An Examination of the Relationship Between

Electroglottographic Contact Quotient,

Electroglottographic Decontacting Phase Profile,
and Acoustical Spectral Moments
*Shaheen N. Awan, *Andrew R. Krauss, and Christian T. Herbst, *Bloomsburg, Pennsylvania, and yOlomouc, Czech Republic
Summary: Objectives. To date, only a few studies have examined the possible relationship between electroglottographic (EGG) data and spectral characteristics of the voice. This study examined the possible association between EGG
signal data (contact quotient [CQ] and decontacting phase profile) and spectral moments of the acoustic signal (spectral
mean, spectral standard deviation (SD), spectral skewness, and spectral kurtosis). Furthermore, the possible effects of
gender on these measurements were analyzed.
Methods. Sustained vowel // productions were obtained from 48 normophonic individuals (24 adult males and 24
adult females). The central 1-second portions of the acoustic vowel samples were analyzed for spectral moments, and
the EGG signal was analyzed for CQ (CQEGG), fundamental frequency (F0), and decontacting phase profile.
Results. Across all subjects, the spectral characteristics of the voice (in particular, spectral SD, skewness, and kurtosis) are significantly related to changes in the relative duration of vocal fold contact (as measured via CQEGG). In addition, significant effects of the profile of the EGG decontacting phase (ie, concave down/knee vs concave up/no
knee) on spectral SD were also observed, as well as a strong trend for decontacting phase profile to influence the
spectral mean.
Discussion. Although the degree of vocal fold contact and differences in decontacting phase profile may have an
influence on the spectral characteristics of the acoustic voice signal, the strength of correlations between CQEGG values
and measures of spectral moments only accounted for approximately 1316% of the variation in spectral distribution
characteristics. These results stress the importance of the transformative role of the supraglottal vocal tract in producing
an acoustic output that maintains some of the characteristics of the glottal source, but which modifies the source
characteristics in ways not completely accounted for by single parameters such as CQEGG or EGG profile.
Key Words: ElectroglottographySpectral momentsContact quotient.
INTRODUCTION
When analyzing vocal fold movement during phonation, adult
males and females have been described as exhibiting physiological differences throughout the glottal cycle.1 During the contact
phase of vibration, men have been reported to approximate more
complete glottal closure than women,2,3 corresponding to a
greater relative duration of vocal fold closure in men versus
women.3,4 In contrast to observations for adult male speakers,
multiple researchers have noted that a majority of adult
females produce comfortable pitch and loudness phonation
with incomplete vocal fold closure (posterior glottal gap)
during modal register phonation,58 resulting in a relative
increase in the duration of the open phase and a possible
tendency for a subtle breathy voice quality in some women. In
addition, adult females have been described as having less
abrupt contacting and decontacting behavior than adult males.9
One method for objectively analyzing vocal fold contact during the phonatory cycle is electroglottography (EGG).10,11 EGG
Accepted for publication October 23, 2014.
Portions of this manuscript were presented at the 43rd Annual Symposium: Care of the
Professional Voice, Philadelphia, PA. May 29, 2014.
From the *Department of Audiology & Speech Pathology, Bloomsburg University of
Pennsylvania, Bloomsburg, Pennsylvania; and the yVoice Research Lab, Department of
Biophysics, Faculty of Science, Palacky University, Olomouc, Czech Republic.
Address correspondence and reprint requests to Dr. Shaheen N. Awan, Department of
Audiology & Speech Pathology, Bloomsburg University of Pennsylvania, 400 East 2nd
St., Bloomsburg, PA, 17815. E-mail: sawan@bloomu.edu
Journal of Voice, Vol. 29, No. 5, pp. 519-529
0892-1997/$36.00
2015 The Voice Foundation
http://dx.doi.org/10.1016/j.jvoice.2014.10.016

is a noninvasive method of measuring vocal fold contact during

voice production4 that provides a physiological correlate of
vocal fold vibration.12 To obtain the EGG signal, two electrodes
are placed on either side of the neck at the level of the vocal
folds, and a small high-frequency current is passed between
the electrodes.4 When the current is transmitted during vocal
fold vibration, the EGG signal varies in response to the fluctuations in electrical impedance that are induced by changes in the
contact area of the vocal folds.13 The resulting EGG waveform
can be analyzed for contact quotient (CQEGGthe relative duration of vocal fold contact during the phonatory cycle14). In addition, modeling studies have shown that a description of the
quality of vocal fold contact during the contacting and decontacting phases may be made from the EGG waveform.15,16
Previous research has described key characteristics of adult
male versus female EGG waveforms. As expected, men have
been observed to produce greater CQEGGs than women3,17
although this observation may depend on the method used for
computation of the CQEGG.18 In addition, description of the
EGG waveform profile has indicated that many adult male
EGG waveforms are characterized as having a concave-down
profile or knee in the decontacting phase1,2,19 that corresponds
to a bulging or skirting motion of the vocal fold(s) in the
vertical dimension,1 resulting in a tendency for increased vocal
fold contact per glottal cycle.2 Although the knee tends to occur
in the majority of adult male EGG waveforms, the presence of a
decontacting knee has also been reported in a smaller percentage
of adult female EGG waveforms at habitual speaking pitch/

520
frequency levels17,20,21 (eg, Awan and Awan17 recently reported
that the knee was present in 76% of the males and 36% of the females that they studied).
In humans, most of the acoustic energy is generated by the
cessation of the transglottal airflow.22 The relative duration of
the closed phase during a glottal cycle has a direct influence on
the shape of the glottal airflow pulse and hence the generation of
stronger high-frequency partials in the radiated sound source spectrum.23,24 Therefore, it may be expected that differences in the
relative duration and the quality or type of vocal fold contact
during the glottal cycle (as discussed in the aforementioned
literature review regarding adult male vs female EGG
waveforms) will have an effect on the spectral characteristics of
the acoustic voice signal. Holmberg et al25 examined the relationships between aerodynamic, acoustic, and EGG measures of voice
in consonant-vowel syllable and vowel production. Results
showed weak but significant negative relationships between
EGG adduction quotient and harmonic-to-harmonic (H1-H2)
and harmonic-to-formant (H1-F1 and H1-F3) amplitude differences, whereas weak positive correlations were observed between
H1-F3 noise and F1-F3 noise. These authors hypothesized that
gradual closing movements of a somewhat abducted membranous
portion of the vocal folds would result in relatively sinusoidal
glottal waveforms and small adduction quotients and that the corresponding acoustic spectrum would be characterized by a first
harmonic with relatively high amplitude, a steep overall spectral
slope, and an attenuated F1 peak amplitude (because of a relatively
wide F1 bandwidth). However, it was reported that strong relationships (rs > 0.70) between EGG adduction quotients and the aforementioned spectral characteristics were observed in 35% of the
subjects. Winkler et al26 examined the effect of aging on EGG,
acoustic, and perceptual analyses of the voice. Results indicated
a characteristic spectral distribution in the // vowel of male voices. Long-term average spectra (LTAS) of the male voices
perceived as old had relatively little harmonic energy between 2
and 4 kHz, whereas younger male voices showed distinct spectral
harmonics up to 4 kHz. The spectral characteristics for the young
males coincided with EGG waveforms that showed a steep increase of vocal fold contact in the closing and a more extended
opening phase with the instant of glottal opening clearly visible
in the EGG signal as a knee. In contrast, the EGG waveforms
for the older males were nearly sinusoidal with reduced steepness
of the increasing signal slope. Bestebreurtje and Schutte27 examined EGG and spectral characteristics of the belting singing style
in a single subject study. Results indicated that the loud bright
sound of the belting style is achieved by enhanced higher harmonics in the acoustic spectrum and a corresponding raising of
the CQEGG above 52% (the authors speculated that this was an
apparent threshold value for belting).
In addition to the examination of isolated spectral harmonic
amplitudes, one of the methods that may be used to summarize
the spectral characteristics of the acoustic voice signal is by
examining the spectral moments of the LTAS.28 In the computation of spectral moments, the spectral values are treated as a
random probability distribution, with the first spectral moment
being the spectral mean (a measure of the spectrums central
tendency in hertz; the center of gravity); the second spectral

