Content uploaded by Anika S. Smeijers
Author content
All content in this area was uploaded by Anika S. Smeijers on Dec 30, 2016
Content may be subject to copyright.
Scienti c-Based Translation of Standardized
Questionnaires into Sign Language of the
Netherlands
Anika. S. Smeijers, MD, Beppie van den Bogaerde,
M. Ens-Dokkum, and A. M. Oudesluys-Murphy
ABSTRACT
Specialized psychological and psychiatric health care for deaf and
hard of hearing clients has emerged during the last 50 years. It has
long been known that deaf and hard of hearing clients are often mis-
diagnosed with psychiatric disorders, but little scienti c attention has
been paid to the tests used with this group. Although these clients
may have poor spoken-language skills and a different (cultural) back-
ground from mainstream clients, regular diagnostic tests are used even
in specialized settings.
To enable the use of standardized questionnaires without language
barriers, we have developed a guideline for adapting internationally
validated questionnaires and translating them into sign language. We
used this guideline to adapt and translate four questionnaires into Sign
Language of the Netherlands (Nederlandse Gebarentaal, NGT). In
this chapter we introduce our guideline and describe the selection and
translation process of research instruments for use with deaf and hard
of hearing individuals. The problems, dilemmas, and ethical issues
encountered are discussed.
One in a thousand people worldwide is born deaf or severely hard
of hearing (Kennedy & McCann, 2004), and this number increases
with age. These children face many challenges in acquiring the spoken
and written language of their environment. Many of them have great
dif culty in achieving a literacy level comparable to that of their peers
(Musselman, 2005). One of the reasons for a prevalence of low literacy
among deaf people is that many Western languages have an alpha-
betical writing system (consisting of letters or graphemes) that is based
: S .
on phonemes. Phonemes are the sounds (e.g., vowels, consonants)
of a language that change the meaning of words (e.g., hat versus bat
[h/b distinction] versus hit [a/i distinction]). When you cannot hear
these phonemes, you have to memorize—for all words—which let-
ters/graphemes represent them (e.g., what combination of characters
and in which order they are used for a certain concept). An English
example would be the verb “know,” of which the visual image on the
mouth resembles the pronunciation of the word “no”; these two writ-
ten forms of the sound [no] thus need to be explicitly learned. Deaf
people also cannot automatically use vocalizations when reading. This
means that deaf people are able to read uently only those words that
they have read before and whose written graphemic construction they
have memorized.
For many deaf and hard of hearing individuals a sign language is their
natural language because they have full access to it. In the Netherlands,
Sign Language of the Netherlands (Nederlandse Gebarentaal, NGT)
is used. In contrast to certain other sign languages like American Sign
Language and Flemish Sign Language, NGT up to now has not been
recognized by the Dutch government as an of cial language of the
Netherlands.
Over the past 25 years there have been discussions about whether deaf
people have more in common with each other than just their medical
condition (in casu their hearing status) plus the fact that many of them
are sign language users. Researchers from Great Britain (Ladd, 2003)
and the United States (Padden & Humphries, 1988, 2005) have convinc-
ingly defended the existence of a Deaf culture. These studies show that
the Deaf communities constitute social and linguistic minorities within
many Western hearing cultures. Such a Deaf cultural minority is charac-
terized by shared experiences, values, traditions, behavioral rules, and,
most important, the use of a sign language as the main mode of com-
munication. Having a cultural identity different from that of the majority
in society may have a negative effect on communication and healthcare
provision (Van Wieringen, Harmsen, & Bruijnzeels, 2002). To distinguish
between the audiological concept of deaf and the cultural Deaf, a capital
letter is used for the latter. The designation “Deaf” is used here to include
people who see themselves as culturally Deaf and as belonging to a lin-
guistic minority group. The designations “deaf” and “hard of hearing”
are used here for people who developed a hearing loss, including Deaf,
deaf, and hard of hearing.
