Voice Perceptions and Quality of Life

of Transgender People
*Adrienne B. Hancock, †Julianne Krissinger, and ‡Kelly Owen, *yzWashington, District of Columbia

Summary: Despite the plethora of research documenting that the voice and quality of life (QoL) are related, the exact
nature of this relationship is vague. Studies have not addressed people who consider their voice to influence their life and
identity, but would not be considered to have a voice ‘‘disorder’’ (e.g., transgender individuals). Individuals seeking
vocal feminization may or may not have vocal pathology and often have concerns not addressed on the standard
psychosocial measures of voice impact. Recent development of a voice-related QoL measure specific to the needs of
transgender care (Transgender Self-Evaluation Questionnaire [TSEQ]) affords opportunity to explore relationships
between self-perceived QoL and perceptions of femininity and likability associated with transgender voice. Twenty
male-to-female transgender individuals living as a female 100% of the time completed the TSEQ and contributed
a speech sample describing Norman Rockwell’s ‘‘The Waiting Room’’ picture. Twenty-five undergraduate listeners
rated voice femininity and voice likability after audio-only presentation of each speech sample. Speakers also self-
rated their voices on these parameters. For male-to-female transgender clients, QoL is moderately correlated with
how others perceive their voice. QoL ratings correlate more strongly with speaker’s self-rated perception of voice
compared with others’ perceptions, more so for likability than femininity. This study complements previous research
reports that subjective measures from clients and listeners may be valuable for evaluating the effectiveness of treatment
in terms of how treatment influences voice-related QoL issues for transgender people.
Key Words: Voice femininity–Voice likability–Transgender–Quality of life.

INTRODUCTION
Male-to-female transgender adults seek voice therapy to gain
a more gender-appropriate (ie, feminine) voice.1 Transgender in-
dividuals commonly comment that they ‘‘pass’’ until they start
talking; often this obstacle can be minimized with voice and
communication therapy. Feminine voice can be achieved
through many modalities, such as raising fundamental frequency,
changing resonance, and using more feminine nonverbal com-
munication.1,2 The relationship between listener’s perception
of voice and speaker’s quality of life (QoL) is a topic that has
been evaluated for various populations with voice disorders;
however, this research does not exist for the male-to-female
transgender population seeking vocal feminization.1,3 The
relationship between a transgender individual’s voice (as rated
by others and themselves) and the impact it has on their
physical, social, and emotional well-being remains unclear.
Voice-related quality of life
Several studies have been conducted regarding the correlation be-
tween QoL and how one’s voice is judged by others, demonstrat-
ing that the two measures are intricately related.3–5 For example,
Murry et al (2004) found a correlation between Voice-Related
Quality of Life (V-RQOL) and perceived voice quality (as
judged by the Grade Roughness Breathiness Asthenic Strain
(GRBAS) perceptual scale).3 This and other studies were
performed with participants who were diagnosed with unilateral
Accepted for publication July 27, 2010.
From the *Department of Speech and Hearing Sciences, The George Washington
University, Washington, District of Columbia; yAegis Therapies, Washington, District
of Columbia; and the zDepartment of Speech-Language Pathology, National Rehabilita-
tion Hospital, Washington, District of Columbia.
Address correspondence and reprint requests to Adrienne B. Hancock, Department of
Speech and Hearing Sciences, The George Washington University, 2115 G Street NW,
Suite 201, Washington, DC 20052. E-mail: hancock@gwu.edu
Journal of Voice, Vol. 25, No. 5, pp. 553-558
0892-1997/$36.00
Ó 2011 The Voice Foundation
doi:10.1016/j.jvoice.2010.07.013
vocal fold paralysis (UVFP), benign vocal fold lesions, muscle
tension dysphonia, dysphonia related to aging, and vocal fold
granulomas.3,5–9 Most studies examining interactions of voice
and QoL use scales judging the perceptual qualities of the
voice (eg, breathiness, grade, roughness, asthenia, and strain)
or acoustic analysis using computer systems to measure
fundamental frequency, intensity, frequency, and intensity
ranges, and perturbation analysis.
