and
was most frequently found in younger children, particularly on the
pitch-direction discrimination task. The noncompliant performers
were at floor level and typically did not progress from the first two
levels of the task.
To check for interrater reliability of the track classification, we
correlated the number of “good,” “compliant,” and “noncompli-
ant” performers identified by each rater for each age-group. These
correlational analyses demonstrated that there was good interrater
reliability for the classification of “good” performers (r ⫽ .908,
p ⬍ .0001), “compliant” performers (r ⫽ .751, p ⬍ .01), and
“noncompliant” performers (r ⫽ 1, p ⬍ .0000) across age-groups
for the pitch-change detection task. There was also good interrater
reliability for the classification of “good” performers (r ⫽ .896,
p ⬍ .001), “compliant” performers (r ⫽ .815, p ⬍ .01), and
“noncompliant” performers (r ⫽ 1, p ⬍ .000) across age-groups
for the pitch-direction discrimination task.
“Noncompliant” performance may be owing to fluctuating at-
tention (Moore et al., 2008), or a failure to comprehend the task,
and so all participants with “noncompliant” tracks were removed
before the main analyses. This screening process resulted in the
exclusion of five data sets for the pitch-change detection task and
eight data sets for the pitch-direction discrimination task. The data
from the 5-year-old group was not included in any further cate-
gorical analysis, as 8 out of 13 children in this group either failed
to complete both threshold tasks or were excluded from the anal-
ysis for having “noncompliant” tracks.
Following this classification, we investigated whether the pro-
portion of participants classified as “good performers” or “com-
pliant performers” differed across the pitch-change detection and
pitch-direction discrimination tasks. Using a Cochran’s Q test, we
found that there was no significant difference in the classification
of performers observed for the two threshold tasks,
2
(1) ⫽ 3.596,
p ⬎ .05. Further analyses by age-group revealed that the propor-
tion of “good performers” and “compliant performers” was not
significantly different for any age-group across the pitch-change
detection and pitch-direction discrimination tasks (see Table 1).
Of those listeners excluded for having noncompliant tracks on
the pitch-direction discrimination task, six of the eight participants
with “noncompliant” tracks for the pitch-direction discrimination
task showed markedly higher mean thresholds for pitch-direction
discrimination as compared with pitch-change detection. It may be
that the noncompliant performers on the pitch-direction discrimi-
nation task show this particular profile as the result of inattention
(Moore et al., 2008), alternatively it may reflect a selective deficit
in the capacity to discriminate the direction of changes in pitch as
seen in adults with congenital amusia (Foxton et al., 2004; Wil-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
76
FANCOURT, DICK, AND STEWART
liamson & Stewart, 2010). For the pitch-change detection task,
four out of the five participants identified as having “noncompli-
ant” tracks either also showed “noncompliant” tracks for the
pitch-direction discrimination task, or failed to complete it alto-
gether. This may indicate that these participants were unable to
meet the demands of the tasks or were simply inattentive across
tasks. Only one participant demonstrated a “noncompliant” track
for only the pitch-change detection task, which may reflect inat-
tention (Moore et al., 2008) or may be indicative of a selective
impairment in the capacity to detect changes in pitch.
Figure 1. Examples of three types of “tracks” obtained from three different participants (all aged 6 years) on
the pitch-direction discrimination task. The y-axis corresponds to the level of the task and the x-axis to the
number of trials. Track 1A was classified as a “good performer,” track 1B as a “compliant” performer, and track
1C as a “noncompliant” performer.
Table 1
Rater 1 (R1) and Rater 2 (R2) Classifications of “Good” and “Compliant” Performers for the Pitch-Change Detection and
Pitch-Direction Discrimination Tasks
Age years
Pitch-change detection Pitch-direction discrimination
Significance
Good Compliant Good Compliant
% R1 R2 % R1 R2 % R1 R2 % R1 R2
5 100 (n⫽5) (3) 0 (0) (2) 60 (3) (4) 40 (2) (1) 0.157
6 56 (10) (15) 41 (7) (4) 47 (8) (11) 53 (9) (5) 0.480
7 83 (15) (14) 17 (3) (4) 61 (11) (9) 39 (7) (9) 0.206
8 78 (18) (17) 22 (5) (6) 74 (17) (14) 26 (6) (9) 0.739
9 76 (13) (14) 24 (4) (3) 65 (11) (14) 45 (6) (3) 0.414
10 65 (13) (16) 35 (7) (4) 55 (11) (11) 45 (9) (9) 0.564
11 82 (9) (8) 18 (2) (3) 73 (8) (8) 27 (3) (3) 0.317
13 100 (4) (4) 0 (0) (0) 100 (4) (4) 0 (0) (0) N/A
18⫹ 93 (14) (15) 7 (1) (0) 93 (14) (14) 7 (1) (1) 1.000
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
77
PITCH DISCRIMINATION
Investigation of the individual differences in performance across
the two threshold tasks showed that for the pitch-change detection
task, 50% of the 6-year-olds had thresholds within the 95% con-
fidence intervals for the 11-year-olds, whereas this was not the
case for any of the 6-year-olds on the pitch-direction discrimina-
tion task.