Journal of Voice, Vol. 29, No. 5, 2015

moment the standard deviation (SD; a measure linked with the

variance around first moment); the third spectral moment being
skewness (a dimensionless measure reflecting the average symmetry of the spectral distribution; positive skewness coincides
with a higher concentration of energy in the lower spectral regions and negative vice versa); and the fourth spectral moment
being kurtosis (a dimensionless measure of the shape of the
spectrum, with positive values representative of a compact spectrum and energy concentrated within a very small frequency
range versus low or negative values indicative of a flattened
spectrum; may be interpreted via peakedness (width of the
peak), as well as tail weight (heavy vs light tails), and the presence of infrequent extreme deviations).2933 Spectral moments
analyses have had various applications to speech/voice
analysis and have been used to assess the impact of vocal
effort on voice characteristics31; to distinguish between typical
versus disordered voice subjects during continuous speech34;
to evaluate the impact of unilateral vocal fold paralysis on pre
versus post-therapy voice characteristics35; to provide a voice
treatment outcomes measure36; to aid in the differential diagnosis of adductor spasmodic dysphonia versus muscle tension
dysphonia patients37; to discriminate between hypokinetic dysarthric and neurologically normal subjects38; and to assess the
effects of cochlear implant on speech characteristics.39
Investigation of the possible relationships between physiological measures of voice (such as EGG) and the acoustic
output of the speech/voice mechanism is valuable in our attempts to gain a deeper understanding of normal voice production.25 Because acoustic measures are readily available, low
cost, noninvasive, and familiar to researchers and clinicians
alike, they are probably the most frequently used objective
measures of voice. Research that can establish the relationship(s) between acoustic and physiological measures of voice
are important in demonstrating the validity of these methods.
Because EGG is a noninvasive (and hence easily obtained)
physiological correlate of vocal fold vibration, it is important
to relate acoustic measures (spectral or time based) to data
from the EGG waveform and thus possibly gain a better understanding of the causal relationship between the quality of vocal
fold vibration and the resulting sound output. Spectral-based
measures (such as spectral moments) are of particular interest
because (1) they have been reported to be effective in documenting voice types that are problematic for traditional perturbation measures (eg, strained voice) and (2) (although not the
focus of this study) they have been shown to be applicable to
the analysis of continuous speech samples. Therefore, the purpose of this study was to examine the possible association between EGG characteristics (including CQEGG and EGG
profile) and spectral characteristics referred to as spectral moments from corresponding acoustic voice signals. In addition,
possible gender differences in EGG and spectral characteristics
of the voice were also investigated.

METHODOLOGY
The Bloomsburg University of Pennsylvania Institutional
Review Board approved the methodology used in this study.

Shaheen N. Awan, et al

Relationship Between EGG CQ, EGG Profile, & Spectral Moments

Participants
Participants consisted of 24 adult males and 24 adult females
between the ages of 18 and 30 years (mean age 21.33 years,
SD 1.49), with typical voice characteristics of pitch, loudness, and quality as determined by a certified speechlanguage pathologist. The mean fundamental frequency (F0)
for the adult male subjects was 112.22 Hz (SD 16.99);
mean F0 for the adult females was 210.00 Hz (SD 16.16).
All subjects passed a hearing screening (25 dB at 0.5, 1, 2,
and 4 kHz).
Task and equipment
Participants produced the vowel // at their comfortable pitch
and loudness for approximately 3 seconds in duration. Each
vowel was repeated three times. During vowel productions,
EGG signals were recorded using a KayPENTAX Model
6103 Electroglottograph (KayPENTAX Inc., Montvale, NJ).
During EGG recording, the time-varying degree of admittance
of an electrical signal presented at the level of the vocal folds
was recorded via two gold-plated EGG electrodes placed on
either side of subjects laryngeal prominence. The signal was
high-pass filtered at 20 Hz to remove direct current (DC) float
and low-frequency artifacts. Sigma Creme electrode gel (Parker
Laboratories, Fairfield, NJ) was used to enhance conductivity,
and optimal positioning of the electrodes was checked by monitoring the EGG signal amplitude in real-time using a Protek
6502 Oscilloscope (Protek Test & Measurement, Norwood,
NJ). During vowel production, the acoustic signal was recorded
using a Shure SM10 A microphone (Shure Inc., Niles, IL)
placed approximately 45 off-center at a constant one inch
(2.5 cm) distance from the participants mouth. EGG and acoustic signals were digitized simultaneously using KayPENTAX
Model 4500 Computerized Speech Lab (CSL) version 3.4.1
and Real-Time EGG software (KayPENTAX) at 44.1 kHz,
16 bits of resolution per channel.
Signal analyses
For each trial of the vowel //, a central 1 second portion of the
combined two-channel acoustic + EGG signal was edited for
further analyses. Because the EGG signal can be adversely
affected by the presence of low-frequency modulations as well
as random broadband noise, the channel containing the EGG
signal was bandpass filtered using a phase-preserving Blackman
filter (pass band of 201500 Hz) using the CSL main program,
consistent with filtering procedures used in a number of previous
studies.18,40 EGG signals were then analyzed using a custom
program written using SciLab version 5.3.3 open source
software for numerical computation (Scilab Enterprises,
Versailles, France) as previously described and validated in
Awan and Awan.17 The analysis algorithm uses the derivative
method for determining the CQEGG, in which the derivative
maxima corresponds to the vocal fold contacting point and the
derivative minima corresponds to the vocal fold decontacting
point.17,18,41 In addition to CQEGG, mean fundamental
frequency (F0) estimates were also obtained for each trial of
the vowel // via autocorrelation-guided peak picking of the
CQEGG maxima.