In these studies, the strength of the relationship varied by the
voice disorder under study as well as the voice quality and QoL
measures. For example, Karnell et al (2007)4 found a strong re-
lationship (r ¼ 0.80) between measurements of QoL and voice
quality using the V-RQOL, the Concensus Auditory Perceptual
Evaluation for Voice (CAPE-V), and GRBAS, but included a va-
riety of vocal disorders (ie, degenerative, inflammatory, mass
changes, muscle tension dysphonia, neuromotor). In a study
specific to patients with adductor spasmodic dysphonia, Liu
et al7 found that patients had more anxiety, depression, and
poor life quality than the control subjects. The patient’s emo-
tional status and life quality improved after botulinum toxin
treatment. As vocal quality improved, anxiety levels decreased
on the Hamilton Anxiety Rating Scale (r ¼ 0.799), but the rela-
tionship with QoL was not statistically significant (r value not
reported). Schindler et al (2008)5 analyzed voice quality
changes before and after voice therapy in a group of patients
with UVFP. Both Grade Instability Roughness Breathiness
Asthenic Strain (GIRBAS) scores and Voice Handicap Index
(VHI) scores improved after behavioral voice therapy; how-
ever, the relationship between the perceptual and well-being
measures was not reported. Despite the plethora of research
documenting that the voice and QoL are related, the exact
nature and degree of this relationship appears to have multiple
factors. Notably absent from this literature are populations who
do not necessarily have voice disorders (eg, transgender) but
consider their voice to influence their life and identity.
554
Measuring voice-related quality of life
Rating scales have been developed to provide subjective mea-
sures of voice disorder severity in terms of self-perceived dis-
ability and handicap resulting from voice disorders; most
rating scales have normative data, at least for some popula-
tions.1,10 This study will use the VHI and the Transgender
Self-Evaluation Questionnaire (TSEQ)1 to measure transgender
individuals’ QoL, specifically concerning the influence of their
voices. The VHI measures functional, emotional, and physical
aspects of voice with underlying assumption that there is an ex-
isting vocal pathology. The VHI has been shown to be a valid
tool used to assess psychosocial handicap of voice disorders rep-
resenting the breadth of pathology seen in most clinical settings.
Using 63 patients on two separate occasions, Jacobson and
White11 found the VHI to exhibit strong internal reliability
and strong test-retest reliability. Although construct validity
was not entirely examined in the study by Jacobson and White
(1993), a moderately strong relationship was found between
the patient’s self-perceived severity and VHI scores.10,12
Individuals seeking vocal feminization may or may not have
vocal pathology and often have concerns not addressed on the
VHI (eg, my laughing and sneezing sound like a man).
Andrews13 emphasizes the relationship between perceived
voice femininity/masculinity and QoL for transgender individ-
uals; yet, only recently has one questionnaire been developed
specific to the needs of transgender care. The Transgender
Self-Evaluation Questionnaire (TSEQ) is a subjective measure
of voice handicap tailored to the transgender population. The
TSEQ is standardized; however, it does not have normative
data at this time. Hancock et al (2009)14 demonstrated a signif-
icant relationship (r ¼ 0.89) between VHI and TSEQ scores
when administered to male-to-female transgender individuals
presenting as females 100% of the time, indicating criterion val-
idity of the TSEQ. Additionally, a strong correlation (r ¼ 0.97)
indicated test-retest reliability of the TSEQ.14 Although the VHI
and TSEQ may be correlated, they are not identical; therefore,
using a measure with greater content validity will provide the
clinician with a richer picture of the client’s feelings and may
be helpful in directing treatment.
Measuring voice
QoL ratings (TSEQ, VHI); voice quality perception scales
(GRBAS, CAPE-V); and nonstandardized interval scales are
used to document how an individual feels about his or her own
voice and how listeners perceive it.6,15 Subjective ratings of
voice quality perception can be administered through informal
assessment, observation, questionnaire, or standardized
subjective rating scales. Standardized scales allow listeners to
rate voice perception on various vocal qualities, including
grade, harshness, breathiness, strain, and asthenic. Interval
scales, such as those used in this study, ask listeners and
speakers to rate the voice on a sliding scale anchored with
certain qualities (in this study, femininity/masculinity and like/
dislike). When properly designed, interval scales can be
reliable, valid, and sensitive.