Parametric analyses were carried out to investigate the differ-
ences in performance across age-groups as compared with an adult
control group. Owing to the small number of participants in the
5-year-old and 13-year-old age-groups for both the pitch-change
detection and pitch-direction discrimination tasks, these partici-
pants were not included in the following categorical analyses (see
Table 2).
Logarithmically transformed mean thresholds were analyzed
using a mixed-design ANOVA, with Task as a within-subject
factor (two levels: pitch-change detection and pitch-direction dis-
crimination) and Age as the between-subjects factor (seven levels:
ages 6, 7, 8, 9, 10, and 11 years, and adults).
The ANOVA analysis revealed a significant main effect of
Task, F(1, 111) ⫽ 36.877, p ⬍ .0001, Age, F(6, 111) ⫽ 7.714, p ⬍
.0001, and a significant interaction between Task and Age, F(6,
111) ⫽ 3.928, p ⬍ .005 (see Figure 2).
Further one-way ANOVAs demonstrated that the mean perfor-
mance thresholds of the different age-groups were not significantly
different on the pitch-change detection task, F(6, 48.866) ⫽ 1.948,
p ⬎ .05, but were significantly different on the pitch-direction
discrimination task, F(6, 48.308) ⫽ 12.284, p ⬍ .001. The same
pattern of results was observed when the adult control data was
removed from the analysis (pitch-change detection, F(5, 46.118) ⫽
1.480, p ⬎ .05; pitch-direction discrimination, F(5, 45.164) ⫽
7.568, p ⬍ .001). Games–Howell post hoc procedures revealed
that the 6- to 10-year-old age-groups differed significantly from
the 11-year-old and adult groups in the thresholds attained on the
pitch-direction discrimination task. However, the 11-year-old
group did not differ significantly from the adult control group on
this task (see Table 3).
To reveal potentially more subtle effects, the ages of the par-
ticipants were converted into months, allowing for treatment as a
continuous regressor as opposed to categorical variable in further
analysis. Participants in the 5-year-old and 13-year-old age-groups
were included, and regression analyses were run to investigate the
relationship between age in months and thresholds attained on the
pitch-change detection and pitch-direction discrimination tasks.
Owing to the presence of large standardized residuals in both the
pitch-change detection and pitch-direction discrimination data sets,
Theil–Sen robust regression analyses were carried out (Sen, 1968;
Theil, 1950; Wright & Field, 2009). For the pitch-change detection
task, Theil–Sen robust linear regression analysis indicated that age
was not a significant predictor of thresholds attained, ⫽⫺.002,
p ⬎ .05, R
2
⫽ .038, whereas it was found to be a significant
predictor of thresholds in the pitch-direction discrimination task,
⫽⫺.0089, p ⬍ .0001, R
2
⫽ .22, (see Figure 3
).
Discussion
The present study compared the developmental trajectories of
pitch-change detection and pitch-direction discrimination in chil-
dren aged 5 to 13 years. The overall pattern of results indicates that
fine-grained pitch-change detection is adult-like in children aged 6
to 7 years, while the sensitivity with which the direction of such
pitch changes can be discriminated does not become adult-like
until around 11 years of age. These findings are somewhat con-
sistent with those of Cooper (1994), in that the greatest age-related
improvements in pitch processing were observed for pitch-
direction discrimination. One of the difficulties with interpreting
the findings from the Cooper (1994) study is that the measures of
pitch-change detection and pitch-direction discrimination were not
independent and so any error in detecting a change in pitch would
also lead to an error in direction discrimination. The present study
investigated these two capacities independently from one another
and demonstrated that thresholds for pitch-direction discrimination
significantly improve over age whereas thresholds for detecting
equivalently small changes in pitch do not.