Scienti c-Based Translation of Standardized Questionnaires :
In the Netherlands, as in many other Western countries, there are lim-
ited facilities for Deaf people to ensure their full participation in the wider
society. For instance, sign language support by interpreters is restricted,
and medical information in NGT is scarce. In contrast to the United
States, we in the Netherlands do not have a disability act. An interna-
tional UN convention handles the rights of people with a disability, but
the Dutch government has not yet rati ed this convention. These aspects,
together with lack of access to the spoken language (e.g., no incidental
learning), explain why deaf and hard of hearing individuals often have
less general and medical knowledge than hearing people (Barnett, 1999;
Jones, Renger, & Firestone, 2005; Vernon& Andrews, 1990).
MENTAL HEALTH CARE
Little information is available on the impact of cultural and linguistic
barriers on the medical care offered to and received by deaf and hard of
hearing individuals (Smeijers & Pfau, 2009; Van Wieringen etal., 2002).
The incidence of psychological problems is higher among these individu-
als than among the hearing population (Fellinger etal., 2005b, 2007).
This is partially explained by the fact that people with severe hearing
impairment often face social barriers due to communication problems.
Another explanation is that this might be a result of the two to three
times greater prevalence of sexual abuse (possibly due to communication
barriers and poorer social skills) in this group than in the hearing popula-
tion (Hoem Kvam, 2004).
The incidence of psychiatric disorders in deaf and hard of hearing
individuals is an important subject of discussion in the medical litera-
ture. In the 1950s deaf and hard of hearing individuals in the United
States were relatively overrepresented in psychiatric clinics (Pollard,
1994; Stein, Mindel, & Jabaley, 1981). These ndings ultimately led to
the development of specialized mental health care for deaf and hard of
hearing patients. Within these specialized settings extra attention is given
to the patients’ cultural background and language skills. After the trans-
fer of patients to these specialized facilities, it became evident that many
of them had no psychiatric disorder. Some had a mild cognitive impair-
ment, and some had a severe language impairment caused either by weak
language skills, rst-language deprivation, or a primary language dis-
order. Although nonspecialized clinics still report a higher incidence of
: S .
psychiatric disorders within the deaf and hard of hearing population, no
evidence for this has been found in specialized settings (Pollard, 1994).
Since the emergence of specialized healthcare facilities for deaf and hard
of hearing individuals is only recent, the number of research instruments
especially developed for or adapted to deaf and hard of hearing individu-
als is still low.
Psychological Tests
Psychological testing of deaf and hard of hearing sign language users
is usually done via written questionnaires. However, as explained ear-
lier, using the written language of the hearing minority may be problem-
atic in this population because its members have an inadequate mastery
of the local spoken/written language (Musselman, 2005). More often
than not these questionnaires are translated ad hoc by sign language
interpreters because very few test instruments are available in sign lan-
guages (Munro & Rodwell, 2009). If an interpreter interprets a written
questionnaire ad hoc into sign language, the interpreter may make non-
standardized linguistic and cultural adaptations. Such adaptations can
of course in uence the replies to the questionnaire and thus the general
outcomes. In most situations the interpreter will also interact with the
deaf or hard of hearing test participant who is lling out the question-
naire. A common pitfall is that the results of questionnaires that were
administered by ad hoc interpreters are analyzed as if they are standard-
ized questionnaires, while the questionnaire was actually converted into
a sort of interview.
Research on ethnic minority groups has demonstrated that the mental
and physical well-being of the group members is in uenced by their
cultural identity (Guillemin, Bombardier, & Beaton, 1993). Therefore,
health-related quality of life (HRQoL) questionnaires can be used for
people whose linguistic and cultural identity are different from that of
the original target group only after an accurate process of translation and
cultural adaptation (Guillemin etal., 1993; Hocker, 2010). Furthermore,
online surveys that use a recorded sign language translation of a written
questionnaire are more suited to reach deaf persons (Graybill etal., 2010;
Hocker, 2010). However, as far as we know, only one adapted and trans-
lated HRQoL questionnaire for deaf people exists, and this instrument is
in use in Austria (Fellinger etal., 2005a). A limited guideline is available
for translating questionnaires into sign language (Crowe Mason, 2005).