The listener is another factor to consider when reporting
perceptual data. McNeill et al (2008)15 used interval scales to
Journal of Voice, Vol. 25, No. 5, 2011
measure listener perception of transgender voice and compared
ratings by speech-language pathologists (SLPs) with those of
naive listeners. Ratings of overall quality by naive listeners
and SLP listeners were strongly correlated (r ¼ 0.97). Naive
judges had more variability in their ability to judge, but their
scores were still consistent with the SLP judges.
Voice perception and quality of life
Transgender individuals’ perception of their own voice quality
is an essential aspect of a successful transition. McNeill et al
(2008)15 measured the correlation between a speaker’s self-
ratings of voice femininity and happiness. The perception trans-
gender individuals had of the femininity of their own voice
showed a moderate correlation with their feelings of happiness
with their voice (r ¼ 0.55). As the study by McNeill et al
(2008)15 demonstrated a potential link between self-perceived
femininity and happiness in regard to voice, there may be mea-
sures of listener perceptions that relate to the transgender
speaker’s QoL.
In general, a relationship between voice quality and self-
perception of QoL is fairly robust in the literature and among
expert opinion, but further exploration is needed with transgen-
der clients to examine whether others’ perceptions of their
voice is associated with patients’ QoL reports. External lis-
teners’ and observers’ perception of feminine voice is helpful
in completing the male-to-female transition and feeling truly
accepted in society.9 Previous research has emphasized that so-
ciety’s negative and prejudiced view of transgender individuals
influences the person’s view of him- or herself.16 In one study of
transgender speakers, QoL moderately correlated with how
feminine voice is perceived by SLPs (r ¼ 0.46) and naive lis-
teners (r ¼ 0.45).15 Although feminine voice and communica-
tion is a prominent goal for many transgender individuals,
voice likability is a factor to consider in any voice therapy.
A voice perceived as feminine may improve one’s self-
perception of life quality, but if the feminine voice comes
with a cost of an unpleasant sound or a voice disorder, the over-
all perception of life may decrease because of the discomfort of
the voice and/or the decreased likability of their voice quality.
As we explore the contributing factors to QoL for transgender
individuals, it may be valuable to distinguish voice femininity
from voice likability.
Likability is a difficult term to narrowly define and does not
exist as a single unit. According to Kent and Ball (2000),17 there
seems ‘‘no viable definitions . about ‘ideal’ quality have
emerged from efforts of speech pathologists as a group.’’ How-
ever, one can speculate that if there is not an ‘‘ideal’’ voice, then
there at least may be a preferred or less preferred voice. For
example, Hollien et al (1991),18 using 80 speaking subjects
(40 females and 40 males), found that listeners of speech sam-
ples ‘‘liked’’ voices that were of relatively lower frequencies
and intensities.
Purpose
Correlations between voice perceptions, as judged by external
listeners, and QoL for transgender persons has been investi-
gated in only one study in Europe, which only measured
Adrienne B. Hancock, et al Voice and Quality of Life
TABLE 1.
Demographic Data for 20 MTF Speakers With Mean and Standard Deviation
Demographics
Age (years)
% of participants with completed sexual reassignment surgery
Time presenting as female any % of the day (years)
Time presenting as female 100% of the day (years)
Duration of voice treatment (months)
Time since treatment (months)
TSEQ*
* TSEQ scores range from 30 to 150, with higher score indicating greater voice handicap.
other- and self-perceived ‘‘femininity’’ as it correlated with
‘‘happiness.’’15 One may predict that the femininity and likabil-
ity of voice have significant impacts on male-to-female trans-
gender individuals’ acceptance in society and of their own
QoL; however, these variables have not yet been thoroughly
explored using standard measures of QoL. Therefore, the
correlation between QoL and perception of male-to-female
transgender voices’ femininity and likability is a relationship
that will be examined in this study. Furthermore, resulting
data could inform us whether it may be appropriate to use
self and naive-listener perception rating scales of femininity
and likability of voice in addition to established treatment out-
comes, such as increased fundamental frequency, increased
feminine intonation patterns, and forward oral resonance. The
purpose of this study was to explore voice likability, femininity,
and QoL for transgender individuals. There are three primary
research questions:
1. Is there a relationship between transgender speaker’s
TSEQ and self-ratings of voice (likability and feminin-
ity)?