Previous studies have shown that pitch-change detection is
adult-like by the age of around 7 years (Jensen & Neff, 1993;
Maxon & Hochburg, 1982; Olsho, 1984; Olsho et al., 1982, 1987;
Soderquist & Moore, 1970; Thompson et al., 1999; Trehub et al.,
1986). The findings from the current study also indicate that the
capacity to detect a change in pitch does not improve significantly
through middle childhood and is already adult-like in children of
around 6 to 7 years. The interpretation of results pertains specifi-
cally to children aged ⱖ6, as the data from 5-year-olds was not
reliable enough to be included in the analysis by age-group.
Table 2
Mean Thresholds and Standard Deviations (Semitones) for the Pitch-Change Detection and Pitch-Direction Discrimination Tasks
Age years
Pitch-change detection Pitch-direction discrimination
N N
5 10 1.97 2.65 5 2.7 2.34
6 19 0.62 0.68 17 1.86 2.17
7 18 0.28 0.15 19 1.03 1.27
8 23 0.36 0.26 22 1.09 1.06
9 18 0.27 0.14 17 1.00 1.5
10 20 0.32 0.12 20 0.68 0.62
11 13 0.38 0.28 12 0.28 0.24
13 4 0.2 0.05 4 0.23 0.13
18⫹ 13 0.23 0.13 13 0.22 0.13
Total 138 129
Note. ⫽mean; ⫽standard deviation.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
78
FANCOURT, DICK, AND STEWART
In terms of pitch-direction discrimination, the findings from the
current study lend further support to earlier work demonstrating
that infants and young children are able to discriminate the direc-
tion of changes in pitch (Stalinski et al., 2008). For example, the
6-year-olds in the present study showed an average pitch-direction
discrimination threshold of around 1.9 ST, indicating that they
were able to discriminate the direction of changes in pitch as small
as a single tone. Although the findings from the present study
demonstrate that the capacity to discriminate the direction of pitch
changes is present in young children, the findings also indicate that
the sensitivity with which those discriminations can be made
continues to improve over age. In the present study, steady im-
provements in the thresholds for pitch-direction discriminations
were observed in children up to the age of around 11 years, at
which point they became similar to the thresholds observed in
adults.
We suggest that our findings cannot be explained simply on the
basis of task difficulty (which may relate to attentional/memory
demands), which are reduced with increasing age. Although no
psychoacoustic procedure can be completely free of influences
such as attention and memory, the AXB method used in the present
study minimizes the memory demands of the task and has been
successfully used to measure pitch-direction discrimination in typ-
ical children and those with impaired language (Bishop et al.,
2005). In the present study, a comparison of the performance
profile plots for the pitch-direction discrimination task and the
pitch-change detection task revealed that all participants who were
retained in the analysis showed a “steady progression toward
threshold” across both tasks. Likewise, the deliberate use of glid-
ing stimuli, to reduce memory demands, does not seem to have
posed particular difficulties for participants, as screening of the
performance plots did not reveal greater track variability for the
pitch-direction discrimination task (where all stimuli were glides)
as compared with the pitch-change detection task (where only the
target was a glide).
One issue that is pertinent to the interpretation of our study’s
findings is that the pitch-direction discrimination task always
followed the pitch-change detection task. The tasks were run in
this fixed order so that the younger participants could familiarize
themselves with the general task demands. While there is no way
to eliminate the possibility that the different patterns of change
over age observed across the two tasks may be partly attributed to
the effects of fatigue, once again we would expect fatigue to result
in greater variability in performance tracks for the pitch-direction
discrimination task as compared with the pitch-change detection
task and this is not what we found. Given that track variability was
equivalent across tasks, it seems unlikely that fatigue was a major
contributory factor to performance on the pitch-direction discrim-
ination task.
A number of studies investigating pitch-change detection and
pitch-direction discrimination have indicated that these two pro-
cesses can dissociate in typical adults (Micheyl, Kaernbach, &
Demany, 2008; Mathias et al., 2010; Semal & Demany, 2006),
Table 3
Games–Howell Post-Hoc Comparisons for the Pitch-Direction Discrimination Task
Age years
(I) (J) Mean difference (I–J) Standard error Significance
11 6 ⫺0.70569 0.12956 .000
ⴱⴱⴱ
7 ⫺0.50123 0.10890 .001
ⴱⴱ
8 ⫺0.50174 0.10850 .001
ⴱⴱ
9 ⫺0.37451 0.13962 .144
10 ⫺0.35333 0.10856 .041
ⴱ
18 ⫹ 0.04859 0.09506 .998
18 ⫹ 6 ⫺0.75428 0.12081 .000
ⴱⴱⴱ
7 ⫺0.54982 0.09832 .000
ⴱⴱⴱ
8 ⫺0.55033 0.09788 .000
ⴱⴱⴱ
9 ⫺0.42310 0.13154 .05
ⴱ
10 ⫺0.40192 0.09794 .005
ⴱ
11 ⫺0.04859 0.09506 .998
ⴱ
p ⬍ .05.