521

The unfiltered channel containing the acoustic signal for the

vowel // was low-pass filtered at 11 kHz (Blackman window;
filter order 100), downsampled to 22.05 kHz, and analyzed
using the LTAS analysis (1024 points; Blackman window) option in the CSL main software program (KayPENTAX Inc.).
Similar analysis bandwidths and LTAS parameters have been
previously used for the analysis of spectral moments in Tanner
et al.36 and Houtz et al37 The estimated time delay of 0.5 milliseconds of the acoustic signal vis-
a-vis the EGG signal was
considered to be negligible for LTAS analysis. After LTAS analysis, spectral moments (spectral mean, spectral SD, spectral
skewness, and spectral kurtosis) were computed.
In addition to EGG and spectral moments analyses, zoomed
images of the waveform (using screen capture software Grab;
Apple Inc., Cupertino, CA) showing three representative vibratory cycles were scored by two independent judges to fall into
one of two categories for the profile of the decontacting phase of
the EGG waveform: concave down (knee is present in the decontacting phase) or other (no knee is present during the decontacting phase; concave up or other profile). Before the judgment
task, judges were provided with examples of concave down/
knee versus concave up/no knee examples (Figure 1). Judges
were blinded regarding the gender of speaker and were presented the EGG waveform images in randomized order.
Because a review of the data indicated very little if any variation
in EGG waveform profile from trial to trial, only EGG waveforms from the second trial of the vowel // were judged. An
evaluation of inter-rater reliability showed an exact agreement
between judges on 40 of the 48 profiles (83.3% exact agreement). One of the researchers (A.R.K.) served as a third judge
to settle disagreements between the original judges (n 8).
RESULTS
All statistical analyses were computed using SPSS version. 16.0
(SPSS Inc., Chicago, IL). Analyses were focused on the second
trial of the sustained vowel // because (1) preliminary analyses
had indicated no significant difference between trials for measures of CQEGG, F0, or the various spectral moments and (2) ratings of EGG profile (knee vs no knee) were obtained from the
second trial productions. Because the shift in harmonics due to
subject-to-subject differences in F0 may influence spectral measures, additional preliminary analyses were conducted to ascertain the degree of correlation between measures of F0 and
spectral moments. Results indicated that F0 was significantly

FIGURE 1. Examples of (A) concave up/no knee versus (B) concave

down/knee-shaped EGG decontacting profiles.

522

Journal of Voice, Vol. 29, No. 5, 2015

and directly correlated with spectral SD (r 0.43, r2 0.18;

P 0.002), and inversely correlated with spectral skewness
(r 0.30, r2 0.09; P 0.037). However, being that these correlations were considered weak, F0 was not included as a partial
correlate in subsequent correlational analyses or as a covariate in
subsequent analyses of variance (ANOVA) because covariate
adjustment with weakly correlated variables will not provide
any noticeable improvement in power and may actually reduce
statistical power due to a loss of degrees of freedom.42 Interpretation of the strength of ANOVA effect sizes using h2 was based
on Cohen,43 who proposed that a small effect 0.01, a moderate
effect 0.06, and a strong effect 0.14. Because effect size measures from nonsignificant results may still indicate the potential
for a subtle but important effects,4447 examination of findings
that were statistically nonsignificant but had effect sizes >0.06
(moderate effect) were also conducted. Table 1 presents the correlation matrix across all subjects, and Tables 2 and 3 present
means for the various dependent variables by gender and by
EGG decontacting phase profile, respectively.
Relationships among EGG and spectral variables
Correlational analyses were computed to ascertain the strength
and direction (direct vs inverse) between measures of CQEGG
and spectral moments computed from acoustic waveforms
(spectral mean, SD, skewness, and kurtosis). Across all subjects
(total N 48), statistically significant correlations were
observed between measures of CQEGG and spectral SD
(r 0.40; r2 0.16; P 0.005) and between measures of
CQEGG and spectral skewness (r 0.37; r2 0.13; P 0.01)

and kurtosis (r 0.36; r2 0.13; P 0.011). Table 1 provides

the correlation matrix as computed across all 48 adult male and
female subjects combined.
In addition to significant correlations between CQEGG and
the aforementioned spectral moments, several other statistically
significant correlations were also observed (Table 1): spectral
mean with spectral skewness (r 0.44; r2 0.19;
P 0.002) and spectral kurtosis (r 0.37; r2 0.13;
P 0.01), spectral SD with spectral kurtosis (r 0.29;
r2 0.08; P 0.045), and spectral skewness with spectral kurtosis (r 0.87; r2 0.76; P < 0.001).
Correlational analyses were also conducted within each adult
male and female subgroup. CQEGG was not observed to correlate significantly with any of the spectral moments within either
of the gender subgroups although the directions of observed
correlations between CQEGG and spectral SD, skewness, and
kurtosis were the same as observed for all subjects combined.
Within the adult male subgroup, the only significant correlation
observed was between spectral skewness and spectral kurtosis
(r 0.85; r2 0.72; P < 0.001). Within the adult female subgroup, spectral mean was observed to correlate inversely with
spectral skewness (r 0.64; r2 0.41; P 0.001) and spectral kurtosis (r 0.49; r2 0.24; P 0.016), and as in the
adult male subgroup, spectral skewness was strongly correlated
with spectral kurtosis (r 0.91; r2 0.83; P < 0.001).
The general relationship trends observed via correlational
analyses were also investigated by categorizing subjects into
low CQ (20 subjects; 15 women and five men) versus high CQ
(28 subjects; nine women and 19 men) groups. CQ categorization

TABLE 1.
Correlations Between the CQEGG and Spectral Moments Across All subjects (N 48)
Dependent Variable
Contact quotient
Pearson correlation
Sig. (2-tailed)
Index of determination
Spectral mean
Pearson correlation
Sig. (2-tailed)
Index of determination
Spectral standard deviation
Pearson correlation
Sig. (2-tailed)
Index of determination
Spectral skewness
Pearson correlation
Sig. (2-tailed)
Index of determination
Spectral kurtosis
Pearson correlation
Sig. (2-tailed)
Index of determination

Contact
Quotient
1

Spectral Mean

Spectral Standard
Deviation

0.24
0.094
0.06

0.40*
0.006
0.16

0.37y
0.01
0.14

0.36y
0.011
0.13

0.06
0.672
<0.01

0.44*
0.002
0.19

0.37y
0.01
0.14

0.003
0.98
<0.01

0.29y
0.045
0.08

* Correlation is significant at the P < 0.01 level (two-tailed).

y
Correlation is significant at the P < 0.05 level (two-tailed).