2. Is there a relationship between transgender speaker’s
TSEQ and listener ratings of voice (likability and
femininity)?
3. Is femininity of voice related to likability of voice
a. when rated by the speaker?
b. when rated by the listener?
METHOD
Subject population
Speakers. Twenty male-to-female transgender individuals
(mean age, 45.8 years; standard deviation [SD] ¼ 10.6; range,
23–63) were recruited through phone calls, flyers, or word of
mouth. Exclusion criteria included self-reported diagnosis of
hearing loss and/or history of mental, psychological, or cogni-
tive disorder (with the exception of gender identity disorder and
depression). All participants received vocal feminization treat-
ment from speech-language pathology graduate students, under
the supervision of SLPs certified by the American Speech-
Language and Hearing Association, at the George Washington
University Speech and Hearing Center (GWUSHC). Three
555
Mean Standard Deviation
45.8 12.6
45 (9/20)
8.9 7.8
4.6 3.6
11 6
2.6 2.2
68.8 27.7
participants were currently enrolled in treatment at GWUSHC;
the remaining 17 received vocal feminization therapy at
GWUSHC within the last 7 years. Therapeutic techniques
were similar across participants with regard to vocal feminiza-
tion (ie, raising fundamental frequency and using feminine in-
tonation and resonance patterns). Participants were in various
stages of the transition process; however, all were living as a fe-
male 100% of the time (eg, consistently used a feminine voice
during life) at the time of testing. Participants completed a med-
ical history questionnaire to gather additional information
(Table 1).
Additionally, five cisgender women aged 30–58 years (mean
age, 46.8 years; SD ¼ 10.6) and five cisgender men aged 35–52
years (mean age, 40.8 years; SD ¼ 7.19), who did not self-
identify as transgender, participated. These speakers self-
reported no prior history of gender identity disorders, no history
of voice therapy, and no report of diagnosed hearing loss. All
speakers received $10 for their participation in this study.
Listeners. Listeners included 12 men and 13 women between
the ages of 18 and 22 years, recruited from undergraduate
liberal arts courses at GWU. Mean age was 18.8 years
(SD ¼ 0.99) for male listeners and 19.65 years (SD ¼ 1.45)
for female listeners. Listeners received $10 for participation.
Materials
Demographic questionnaire. All participants completed
demographic questionnaires to document potentially influential
factors, such as age, gender, psychological history, and aca-
demic major. Speaker questionnaires also included questions
with regard to duration of voice therapy received (months),
time since treatment (months), whether they had received sex-
ual reassignment surgery, the percent of time presenting as fe-
male for a percentage of the day (years), and the time
presenting as female 100% of the day (years) (Table 1).
Speech sample. A speech sample was collected from the 20
male-to-female transgender participants, five cisgender fe-
males, and five cisgender males. Audio recordings were re-
corded with the Computerized Speech Lab (CSL; Kay Pentax,
NJ) software with a Shure SM48 microphone stand 4 inches
from the client’s lips. All audio recordings were made in the
GWUSHC in a quiet room with minimal ambient noise. Each
participant was asked to describe aloud Norman Rockwell’s
556
‘‘The Waiting Room’’ for 20–25 seconds.19 After collection of
all participants’ picture descriptions, each sample was assessed
for length (at least 20 seconds), sound quality (no laughing or
coughing), on-topic language, and appropriate theme (ie, did
the speaker actually describe the picture?).
Speaker self-perception measures. Rating scales were
presented using a computer-presented visual analog scale
(VAS); each rated how masculine or feminine they perceived
their own voice. Clinicians instructed the participants to move
the cursor from midline toward the anchor word on VAS (‘‘mas-
culine’’/‘‘feminine’’ and ‘‘dislike’’/‘‘like’’) that described their
perception. No numbers were visually presented on the scale;
however, the computerized scoring scale ranged from 0 to
1000. this was to ensure a highly sensitive measure. The VAS
is a common tool used in medical and speech-language pathol-
ogy research.15,20–23
Speakers were instructed to evaluate their voice based on
their own perceptions of their voice and not from the recording.