ⴱⴱ
p ⬍ .001.
ⴱⴱⴱ
p ⬍ .0001.
Figure 2. Mean thresholds (logarithmically transformed) for the pitch-
change detection and pitch-direction discrimination tasks. Error bars de-
note standard error of the mean.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
79
PITCH DISCRIMINATION
adults with congenital amusia (Foxton et al., 2004; Liu et al., 2010;
Williamson & Stewart, 2010), and lobectomy patients (Johnsrude
et al., 2000; Tramo et al., 2002). The findings from our study
indicate that the capacity to discriminate directional changes in
pitch matures later in typical development than the capacity to
detect equivalently small changes in pitch. The higher perceptual
thresholds observed in the youngest participants in the current
study for pitch-direction discrimination cannot be attributed to
conceptual difficulties relating pitch to the terms “high and low”
(Andrews & Madeira, 1977), and the observation that the variabil-
ity in performance tracks was equivalent across the pitch-change
detection and pitch-direction discrimination tasks reduces the like-
lihood that inattention, fatigue, or confusion dramatically influ-
enced performance on the pitch-direction discrimination task.
Thus, the adult-like thresholds achieved by the majority of children
on the pitch-change detection task together with the monotonically
decreasing thresholds for the pitch-direction discrimination task
supports the hypothesis that these two auditory abilities differ in
their developmental trajectories.
The prolonged trajectory for the development of pitch-direction
discrimination may make this process especially prone to devel-
opmental perturbation, that is, more susceptible to developmental
delay or arrest. It has been consistently demonstrated that adult
congenital amusics show higher thresholds for pitch-direction dis-
crimination as compared with pitch-change detection (Foxton et
al., 2004; Liu et al., 2010; Williamson & Stewart, 2010). The
impairments in pitch perception observed in congenital amusia
may stem from a failure to follow the normal developmental
trajectory for the cortical analysis of pitch pattern information, of
which the capacity to discriminate the direction of a change in
pitch is an integral part.
References
Andrews, M. L., & Madeira, S. S. (1977). The assessment of pitch
discrimination ability in young children. Journal of Speech & Hearing
Disorders, 42, 279 –286.
Bartlett, J. C., & Dowling, W. J. (1980). Recognition of transposed mel-
odies: A key distance effect in developmental perspective. Journal of
Experimental Psychology: Human Perception and Performance, 6, 501–
515. doi:10.1037/0096-1523.6.3.501
Bishop, D. V. M., Adams, C. V., Nation, K., & Rosen, S. (2005). Percep-
tion of transient nonspeech stimuli is normal in children with specific
language impairment. Applied Psycholinguistics, 26, 175–194. doi:
10.1017/S0142716405050137
Cooper, N. (1994). An exploratory study in the measurement of children’s
pitch discrimination ability. Psychology of Music, 22, 56 –62. doi:
10.1177/0305735694221005
Demany, L., Carlyon, L. P., & Semal, C. (2009). Continuous versus
discrete frequency changes: Different detection mechanisms? Journal of
the Acoustical Society of America, 125, 1082–1090. doi:10.1121/1
.3050271
Dowling, W. J., & Bartlett, J. C. (1981). The importance of interval
information in long-term memory for melodies. Psychomusicology: A
Journal of Research in Music Cognition, 1, 30 –49. doi:10.1037/
h0094275
Dowling, W. J., & Fujitani, D. S. (1971). Contour, interval, and pitch
recognition in memory for melodies. Journal of the Acoustical Society of
America, 49, 524 –531. doi:10.1121/1.1912382
Dowling, W. J., & Harwood, D. L. (1986). Music cognition (pp. 124 –152).
London: Academic Press.
Dowling, W. J., Kwak, S., & Andrews, M. W. (1995). The time course of
recognition of novel melodies. Perception and Psychophysics, 57, 136 –
149. doi:10.3758/BF03206500
Field, A. (2005). Discovering statistics using SPSS (pp. 63–106). London:
SAGE Publications Ltd.