Spectral Skewness

Spectral Kurtosis

0.87*
<0.001
0.76
1

Shaheen N. Awan, et al

523

Relationship Between EGG CQ, EGG Profile, & Spectral Moments

TABLE 2.
Summary Statistics for Contact Quotient (CQEGG) and Spectral Moments in the Adult Male and Female Subgroups
Dependent Variable

Gender

Mean

Standard Deviation

Contact quotient (%)

Female
Male
Female
Male
Female
Male
Female
Male
Female
Male

24
24
24
24
24
24
24
24
24
24

36.22
46.96
736.04
670.63
376.97
307.13
2.05
3.21
23.35
42.84

10.63
9.17
120.44
101.57
70.22
66.58
1.41
1.82
23.17
40.24

Spectral mean (Hz)

Spectral standard deviation (Hz)
Spectral skewness
Spectral kurtosis

was determined using the overall mean CQ across all subjects

(CQEGG 41.59%) as the cutoff point. A series of independent
t tests were then conducted to determine if there was a significant
difference between high CQ (mean CQ 49.34 [SD 5.69])
versus low CQ subjects (mean CQ 30.74 [SD 7.28]) on
the various spectral moment variables (adjustment of P levels
for significant results was conducted using the HolmBonferroni method48). Results indicated no significant difference
between high CQ versus low CQ subjects for spectral mean
(689.74 Hz [SD 112.20] vs 722.37 Hz [SD 118.86]; equal
variances not assumed: t(39.65) 0.97, P 0.34). As expected
from the aforementioned correlational analyses, high CQ subjects were observed to have marginal but significantly lower
mean spectral SD versus low CQ subjects (321.05 Hz
[SD 69.93] vs 371.44 Hz [SD 76.87]; Equal variances not
assumed: t(38.61) 2.32, P 0.025; Holm-Bonferroni
adjusted P 0.048), as well as significantly higher spectral skewness (3.12 [SD 1.91] vs 1.95 [SD 1.11]; Equal variances not
assumed: t(44.38) 2.68, P 0.01; Holm-Bonferroni adjusted
P 0.036) and significantly higher spectral kurtosis (42.66
[SD 40.24] vs 19.70 [SD 15.01]; Equal variances not
assumed: t(36.58) 2.76, P 0.009; Holm-Bonferroni adjusted
P 0.036). Figure 2 provides mean LTAS for the high versus low
CQ groups.

Effect of EGG profile

Overall, 24 of 48 participants were judged to display a concavedown profile (ie, knee) in the decontacting phase of the EGG
waveform. In women, seven participants fell into the concave
down/knee group (7/24 29.1%), and 17 in the concave up/other
group (ie, no knee) (17/24 70.9%). In this data set, men were
found to have the exact opposite proportions, with 17 being
scored in the concave down/knee group (17/24 70.9%) and
seven in the concave up/other group (7/24 29.1%).
ANOVAs were conducted to determine if there was a significant difference between concave down/knee versus concave
up/other (ie, no knee) subjects on the various EGG and spectral
moment variables. A series of two-way ANOVAs (gender two
levels; profile two levels) were computed, and the following
results were observed.
Contact quotient. ANOVA results indicated a significant
main effect of gender (F(1, 44) 9.81; P 0.030;
h2 0.17), but no significant main effect of EGG profile
(F(1, 44) 0.26; P 0.62; h2 0.004). In addition, there
was no significant interaction between gender and profile
(F(1, 44) 0.02; P 0.90; h2 < 0.001). The h2 value for the
gender effect was considered to have a strong degree of practical significance with men observed to have greater CQEGGs
than women (46.96% vs 36.22%, respectively; Table 2).

TABLE 3.
Summary Statistics for Contact Quotient (CQEGG) and Spectral Moments in the Concave Down (Knee) and Concave Up (No
Knee) Subgroups
Dependent Variable

Profile

Mean

Standard Deviation

Contact quotient (%)

Concave up
Concave down
Concave up
Concave down
Concave up
Concave down
Concave up
Concave down
Concave up
Concave down

24
24
24
24
24
24
24
24
24
24

38.68
44.50
745.16
661.51
336.9
347.14
2.31
2.96
33.46
32.73

12.74
8.70
101.17
114.76
80.97
72.79
1.64
1.76
39.31
28.45

Spectral mean (Hz)

Spectral standard deviation (Hz)
Spectral skewness
Spectral kurtosis

524

Journal of Voice, Vol. 29, No. 5, 2015

FIGURE 2. Mean LTAS for high versus low CQ subjects (CQ categorization was determined using the overall mean CQ across all subjects
(CQEGG 41.59%) as the cutoff point). Raw LTAS (A) and Loess smoothed LTAS (B) are displayed. High CQ subjects were observed to have significantly lower mean spectral standard deviation, significantly higher spectral skewness and significantly higher spectral kurtosis versus low CQ
subjects.
Figure 3 provides a histogram showing the range of CQEGG
values within each adult male and female subgroup.
Spectral mean. No significant main effects of profile (F(1,
44) 3.97, P 0.053; h2 0.074) or gender (F(1,
44) 1.17, P 0.29; h2 0.022) and no significant interaction
between gender and profile (F(1, 44) 1.13, P 0.29;
h2 0.021). Although nonsignificant, the main effect of EGG
profile was considered to have an effect size that indicated a
moderate degree of practical significance, with the concave
down/knee group observed to have lower average spectral
mean than the no knee profile group (661.51 Hz vs
745.16 Hz, respectively; Table 3).
Spectral SD. Analyses showed significant main effects of
gender (F(1, 44) 18.19, P < 0.001; h2 0.29) and profile
(F(1, 44) 5.11, P 0.029; h2 0.082), whereas the interaction
of gender and profile was statistically nonsignificant (F(1,

44) 0.07, P 0.79; h2 0.001). Review of the strong effect

of gender showed that men had a significantly lower mean spectral SD than women (adult male vs female: 307.13 Hz vs 376.97).
The effect size for profile was considered to be of moderate
strength, with the concave down profile/knee group having a
significantly greater mean spectral SD than the concave up/no
knee profile group (347.14 Hz vs 336.96 Hz).
Spectral skewness. A significant main effect of gender
was observed (F(1, 44) 4.19, P 0.047; h2 0.084),
but no significant main effect of profile (F(1, 44) 0.14,
P 0.71; h2 0.003) or interaction between gender and profile (F(1, 44) 0.01, P 0.92; h2 < 0.001). Further review
of the gender subgroup means (moderate effect size) was
conducted and indicated that men were observed to have
increased positive skewness than women (adult male vs female: 3.21 vs 2.05).