They completed two scales to measure subjective impression of
their voice. Each speaker first marked how feminine he or she
felt his or her voice was on a scale of 0 (very masculine) to
1000 (very feminine). The second scale measured how likable
his or her voice was (0: dislike; 1000: like).
Listener perception measures. Listeners were presented
the recorded speech samples, described earlier, in a quiet
room in the GWUSHC with minimal ambient noise. Record-
ings were presented in a sound field by means of speakers 2 ft
from the listener and a comfortable listening loudness as
determined by the listener. The listeners were shown
computer-presented VAS identical to the one the transgender
speaker participants were shown. They used the cursor on the
VAS to record their perception of the speaker’s femininity/mas-
culinity and if they liked/disliked the voice after hearing each
20-second speech sample. Listeners were only able to listen
to each sample once and had as much time as necessary to
rate the voice. Pilot testing demonstrated that 20 seconds was
an adequate amount of time to make judgments of a voice pre-
sented by audio only.
Procedures
All participants gave written consent and were assured of con-
fidentiality and anonymity measures. The Institutional Review
Board (IRB) at GWU approved all materials and procedures
(IRB #050834). There were two phases to this study. First,
the demographic information, picture descriptions, and self-
ratings were collected from all of the speakers. Second, the lis-
teners rated the voice femininity and likability of each sample.
In the first phase with the speakers, all assessments were con-
ducted by a graduate clinician and completed in a single, 1-hour
session. Twenty-second segments of sound files were acousti-
cally collected, analyzed, and edited using a Kay Pentax CSL
software. Twenty transgender individuals, five cisgender males,
five cisgender females, and six random samples, repeated for
intrarater reliability (36 in total), were randomly compiled for
the listening phase of the study. To control for possible order
effects, the speech samples were counterbalanced in split half
Journal of Voice, Vol. 25, No. 5, 2011
order (ie, half of the listeners were presented samples 19–36
before 1–18).
The second phase consisted of the listeners using the VAS
rating scales, described earlier, to judge the femininity and
likability of the speakers’ voice. During a 45-minute session
at the GWUSHC, listeners were presented all 36 samples and
were allowed to ask for a break when needed. Listeners were
not told the gender identity of the speakers until the conclusion
of the study. At the completion of the study, the listeners
were debriefed. After the debriefing, no listeners objected to
participation.
RESULTS
Listener reliability and bias
To examine listener intrarater reliability in femininity and in
likability scoring, the difference between a listener’s two rat-
ings of each repeated sample was calculated. Criterion for reli-
ability was set at an average difference of 25% (eg, 250 out of
a possible 1000-point score) across six samples. All listeners
showed intrarater reliability for femininity, and one rater was
excluded from analysis involving likability because he did
not meet this criterion.
Half of the listeners were presented with the samples in linear
order (1–36), and half the listeners heard the samples in split
half order (19–36 and 1–18). The results of this one-way mul-
tivariate analysis of variance were not statistically different,
indicating that there was no statistical difference between
the ratings from the two order presentations (femininity:
F ¼ 2.888 (1,23), P ¼ 0.103; likability: F ¼ 0.003 (1,23),
P ¼ 0.954) In addition, there was no significant difference be-
tween male and female listeners for ratings (femininity:
F ¼ 0.255 (1,23), P ¼ 0.619; likability: F ¼ 0.000 (1,23),
P ¼ 0.999). After statistics confirmed that listener’s gender
and the order of presentation likely did not influence ratings,
listener data were pooled and analyzed as one group.
Voice perceptions and quality of life
Descriptive statistics for the perceptual measures are shown in
Table 2. To answer the stated research questions, QoL mea-
sures, self-ratings of femininity and likability, and listener rat-
ings of femininity and likability were analyzed using Pearson
TABLE 2.
Descriptive Data for Perceptual Measures
Standard
Perceptual Measure Mean Deviation Range
Self-ratings
Femininity of own voice 529 254 10–990
Likability of own voice 552 238 9–978
Listener ratings
Femininity of voice 493 243 94–755
Likability of voice 533 112 340–398
Perceptual scores range from 0 to 1000, with higher score indicating
greater femininity/likability.
Adrienne B. Hancock, et al Voice and Quality of Life
TABLE 3.