Scienti c-Based Translation of Standardized Questionnaires :
However, we could nd no comprehensive guidelines that cover the whole
process of translating and adapting the questionnaires for use by deaf
and hard of hearing respondents. When we planned to conduct an epide-
miological study on the health and healthcare needs of deaf and hard of
hearing individuals in the Netherlands, we became aware of the enormous
problems posed by the lack of suitable instruments for this group. To help
solve these dif culties we developed guidelines that we adapted during
the process of translating four questionnaires into NGT for our research
project.
METHODOLOGY
Procedures
Founded on current guidelines for translating and adapting HRQoL
questionnaires for spoken languages (Hocker, 2010; Pollard, 1994;
Ravens-Sieberer et al., 2005; KIDSCREEN Group Europe, 2006) and
referring to our own trials and experiences, we have developed guidelines
for translating international written questionnaires into a sign language
(Figure1). A group of Deaf NGT communication experts, a physician/
NGT linguist, a second NGT linguist, NGT interpreters, and a master’s
student in Deaf studies translated the selected questionnaires into two
different variants of NGT. After backward translations (i.e., from NGT
into written Dutch), consultation between experts, and reviews by deaf
and hard of hearing test participants, the signed questionnaires were
adapted to the cultural and linguistic needs of deaf and hard of hearing
individuals.
Besides the two NGT versions, a sign-supported version (spoken
Dutch with simultaneously produced NGT signs) and a written Dutch
version were also provided. All of the questionnaires were placed in a
secure online environment. We used Unipark software, which allowed
us to create our own layout and has a direct link to a database (Hocker,
2010; www.unipark.de). The guidelines are presented in appendix 1,
while their development is discussed in this chapter.
Instruments
One of our rst challenges was to select test instruments that were
suited for translation into NGT and that were reliable also when used by
: S .
people who have weak language skills and minority cultural backgrounds.
The selected questionnaires for this study are as follows:
– World Health Organization Quality of Life-BREF (WHOQoL-
BREF): This is an internationally standardized questionnaire that
has been translated and validated in more than a dozen languages
and has been widely eld-tested, which makes the instrument
methodologically strong (WHOQoL Group, 1998). Numerous
questionnaires have been designed to measure all sorts of aspects of
quality of life. Quality of life questionnaires that explicitly evaluate
(1) If necessary rst perform an ofcial Forward-Backward Translation
from the questionnaire in the original written language (source language),
e.g. written English into the written language of region of the target language
(intermediate language), e.g. written Dutch.
Original questionnaire or questionnaire in intermediate language (written)
(2) TranslationF-1 Forward translation TranslationF-2 (sign language)
(3a) Reconciliation of Problematic Items (sign language)
(3b) Reconciled Forward Translation (sign language)
(4) Backward translation TranslationB-2 (written)
(5a) Review of Forward (sign language) and Backward (intermediate language) translation
(5b) Final Forward Translations after Review (sign language)
(6) Pre-test and review (sign language)
Final Questionnaires (sign language)
TranslationB-1
. Guidelines for translating questionnaires into a sign language.
Scienti c-Based Translation of Standardized Questionnaires :
participants’ physical health are less common. We chose the
WHOQoL-BREF because it was the only short questionnaire that
met both our content demands and our methodological demands.
– General Health Questionnaire (GHQ): This is a screening instru-
ment to identify minor psychiatric disorders. It can be used in
the general population or with clients in nonpsychiatric clinical
or primary care settings. We have used the GHQ-12, the short-
est version, especially designed for research studies (Goldberg &
Williams, 1988). One of its strong points is that it is a short, reli-
able questionnaire.