FIGURE 3. Histogram showing the relative distribution of CQEGG values for the adult female and male subgroups.

Shaheen N. Awan, et al

Relationship Between EGG CQ, EGG Profile, & Spectral Moments

525

FIGURE 4. Mean LTAS for concave down/knee versus concave up/no knee EGG decontacting phase profile subjects. Raw LTAS (A) and Loess
smoothed LTAS (B) are displayed. The concave down/knee group was observed to have significantly greater spectral standard deviation and a strong
trend (moderate strength effect size) for a lower spectral mean than concave up/no knee subjects.

Spectral kurtosis. A significant main effect of gender was

observed (F(1, 44) 5.29, P 0.026; h2 0.10), but there
were no significant main effects of profile (F(1, 44) 1.06,
P 0.31; h2 0.021) or interaction between gender and profile
(F(1, 44) 0.99, P 0.32; h2 0.02). Review of the gender
subgroup means (moderate effect size) showed that men had
significantly greater mean spectral kurtosis than women (adult
male vs female: 42.84 vs 23.35).
Figures 4 and 5 provide mean LTAS for the concave down/
knee versus concave up/no knee and adult male versus female
subjects, respectively.
DISCUSSION
The results of this study show that, when evaluated across
gender groups, the spectral characteristics of the voice (in
particular, spectral SD, skewness, and kurtosis) are significantly
related to changes in the relative duration of vocal fold contact
(as measured via CQEGG). However, when analyzed within
gender subgroups, these correlations were observed to be
weaker and nonsignificant. In addition, significant effects of
the profile of the EGG decontacting phase (ie, concave down/
knee vs concave up/no knee) on spectral SD were also
observed, as well as a trend (as determined via the presence
of h2 values >0.06) for decontacting phase profile to influence
spectral mean.
Relationships between CQEGG and spectral
moments
As stated, the CQ of the glottal cycle was shown to negatively
correlate with spectral SD and positively correlate with spectral
skewness and spectral kurtosis of the acoustic voice signal across
all subjects. Further analyses showed that these correlations
were only significant when analyzing the increased range of
CQEGG and spectral moment values produced by combining
adult males and females together versus weaker and nonsignificant correlations (due to restricted ranges) within individual
gender groups. Larger CQEGGs are a reflection of longer duration

vocal fold contact, strong excitation of the vocal tract, and

increased energy in higher frequency partials, as opposed to
phonation produced with smaller CQEGGs. In accord with this
view, the results of this study did show a positive correlation between CQ and spectral kurtosis with increased energy in the tail
of the spectral distribution (approximately 4 kHz and higher) for
high CQ subjects (Figure 2). In addition, the results of this study
indicate that, even when CQEGGs increase, the concentration of
energy may still remain in the lower frequency region of the
spectrum (positive correlation between CQ and spectral skewness). Finally, the degree of variance around the spectral mean
tended to be reduced with increased CQEGGs (inverse correlation
between CQEGG and spectral SD), resulting in more compact
spectral energy distribution in the lower frequency region.
Therefore, although energy in higher partials may increase
with increased CQEGGs, this does not necessarily detract from
energy in the lower harmonics near the fundamental frequency.
It should be noted that, in accord with Holmberg et al,25 the overall strength of relationships between CQEGG and the various
spectral moment measures were relatively weak and only accounted for approximately 1316% of the variance. Alipour
et al49 have reported that an increase in glottal adduction may
result in variations in the balance of spectral energy between harmonics and result in unpredictable increases in the spectral energy of the first few harmonics. Therefore, as observed in this
study, CQEGG may only account for a relatively small portion
of the variance in the spectral distribution characteristics of the
airborne acoustic signal.
In addition to significant correlations between CQEGG and
spectral moments, several significant correlations were also
observed within the spectral moment measures. Across all subjects, increases in the spectral mean were associated with decreases in spectral skewness (r 0.44) and kurtosis
(r 0.37). This finding indicates that, as the center of gravity
of spectral energy moves away from lower frequency harmonics
to higher frequency regions of the spectrum, the overall spectral
distribution approaches symmetry (ie, skewness reduced toward
zero), and the overall spectrum will tend to become flatter. This

526

Journal of Voice, Vol. 29, No. 5, 2015

FIGURE 5. Mean LTAS for concave adult male versus female subjects. Raw LTAS (A) and Loess-smoothed LTAS (B) are displayed. Men were
observed to have significantly greater spectral skewness and spectral kurtosis than women, whereas women presented with significantly greater spectral standard deviation versus men.
finding was particularly prominent and statistically significant in
women (rs 0.64 and 0.49 for correlations between spectral
mean and spectral kurtosis and skewness, respectively) although
weak and nonsignificant in men. Calculation of coefficients of
variation (CV) indicated that there was greater relative variation
within the adult female versus male group for the spectral mean
(CVs 16.36% vs 15.15%, respectively), kurtosis
(CVs 99.23% vs 93.93%, respectively), and particularly skewness (CVs 68.78% vs 56.70%, respectively). As previously
mentioned, increased variance and range of data often aids in
revealing correlational trends and may have resulted in the stronger correlations observed between these aforementioned spectral
moments in adult females versus males. Across all subjects, spectral skewness and spectral kurtosis were significantly and
strongly correlated. Spectral energy concentration in the lower
harmonics (ie, more positively skewed) tends to result in a
more peaked spectrum (ie, higher kurtosis) with the possibility
of a heavier tail in the spectral distribution (spectral energy
may move from the shoulders of the spectral distribution to
the higher frequency region). In contrast, as energy becomes
more evenly distributed throughout the spectrum, the spectrum
becomes flatter and spectral kurtosis is reduced.
Effect of EGG profile on spectral moments
The presence of concave down (knee) versus concave up and
other decontacting profiles (no knee) was observed to be
associated with moderate strength effects (h2 values >0.06)
for spectral mean (statistically nonsignificant at P 0.053)
and spectral SD (statistically significant at P 0.029). Subjects
with concave down/knee profiles were observed to have a
strong trend for lower spectral means than subjects with
concave up and other decontacting profiles. Given the profile
of concave down/knee-shaped EGG waveforms, it may be expected that the decontacting point is more abrupt in these subjects as opposed to the more gradual linear decontacting of the
concave up waveforms.4,20 More abrupt decontacting was also
observed to coincide with longer duration vocal fold contact
(increased CQEGGs) and may also be consistent with