Correlations Between the TSEQ and Perceptual Ratings
Pearson
Correlation
Coefficient (r) P Value
Self-ratings
Femininity of own voice 0.454* 0.044
Likability of own voice 0.615* 0.004
Listener ratings
Femininity of voice 0.375 0.103
Likability of voice 0.311 0.182
As femininity and likability increase, TSEQ score decreases (ie, QoL ratings
are higher).
* Statistically significant at the a ¼ .05 level.
product-moment correlations. Correlation coefficient (r) is re-
ported for each question (Table 3). As reported by Sheskin
(2003),24 0 < r < 0.3 was considered weak, 0.3 < r < 0.7 was
considered moderate, and 0.7 < r < 1.0 was deemed strong.
Alpha was at 0.05 for all analyses.
Voice-related QoL for the transgender (TG) speakers was
measured using the TSEQ. TSEQ scores range from 0 to 150,
with a lower score indicating better QoL. Moderate inverse re-
lationships were established between scores on the TSEQ and
speaker’s self-ratings of femininity of the voice (r ¼ 0.454,
P ¼ 0.044) and likability (r ¼ 0.615, P ¼ 0.004). This means
that 20% and 38% of the variance in TSEQ score can be attrib-
uted to self-ratings of femininity and likability, respectively.
Moderate, but relatively weaker and statistically nonsignificant,
inverse relationships were also found between the TSEQ and
listener ratings of femininity of the voice (r ¼ 0.375,
P ¼ 03) and likability (r ¼ 0.311, P ¼ 0.182). Therefore,
14% and 9% of the variance in TSEQ scores can be attributed
to listener perception of the voice. QoL scores improved as
the voice was rated more feminine or more likable; this relation-
ship with QoL was stronger for the self-ratings compared with
the listener ratings. As for the relationship between femininity
and likability of the voice, a moderately strong positive rela-
tionship was found between the speaker’s self-rated likability
and femininity of the voice (r ¼ 0.704, P ¼ 0.001), but for lis-
teners, there was no relationship between femininity and lik-
ability ratings of the same sample (r ¼ 0.047, P ¼ 0.843).
For analysis comparing self-ratings and listener ratings, please
refer to Owen and Hancock (in press).
DISCUSSION
We set out to explore how femininity and likability, as rated by
speakers and listeners, may be related to male-to-female trans-
gender individuals’ QoL. We examined these relationships to
investigate the potential use of likability and femininity rating
scales in diagnosis and treatment outcomes for transgender peo-
ple. Our findings are similar to those in the study of McNeill
et al (2008)15; they reported a moderate correlation (r ¼ 0.45)
between the transgender speaker’s self-rated happiness with
regard to voice as rated on a 1–10 scale and naive listeners’
557
perceptions of the voice’s femininity, whereas our correlation
for similar measures was r ¼ 0.375. As in their study, QoL cor-
related relatively more strongly with speakers’ ratings than
those of the listeners.
The study by McNeill et al (2008)15 and the present study are
comparable; however, this study’s findings expanded the con-
clusions of McNeill et al. The same moderate correlation was
found; however, perceptual measures in the present study
were more sensitive and also addressed likability. Furthermore,
a QoL measure specific to voice and transgender concerns (ie,
TSEQ) was used to provide a more accurate representation of
how the speaker feels. This study concurred with the conclu-
sions of McNeill et al (2008)15 but with different culture and
measures, thereby providing further support for use of percep-
tual measures to gauge psychosocial impact of transgender
voice therapy. Furthermore, similarities across cultures are en-
couraging for external validity or ability to generalize these
findings to other cultures. It is critical that we evaluate self
and listener perceptions and their role in transgender voice fem-
inization therapy. Further research regarding the TSEQ would
allow us to examine and evaluate QoL as we do for any other
voice population.
As we explore the contributing factors to QoL for transgen-
der individuals, we thought it valuable to distinguish voice fem-
ininity from voice likability. It appears that it may depend on
whom you ask. Listeners—who did not know they were listen-
ing to male-to-female (MTF) speakers—did not relate feminin-
ity to likability, whereas TG speakers—who had or currently
were enrolled in voice therapy—did relate how feminine they
perceived their voice to their voice’s likability. In this study, lis-
teners were blinded to the nature of the voices they were judg-
ing: If they had known they were judging male-to-female
transgender voices, would their judgments of femininity and
likability change? The speakers knew the desired voice quality
and the listener did not.