Foxton, J. M., Dean, J. L., Gee, R., Peretz, I., & Griffiths, T. D. (2004).
Characterization of deficits in pitch perception underlying ‘tone deaf-
ness’. Brain, 127, 801– 810. doi:10.1093/brain/awh105
Geisser, S., & Greenhouse, S. W. (1958). An extension of Box’s results on
the use of the F distribution in multivariate analysis. The Annals of
Mathematical Statistics, 29, 885– 891. doi:10.1214/aoms/1177706545
Griffiths, T. D. (2003). Functional Imaging of Pitch Analysis. Annals of the
New York Academy of Sciences, 999, 40 –49. doi:10.1196/annals.1284
.004
Habib, M., Daquin, G., Milandre, L., Royere, M. L., Rey, M., Lanteri, A.,
. . . Khalil, R. (1995). Mutism and auditory agnosia due to bilateral
insular damage – role of the insula in human communication. Neuro-
psychologia, 33, 327–339. doi:10.1016/0028-3932(94)00108-2
Hattiangadi, N., Poillon, J. P., Slomine, B., Christensen, J., Trovato, M. K.,
& Speedie, L. J. (2005). Characteristics of auditory agnosia in a child
with severe traumatic brain injury: A case report. Brain and Language,
92, 12–25. doi:10.1016/j.bandl.2004.05.003
Figure 3. Pitch-change detection and pitch-direction discrimination
thresholds (log transformed ST) plotted against age in months.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
80
FANCOURT, DICK, AND STEWART
Horváth, J., Czigler, I., Birkas, E., Winkler, I., & Gervais, J. (2009).
Age-related differences in distraction and reorientation in an auditory task.
Neurobiology of Aging, 30, 1157–1172. doi:10.1016/j.neurobiolaging
.2007.10.003
Howell, D. C. (1997). Statistical methods for psychology (4th ed., pp.
317–362). Belmont, CA: Duxbury Press.
Jensen, J. K., & Neff, D. L. (1993). Development of basic auditory
discrimination in preschool children. Psychological Science, 4, 104 –
107. doi:10.1111/j.1467-9280.1993.tb00469.x
Johnsrude, I. S., Penhune, V. B., & Zatorre, R. J. (2000). Functional
specificity in the right human auditory cortex for perceiving pitch
direction. Brain, 123, 155–163. doi:10.1093/brain/123.1.155
Liu, F., Patel, A. D., Fourcin, A., & Stewart, L. (2010). Intonational
identification-imitation dissociation in congenital amusia. Brain, 133,
1682–1693. doi:10.1093/brain/awq089
Mathias, S. R., Micheyl, C., & Bailey, P. J. (2010). Stimulus uncertainty
and insensitivity to pitch-change direction. Journal of the Acoustical
Society of America, 127, 3026 –3037. doi:10.1121/1.3365252
Maxon, A. B., & Hochburg, I. (1982). Development of psychoacoustic
behaviour: Sensitivity and discrimination. Ear and Hearing, 3, 301–308.
doi:10.1097/00003446-198211000-00003
McDermott, J. H., & Oxenham, A. J. (2008). Music perception, pitch, and
the auditory system. Current Opinion in Neurobiology, 18, 452– 463.
doi:10.1016/j.conb.2008.09.005
Micheyl, C., Kaernbach, C., & Demany, L. (2008). An evaluation of
psychophysical models of auditory change perception. Psychological
Review, 115, 1069 –1083.
Moore, D. R., Ferguson, M. A., Halliday, L. F., & Riley, A. (2008).
Frequency discrimination in children: Perception, learning and attention.
Hearing Research, 238, 147–154. doi:10.1016/j.heares.2007.11.013
Olsho, L. W. (1984). Infant frequency discrimination. Infant Behaviour
and Development, 7, 27–35. doi:10.1016/S0163-6383(84)80020-X
Olsho, L. W., Koch, E. G., & Halpern, C. F. (1987). Level and age effects
in infant frequency discrimination. Journal of the Acoustical Society of
America, 82, 454 – 464. doi:10.1121/1.395446
Olsho, L. W., Schoon, C., Sakai, R., Turpin, R., & Sperduto, V. (1982).
Auditory frequency discrimination in infancy. Developmental Psychol-
ogy, 18, 721–726. doi:10.1037/0012-1649.18.5.721
Patel, A. D. (2008). Music, language and the brain (pp. 181–238). New
York: Oxford University Press.