relatively abrupt contacting, greater vocal tract excitation

(associated with increased amplitude of the higher frequency
partials49). Because the concave down/knee profile was associated with significantly longer CQEGG than the concave up/no
knee profile, spectral mean was observed to be significantly
lower in this group, with the spectral center of gravity focused
in the lower harmonics. However, although the general relationship between CQEGG and spectral SD was inverse (greater CQs
associated with lower spectral SD), characterization of EGG
profiles by knee versus no knee showed an opposite trend
with knee-shaped profiles having significantly greater spectral
SD than no knee waveforms. This result appears to indicate
that spectral energy may become increasingly distributed in
subjects with concave down/knee-shaped EGG decontacting
profiles. Although this result was somewhat unexpected in relation to the CQEGG results, a review of the acoustic spectra that
coincided with several concave down/knee EGG waveforms
showed increased spectral energy in the location of the first
and second harmonics (H1 and H2) versus no knee examples,
as well as increased energy in the region of the first formant.
It may have been these regions of energy concentration that resulted in the lower center of gravity/spectral mean for some of
the concave-down subjects and the marginal but significant increase in mean spectral SD for the knee-shaped profiles. The
discrepancy in findings between CQ versus overall EGG profile
results may be because of the fact that CQEGG is a single measure of the EGG waveform while the categorization of EGG
profile into knee versus no knee may actually be associated
with multiple EGG characteristics in addition to decontacting
phase profile, such as degree of peakedness of the EGG profile,
slope of the contacting phase, and relative duration of vocal fold
contact. Therefore, EGG profile categorization that incorporates multiple EGG characteristics may be more closely associated with spectral moments than isolated CQEGG values.
Effect of gender on CQEGG and spectral moments
Men were observed to have significantly greater CQs (consistent with several previous studies examining the effects of

Shaheen N. Awan, et al

Relationship Between EGG CQ, EGG Profile, & Spectral Moments

gender on CQEGG1,4,17) and significantly greater spectral

skewness and spectral kurtosis than women, whereas women
presented with significantly greater spectral SD versus men.
Previous literature has indicated that gender differences in
phonatory physiology should be expected and may be
because of a tendency for a relatively abrupt transition from
contacting to decontacting phases in men versus women4,17
and a tendency for women to produce modal register
phonation with a posterior glottal chink.8 Furthermore, longer
CQEGGs are associated with concave-down decontacting profiles, which were observed in 17 men (70.9% of the adult
male subgroup) as opposed to seven women (29.1% of the adult
female subgroup). In this study, adult male subjects were
observed to have decreased spectral SD compared with adult females, as well as increased positive spectral skewness consistent with a greater concentration of spectral energy in the
lower frequency region. The increased spectral SD observed
in women may be due to the tendency to produce phonation
with lower CQEGGs, greater glottal gap size, and the possibility
of additive noise secondary to air escape3,50 while lower
spectral SDs and more positive spectral skew may be
associated with the tendency for longer CQEGGs. Previous
studies using measures of spectral distribution such as the
relative ratio of low-to-high frequency spectral energy, and
the cepstrum have indicated that adult female voice signals
tend to have increased high frequency spectral energy and lower
cepstral peak amplitudes than adult males.3,51,52 The addition of
higher frequency noise may contribute to increased spectral
variation (and increased spectral SD) compared with men.
Limitations
Several limitations in the methodology of this study should be
noted. First, the sustained vowel productions analyzed in this
study were elicited at comfortable pitch and loudness. Although
this allowed us to investigate the habitual EGG and spectral characteristics of our subjects, the natural variation from subject to
subject in vocal fundamental frequency (F0) and sound pressure
level may have also contributed to variability in CQEGG and measures of spectral moments. As an example, because higher F0s
(basically twice as high for adult females vs males) shifts the harmonics upwards, it is unsurprising that the spectral mean was
higher for women than for men (Figure 5). Because F0 may
have influenced the interpretation of some of the results reported
in this manuscript, spectral normalization methods such as those
reported by Boersma and Kovacic53 and Herbst et al54 that
attempt to normalize F0 into specific frequency bandwidth bins
or by harmonics may be investigated (although these methods
will certainly change the standard interpretation of measures
such as the spectral mean and SD as used in this and previous
studies of spectral moments). Second, the LTAS and spectral moments analyses computed in this study used a relatively wide frequency range up to the 11 kHz. Alternatively, the analyses could
have been focused on more concentrated regions of the spectrum
by choosing a lower sampling frequency. However, Tanner et al36
did not observe substantial changes in the general relationships
between spectral moments measures based on analysis frequency range and examination of several sustained vowel sam-

527

ples using different analysis frequency ranges (eg, 05 kHz)

and sampling rates did not show any substantial effect on the general strength and direction of correlations or mean differences reported here. Third, although our interjudge reliability for the
rating of concave-down versus concave-up decontacting profile
was felt to be more than adequate (83.3% exact agreement), at
least eight of the 40 EGG samples met with disagreement among
the two initial judges, and classification was determined via a
third judge. It is clear that EGG decontacting profiles do not always fall clearly into knee versus no knee categories, and
those samples that met with some degree of indecision between
judges may have influenced our interpretation of the associations
between decontacting profile and CQEGG and spectral moments.
Future studies may either (a) incorporate alternative categories
for the categorization of decontacting profile or (b) use an objective metric for the determination of degree of concavity in the decontacting phase profile. Fourth, the computations of CQEGG
were computed using the derivative method because EGG measures obtained via the differentiated EGG (DEGG) correspond
well to measures obtained via the glottal flow waveform and
high-speed film of vocal fold vibration,41,55 may provide a
method of detecting and classifying variations in glottal
configuration,41 and are not limited by the assumption that closing and opening points occur at the same percentage of the peakto-peak (P-to-P) EGG amplitude.17 However, it is recognized
that criterion threshold measures (eg, 25% of the P-to-P EGG
amplitude13,18) have also been commonly used to determine
CQEGG and may result in somewhat different degrees of
relationship with measures of spectral moments as reported in
this study. In addition, a recent study by Herbst et al56 has indicated that DEGG peaks do not always coincide with the specific
events of glottal closure and initial opening (as used in the
computation of CQEGG) and vocal fold contacting and decontacting should be considered to extend over a certain interval.
Finally, a focus of this study was on the profile of the decontacting phase of the EGG waveform and its possible association with
spectral characteristics. Because the main acoustical event of
sound generation in human phonation is constituted by a strong
negative peak in the derivative of the glottal airflow at the instant
of glottal closure,9,57,58 it may be that spectral measures such as
the various moments are more closely associated with measures
that consider the contacting region of the closing phase such as
speed quotient (a measure of the asymmetry of the glottal
pulse59), relative contact rise time (the slope of the increasingcontact portion of the vibratory cycle60,61), or via the alpha
ratio (a measure of the rate of glottal adduction, computed as
the ratio between the summed energy in the 501000 kHz to
the summed energy in the 15 kHz band62,63). Further research
is required to address this issue.