Voice feminization therapy for transgender individuals is
efficacious; the perception of a female voice can be enhanced
by compensatory techniques, hormone replacement therapy,
and surgical procedures that manipulate aerodynamic and phys-
iological processes of the voice.2,25–28 This study complements
previous research reports that subjective measures from clients
and listeners may be valuable for evaluating the effectiveness of
treatment in terms of how treatment influences voice-related
QoL issues for transgender people. A transgender client’s voice
may have reached appropriate acoustic parameters for gender
norms; however, would one discharge treatment if the client
did not ‘‘like’’ his or her voice or if his or her QoL continued
to be negatively impacted by his or her voice?
Future directions
Questions still remain regarding these complex relationships
between various perceptions of the transgender voice and
one’s feelings of QoL. We found that perceptual measures of
speaker self-rated and listener-rated likability of the voice cor-
relate with QoL measures but to different degrees. This gener-
ates questions in terms of why likability of the voice correlates
to transgender persons’ QoL and brings us back to consider
558
what creates the ‘‘ideal’’ voice. As literature continues to ex-
plore characteristics of the ideal voice, we can use the results
of this study to guide future research suggesting the voice’s re-
lationship with QoL. We know the various physiological and
aerodynamic components that contribute to voice production;
however, it is still unclear whether certain acoustic or aerody-
namic parameters are more indicative of perceptual and psy-
chosocial measures of transgender voice.
Some research had explored the relationship between those
objective measures and voice perceptions. For example, a posi-
tive correlation was found between speech frequency and voice
femininity.1,2,9 There are many considerations that impact
a transgender individual’s QoL, voice quality, and how others
perceive their voice quality.1,9 As listeners perceive an
individual’s voice, they are taking into account voice
characteristics that are influenced by acoustic (eg, pitch) and
aerodynamic (eg, breathiness) parameters of voice quality
that are interdependent.27,29 It is critical to consider these
variables for male-to-female transgender individuals, because
they are the aspects external listeners would be judging during
auditory perceptual assessment. It is possible that one or few of
these variables are more predictive of voice judgment, and, in
turn, may be more related to speaker’s QoL.
Although finding a relationship between speech frequency
and voice femininity provides targets for voice feminization
for male-to-female transgender individuals, it is important to
consider the implications of such findings in light of the present
study. If transgender individuals’ QoL is likely to correlate with
their self-rating of their voice likability, then perhaps we should
incorporate not only objective (ie, frequency) targets but also
subjective measures (ie, perceptual scales and QoL measures)
in vocal feminization goals.
CONCLUSION
For male-to-female transgender clients, QoL is correlated with
how others perceive her voice. QoL ratings correlate more
strongly with speaker’s self-rated perception of voice compared
with others’ perceptions and, more so, for likability than feminin-
ity. These results support using likability and femininity ratings
to supplement objective measures of treatment effectiveness.
REFERENCES
1. Adler RK, Hirsch S, Mordaunt M. Voice and Communication Therapy for
the Transgender/Transsexual Client: A Comprehensive Clinical Guide. San
Diego, CA: Plural Publishing; 2006.
2. Gelfer MP. Voice treatment for the male-to-female transgendered client. Am
J Speech Lang Pathol. 1999;8:201–208.
3. Murry T, Medrado R, Hogikyan ND, Aviv JE. The relationship between rat-
ings of voice quality and quality of life measures. J Voice. 2004;18:
183–192.
4. Karnell MP, Melton SD, Childes JM, Coleman TC, Dailey SA,
Hoffman HT. Reliability of clinician-based (GRBAS and CAPE-V) and
patient-based (V-RQOL and IPVI) documentation of voice disorders.
J Voice. 2007;21:576–590.
Journal of Voice, Vol. 25, No. 5, 2011
5. Schindler A, Bottero A, Capaccio P, Ginocchio D, Adorni F, Ottaviani F.
Vocal improvement after voice therapy in unilateral vocal fold paralysis.
J Voice. 2008;22:113–118.
6. Kasama ST, Brasolotto AG. Vocal perception and life quality. Pro Fono.
2007;19:19–28.