Patterson, R. D., Uppenkamp, S., Johnsrude, I. S., & Griffiths, T. D.
(2002). The processing of temporal pitch and melody information in
auditory cortex. Neuron, 36, 767–776. doi:10.1016/S0896-
6273(02)01060-7
Randel, D. M. (1978). Harvard concise dictionary of music. Cambridge,
MA: Harvard University Press.
Semal, C., & Demany, L. (2006). Individual differences in the sensitivity
to pitch direction. Journal of the Acoustical Society of America, 120,
3907–3915. doi:10.1121/1.2357708
Sen, P. K. (1968). Estimates of the regression coefficient based on Kend-
all’s tau. Journal of the American Statistical Association, 63, 1379 –
1389. doi:10.1080/01621459.1968.10480934
Soderquist, D. R., & Moore, M. (1970). Effect of training on frequency
discrimination in primary school children. Journal of Auditory Research,
10, 185–192.
Stalinski, S. M., Schellenberg, G. E., & Trehub, S. E. (2008). Develop-
mental changes in the perception of pitch contour: Distinguishing up
from down. Journal of the Acoustical Society of America, 124, 1759 –
1763. doi:10.1121/1.2956470
Stewart, L., Kriegstein, V. K., Warren, J. D., & Griffiths, T. D. (2006).
Music and the brain: Disorders of musical listening. Brain, 129, 2533–
2553. doi:10.1093/brain/awl171
Terao, Y., Mizuno, T., Shindoh, M., Sakurai, Y., Ugawa, Y., Kobayashi,
S.,...Tsuji, S. (2006). Vocal amusia in a professional tango singer die
to a right superior-temporal cortex infarction. Neuropsychologia, 44,
479 –488. doi:10.1016/j.neuropsychologia.2005.05.013
Theil, H. (1950). A rank-invariant method of linear and polynomial re-
gression analysis. Nederlandse Akademie Wetenchappen Series A, 53,
386 –392.
Thompson, N. C., Cranford, J. L., & Hoyer, E. (1999). Brief-tone fre-
quency discrimination by children. Journal of Speech Language and
Hearing Research, 42, 1061–1068.
Tramo, M. J., Shah, G. D., & Braida, L. D. (2002). Functional role of
auditory cortex in frequency processing and pitch perception. Journal of
Neurophysiology, 87, 122–139.
Trehub, S. E., Bull, D., & Thorpe, L. A. (1984). Infants’ perception of
melodies: The role of melodic contour. Child Development, 55, 821–
830.
Trehub, S. E., Cohen, A. J., Thorpe, L. A., & Morrongiello, B. A. (1986).
Development of the perception of musical relations: Semitone and
diatonic structure. Journal of Experimental Psychology Human Percep-
tion and Performance, 12, 295–301. doi:10.1037/0096-1523.12.3.295
Trehub, S. E., Thorpe, L. A., & Morrongiello, B. A. (1985). Infants’
perception of melodies: Changes in a single tone. Infant Behaviour and
Development, 8, 213–223. doi:10.1016/S0163-6383(85)80007-2
Van Zee, N. (1976). Responses of kindergarten children to musical stimuli
and terminology. Journal of Research in Music Education, 24, 14 –21.
doi:10.2307/3345062
Wetzel, N., & Schroeger, E. (2007). Modulation of involuntary attention by
the duration of novel and pitch deviant sounds in children and adoles-
cents. Biological Psychology, 75, 24 –31. doi:10.1016/j.biopsycho.2006
.10.006
White, D. J., Dale, P. S., & Carlsen, J. C. (1990). Discrimination and
categorization of pitch direction by young children. Psychomusicology:
A Journal of Research in Music Cognition, 9, 39 –58. doi:10.1037/
h0094159
Williamson, V. J., Liu, F., Peryer, G., Grierson, M., & Stewart, L. (2012).
Perception and action de-coupling in congenital amusia: Sensitivity to
task demands. Neuropsychologia, 50, 172–180. doi:10.1016/j
.neuropsychologia.2011.11.015
Williamson, V. J., & Stewart, L. (2010). Memory for pitch in congenital
amusia: Beyond a fine-grained pitch perception problem. Memory, 18,
657– 669. doi:10.1080/09658211.2010.501339
Wright, D. B., & Field, A. P. (2009). Giving your data the bootstrap. The
Psychologist, 22, 412– 413.
Received July 30, 2012
Revision received March 21, 2013
Accepted March 25, 2013 䡲
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
81
PITCH DISCRIMINATION