CONCLUSIONS
The results of this study clearly show that degree of vocal fold
contact and differences in decontacting phase profile have an
effect on the spectral characteristics of the voice. However,
the strength of correlations between CQEGG values and measures of spectral moments tended to be relatively weak and

528
only accounted for approximately 1316% of the variation in
spectral distribution characteristics. These findings are consistent with the relatively weak relationships between CQEGG
and spectral measures reported by Holmberg et al25 and recent
observations by Alipour et al49 indicating that degree of glottal
adduction may influence certain spectral characteristics in relatively unpredictable ways. Although somewhat stronger associations were observed between EGG profile (knee vs no knee)
and spectral characteristics (such as a tendency for lower spectral means (moderate effect size) but significantly greater spectral SD (moderate effect size) observed in the acoustic spectral
distributions associated with concave down/knee-shaped
EGG waveforms), the results of this study stress the importance
of the transformative role of the supraglottal vocal tract in producing an acoustic output that maintains some of the characteristics of the glottal source, but which modifies the source
characteristics in ways not completely accounted for by single
parameters such as CQEGG or EGG profile. Future studies that
examine EGG and acoustic measures with larger groups of participants may potentially allow for more clarity regarding the
presence of possible correlations between EGG measures and
spectral moments. In addition, future studies may compare
and contrast the spectral characteristics of the EGG waveform
itself (as in the study by Libeaux64) with the spectral characteristics of the airborne acoustic signal that reflect the characteristics of the vocal tract filter.
Acknowledgment
This study was partially supported by the European Social Fund
project OP VK CZ.1.07/2.3.00/30.0004 POST-UP (CTH).

Journal of Voice, Vol. 29, No. 5, 2015

11.
12.

13.

14.
15.
16.

17.
18.
19.
20.

21.

22.
23.
24.
25.

REFERENCES
1. Titze I. Physiologic and acoustic differences between male and female voices. J Acoust Soc Am. 1989;85:16991707.
2. Chen Y, Robb M, Gilbert H. Electroglottographic evaluation of gender and
vowel effects during modal and vocal fry phonation. J Speech, Lang Hear
Res. 2002;45:821830.
3. Hanson HM, Chuang ES. Glottal characteristics of male speakers: acoustic
correlates and comparison with female data. J Acoust Soc Am. 1999;106:
10641077.
4. Higgins MB, Schulte L. Gender differences in vocal fold contact computed
from electroglottographic signals: the influence of measurement criteria. J
Acoust Soc Am. 2002;111:1865.
5. Echternach M, Dippold S, Sundberg J, Arndt S, Zander MF, Richter B.
High-speed imaging and electroglottography measurements of the open
quotient in untrained male voices register transitions. J Voice. 2010;24:
644650.
6. Hanson HM. Glottal characteristics of female speakers: acoustic correlates.
J Acoust Soc Am. 1997;101:466.
7. Orlikoff RF, Golla ME, Deliyski DD. Analysis of longitudinal phase differences in vocal-fold vibration using synchronous high-speed videoendoscopy and electroglottography. J Voice. 2012;26:816.e13816.e20.
8. S
odersten M, Herteg
ard S, Hammarberg B. Glottal closure, transglottal
airflow, and voice quality in healthy middle-aged women. J Voice. 1995;
9:182197.
9. Holmberg EB, Hillman RE, Perkell JS. Glottal airflow and transglottal air
pressure measurements for male and female speakers in soft, normal, and
loud voice. J Acoust Soc Am. 1988;84:511529.
10. Fabre P. Un procede electrique percuntane dinscription de laccolement
glottique au cours de la phonation: glottographie de haute frequence; premiers resultats. [A non-invasive electric method for measuring glottal

26.

27.
28.

29.

30.

31.
32.
33.
34.

35.
36.

closure during phonation: high frequency glottogr]. Bull Acad Natl Med.
1957;141:6669.
Baken RJ. Electroglottography. J Voice. 1992;6:98110.
Scherer R, Druker D, Titze I. Electroglottography and direct measurement
of vocal fold contact area. In: Fujimura O, ed. Vocal Fold Physiology: Voice
production, Mechanisms and Functions, Vol. 2. New York: Raven Press;
1988:279290.
Orlikoff R. Assessment of the dynamics of vocal fold contact from the electroglottogram: data from normal male subjects. J Speech Hear Res. 1991;
34:10661073.
Rothenberg M, Mahshie JJ. Monitoring vocal fold abduction through vocal
fold contact area. J Speech Hear Res. 1988;31:338351.
Titze IR. A four-parameter model of the glottis and vocal fold contact area.
Speech Comm. 1989;8:191201.
Childers DG, Hicks DM, Moore GP, Alsaka YA. A model for vocal fold
vibratory motion, contact area, and the electroglottogram. J Acoust Soc
Am. 1986;80:13091320.
Awan SN, Awan JA. The effect of gender on measures of electroglottographic contact quotient. J Voice. 2013;27:433440.
Herbst C, Ternstrom S. A comparison of different methods to measure the
EGG contact quotient. Logoped Phoniatr Vocol. 2006;31:126138.
Robb MP, Simmons JO. Gender comparisons of childrens vocal fold contact behavior. J Acoust Soc Am. 1990;88:13181322.
Herbst CT, Fitch WTS, Svec JG. Electroglottographic wavegrams: a technique for visualizing vocal fold dynamics noninvasively. J Acoust Soc Am.
2010;128:30703078.
Cecconello LA. Qualitative and quantitative aspects of the electroglottography in singers with different glottic contact patterns. In: 28th World
Congress of the International Association of Logopedics and Phoniatrics.
Cordoba, Argentina: Sanatorio del Salvador; 2010:779782.
Fant G. Quarterly progress and status report: glottal source and excitation
analysis. STL-QPSR. 1979;20:85107.
Doval B, DAlessandro C, Henrich N. The spectrum of glottal flow models.
Acta Acustica. 2006;92:10261046.
Titze I. Principles of Voice Production. 2nd ed. Salt Lake City, UT: National
Center for Voice and Speech; 2000.
Holmberg EB, Hillman RE, Perkell JS, Guiod PC, Goldman SL. Comparisons among aerodynamic, electroglottographic, and acoustic spectral measures of female voice. J Speech Hear Res. 1995;38:12121223.
Winkler R, Bruckl M, Sendlmeier W. The aging voice: an acoustic, electroglottographic and perceptive analysis of male and female voices. In: 15th
ICPhS. Barcelona; 2003:14.
Bestebreurtje ME, Schutte HK. Resonance strategies for the belting style:
results of a single female subject study. J Voice. 2000;14:194204.
Buder E, Kent R, Kent J, Milenkovic P, Workinger M. Formoffa: an automated formant, moment, fundamental frequency, amplitude analysis of
normal and disordered speech. Clin Linguist Phon. 1996;10:3154.
Forrest K, Weismer G, Milenkovic P, Dougall RN. Statistical analysis of
word-initial voiceless obstruents: preliminary data. J Acoust Soc Am.
1988;84:115123.
Kardach J, Wincowski R, Metz DE, Schiavetti N, Whitehead RL,
Hillenbrand J. Preservation of place and manner cues during simultaneous
communication: a spectral moments perspective. J Commun Disord. 2002;
35:533542.
Harwardt C, Gottsman F, Noubours S. On the relationship between vocal
effort and spectral moments. In: SiMPE 2011 Proceedings.;14.
DeCarlo L. On the meaning and use of kurtosis. Psychol Methods. 1997;2:
292307.
Dytham C. Choosing and Using Statistics: A Biologists Guide. 3rd ed. Oxford, UK: Wiley Blackwell; 2011.
Lowell SY, Colton RH, Kelley RT, Hahn YC. Spectral- and cepstral-based
measures during continuous speech: capacity to distinguish dysphonia and
consistency within a speaker. J Voice. 2011;25:e223e232.
Colton RH, Paseman A, Kelley RT, Stepp D, Casper JK. Spectral moment
analysis of unilateral vocal fold paralysis. J Voice. 2011;25:330336.
Tanner K, Roy N, Ash A, Buder EH. Spectral moments of the long-term
average spectrum: sensitive indices of voice change after therapy? J Voice.
2005;19:211222.