7. Liu CY, Yu JM, Wang NM, et al. Emotional symptoms are secondary to the
voice disorder in patients with spasmodic dysphonia. Gen Hosp Psychiatry.
1998;20:255–259.
8. Ohlwein S, Kruse E, Steiner W, Kiese-Himmel C. Voice function and qual-
ity of life—patients with laryngeal carcinoma following minimal-invasive
laser surgery and function voice rehabilitation. Laryngorhinootologie.
2005;84:253–260.
9. Van Borsel J, De Cuypere G, Van den Berghe H. Physical appearance and
voice in male-to-female transsexuals. J Voice. 2001;15:570–575.
10. Jacobson BH, Johnson A, Grywalski C, Silbergliet A, Jacobson G,
Benninger MS, Newman CW. The Voice Handicap Index: development
and validation. Am J Speech Lang Pathol. 1997;6:66–70.
11. Jacobson BH, White J. Multidisciplinary approach to treatment. In:
Benninger MS, Jacobson BH, Johnson AF, eds. Vocal Arts Medicine: The
Care and Prevention of Professional Voice Disorders. New York, NY:
Thieme Medical Publishers; 1993:318–343.
12. Webb AL, Carding PN, Deary IJ, MacKenzie K, Steen IN, Wilson JA.
Optimising outcome assessment of voice interventions, I: reliability and
validity of three self-reported scales. J Laryngol Otol. 2007;121:763–767.
13. Andrews M. Manual of Voice Treatment: Pediatrics Through Geriatrics.
2nd ed. San Diego, CA: Singular Publishing Group; 1999.
14. Hancock A, Owen K, Siegfriedt L, Brundage S. Reliability and validity of
the transgender self-evaluation questionnaire. Paper presented at: Voice
Foundation’s Annual Symposium; 2009; Philadelphia, PA.
15. McNeill EJM, Wilson JA, Clark S, Deakin J. Perception of voice in the
transgender client. J Voice. 2008;22:727–733.
16. Lewis C. Gender: a state of body or a state of mind? GEMSNEWS. 1994;
17:45.
17. Kent R, Ball M. Voice Quality Measurement. San Diego, CA: Singular
Publishing; 2000.
18. Hollien H, Gelfer MP, Carlson T. Listening preferences for voice types as
a function of age. J Commun Disord. 1991;24:157–171.
19. Rockwell N (Painter). The waiting room [online image]. Retrieved October
1, 2008, from Google Images. Available at: http://sladept.com/artwork/
5575.
20. Campbell W, Lewis S. Visual analogue measurement of pain. Ulster Med J.
1990;59:149–154.
21. Grant S, Aitchison T, Henderson E, Christie J, Zare S, McMurray J,
Dargie H. A comparison of the reproducibility and the sensitivity to change
of visual analogue scales, Borg scales, and Likert scales in normal subjects
during submaximal exercise. Chest. 1999;116:1208–1217.
22. Rammstedt B, Rammsayer T. Gender differences in self-estimated intelli-
gence and their relation to gender-role orientation. Eur J Pers. 2002;16:
369–382.
23. Van Borsel J, Van Eynde E, De Cuypere G, Bonte K. Feminine after crico-
thyroid approximation? J Voice. 2008;22:379–384.
24. Sheskin D. Handbook of Parametric and Nonparametric Statistical Proce-
dures. 3rd ed. Boca Raton, LA: Chapman & Hall/CRC; 2003.
25. Owen K, Hancock A. The role of self- and listener perceptions of femininity
in voice therapy. International Journal of Transgenderism. [in press].
26. De Bruin MD, Coerts MJ, Greven AJ. Speech therapy in the management of
male-to-female transsexuals. Folia Phoniatr Logop. 2009;52:220–227.
27. Gorham-Rowan M, Morris R. Aerodynamic analysis of male-to-female
transgender voice. J Voice. 2006;20:251–262.
28. Wolfe VI, Ratusnik DL, Smith FH, Northrop GE. Intonation and fundamen-
tal frequency in male-to-female transsexuals. J Speech Hear Disord. 1990;
55:43–50.
29. Krom G. Consistency and reliability of voice quality ratings for different
types of speech fragments. J Speech Hear Res. 1994;7:985.