Shaheen N. Awan, et al

Relationship Between EGG CQ, EGG Profile, & Spectral Moments

37. Houtz DR, Roy N, Merrill RM, Smith ME. Differential diagnosis of muscle
tension dysphonia and adductor spasmodic dysphonia using spectral moments of the long-term average spectrum. Laryngoscope. 2010;120:749757.
38. Dromey C. Spectral measures and perceptual ratings of hypokinetic dysarthria. J Med Speech Lang Pathol. 2003;11:8594.
39. Uchanski R, Geers A. Acoustic characteristics of the speech of young
cochlear implant users: a comparison with normal-hearing age-mates.
Ear Hear. 2003;24(1 Suppl.):90S1055S.
40. Vieira MN, McInnes FR, Jack MA. Comparative assessment of electroglottographic and acoustic measures of jitter in pathological voices. J Speech
Lang Hear Res. 1997;40:170182.
41. Henrich N, dAlessandro C, Doval B, Castellengo M. On the use of the derivative of electroglottographic signals for characterization of nonpathological phonation. J Acoust Soc Am. 2004;115:1321.
42. Owen S, Froman R. Focus on qualitative methods uses and abuses of the
analysis of covariance. Res Nurs Health. 1998;21:557562.
43. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed.
Hillsdale N.J: L. Erlbaum Associates; 1988.
44. Ferguson CJ. An effect size primer: a guide for clinicians and researchers.
Prof Psychol Res Pract. 2009;40:532538.
45. Kraemer H, Morgan G, Leech N, Gliner J, Vaske J, Harmon R. Measures of clinical significance. J Am Acad Child Adolesc Psychiatry. 2003;42:15241529.
46. Schuele C, Justice L. The importance of effect sizes in the interpretation of
research: primer on research: part 3. ASHA Lead. 2006.
47. Thalheimer W, Cook S. How to calculate effect sizes from published
research: a simplified methodology. Somerville, MA: A Work-Learning
Research; 2002.
48. Holm SA. Simple sequentially rejective multiple test procedure. Scand J
Stat. 1979;6:6570.
49. Alipour F, Scherer R, Finnegan E. Measures of spectral slope using an
excised larynx model. J Voice. 2012;26:403411.
50. Kreiman J, Shue Y-L, Chen G, et al. Variability in the relationships among
voice quality, harmonic amplitudes, open quotient, and glottal area waveform shape in sustained phonation. J Acoust Soc Am. 2012;132:26252632.
51. Awan SN, Giovinco A, Owens J. Effects of vocal intensity and vowel type
on cepstral analysis of voice. J Voice. 2012;26:670.e15670.e20.

529

52. Awan S, Roy N. Toward the development of an objective index of dysphonia

severity: a four-factor acoustic model. Clin Linguist Phon. 2006;20:3549.
53. Boersma P, Kovacic G. Spectral characteristics of three styles of Croatian
folk singing. J Acoust Soc Am. 2006;119:18051816.
54. Herbst CT, Howard D, Schlomicher-Thier J. Using electroglottographic
real-time feedback to control posterior glottal adduction during phonation.
J Voice. 2010;24:7285.
55. Baer T, Lofqvist A, McGarr N. Laryngeal vibrations: a comparison between
high-speed filming and glottographic techniques. J Acoust Soc Am. 1983;
73:13041308.

56. Herbst CT, Lohscheller J, Svec
JG, Henrich N, Weissengruber G,
Fitch WT. Glottal opening and closing events investigated by electroglottography and super-high-speed video recordings. J Exp Biol. 2014;
217(Pt 6):955963.
57. Miller D, Schutte H. Characteristic patterns of sub- and supraglottal pressure variations within the glottal cycle. In: Transcripts of the XIIIth Symposium: Care of the Professional Voice. New York, NY: The Voice
Foundation; 1984.
58. Schutte H, Miller D. Resonanzspiele der Gesangsstimme in ihren Beziehungen zu supra- und subglottalen Druckverlaufen: Konsequenzen fur
die Stimmbildungstheorie. Folia Phoniatr Logop. 1988;40:6573.
59. Marasek K. EGG and Voice Quality 1997. Available at. http://www.ims.
uni-stuttgart.de/phonetik/EGG/frmst1.htm. Accessed January 12, 2013.
60. Kelman A. Vibratory pattern of the vocal folds. Folia Phoniatr (Basel).
1981;33:7399.
61. Baken R, Orlikoff R. Clinical Measurement of Speech and Voice. 2nd ed.
San Diego, CA: Singular Publishing Group; 2000.
62. Patel S, Scherer KR, Sundberg J, Bjorkner E. Acoustic markers of emotions
based on voice physiology. In: Speech Prosody. Chicago, IL; 2010:47.
63. Sundberg J, Nordenberg M. Effects of vocal loudness variation on spectrum
balance as reflected by the alpha measure of long-term-average spectra of
speech. J Acoust Soc Am. 2006;120:453.
64. Libeaux A. The EGG Spectrum Slope in Speakers and Singers: Variations
related to Voice Sound Pressure Level, Vowel and Fundamental Frequency
[Masters Thesis]. Stockholm, Sweden: KTH Royal Institute of Technology; 2010:155.