– KIDSCREEN: This is a generic quality of life instrument that has
been designed and normed for (hearing) children and adolescents
between the ages of 8 and 18 years. KIDSCREEN can be used
as a screening, monitoring, and evaluation tool in health surveys
(Ravens-Sieberer etal., 2005; KIDSCREEN Group Europe, 2006).
It covers ten health-related quality of life dimensions, whereas
many QoL questionnaires for children cover only psychologi-
cal and schooling domains. It is an internationally standardized
questionnaire that does not require a high language level.
– Deaf Acculturation Scale (DAS): This is a 58-item scale that
measures deaf and hard of hearing individuals’ degree of accultur-
ation to both Deaf and hearing cultures. It consists of two overall
acculturation scales: a Deaf acculturation scale (DASd) and a
hearing acculturation scale (DASh). Both measure acculturation
in ve domains (Maxwell-McCaw & Zea, 2011). The DAS is the
only validated international scale that provides information about
the cultural status of deaf and hard of hearing individuals. Having
a cultural identity different from that of the majority in society
may have a negative effect on an individual’s communication
(VanWieringen etal., 2002). The DAS can be used to evaluate
such effects among deaf and hard of hearing individuals.
Issues Encountered during Selection and Translation
of Questionnaires
Our project consisted of ve phases: 1. choosing the questionnaires;
2.producing the forward translation; 3. producing the backward transla-
tion and harmonization; 4. testing; 5. taking the survey. We encountered
: S .
linguistic, cultural, and technical issues during all phases of selection and
translation of questionnaires. Later we describe these issues and how we
dealt with them.
Selection Criteria for Questionnaires
It is dif cult to perform a validation study on questionnaires to be
used by sign language users because of the small number of such persons
in the population, as mentioned earlier. Therefore, one has to be very sure
of the potential and the characteristics of the questionnaire. We preferred
to use only internationally validated questionnaires that had already been
successfully translated into a number of languages and have been used in
various cultural settings, not questionnaires that were validated in only
a limited number of European countries or only the United States. From
those we chose the ones that showed the most potential for cultural and
linguistic translation. For instance, questionnaires that use a great deal
of gurative speech are more dif cult to translate into another language,
as are questionnaires with many semantic weaknesses (see example 1).
We also took the required language level into account. Since the medical
knowledge of deaf and hard of hearing individuals is often limited, we
excluded questionnaires that use a large proportion of medical jargon or
require a high language level.
EXAMPLE 1. (from GHQ)
Have you felt capable of making decisions about “things”?
“Things” is semantically weak and therefore a difficult concept to capture
in translation.
Selecting Signers
The main but elementary difference between a written questionnaire
and a signed one is the need for a signing interviewer on the screen. The
person who is recorded signing the questions will automatically function
as the interviewer. As with any interviewer, this individual may cause
some bias. To minimize the bias, the signers must be carefully selected.
A central point in this is that the respondents to the questionnaire
must feel comfortable with the signer, who may be asking very personal
questions. Although the respondents will be aware that the signer on the
Scienti c-Based Translation of Standardized Questionnaires :
lm will not see the answers, the signer will unconsciously exert in uence
nonetheless. Interviewer bias can also be caused by gender, age, status,
cultural and ethnic background, language, and/or linguistic style.
To minimize potential bias, we decided to make different versions. One
version was signed by a deaf woman who is a teacher of NGT and Deaf
culture and is well known in the Deaf community. The second version
was signed by a hearing male NGT interpreter who has much experience
in working in psychiatric settings but has no personal attachments to the
Deaf community. In the Netherlands the Deaf community is rather small,
and most of its members know each other. By choosing one interviewer
who was well known in this community we created an opportunity for
the participants to be questioned by someone familiar and trusted. By
also selecting an NGT interpreter without personal attachments to the
Deaf community and who works only in a small, specialized setting,
we provided an opportunity for the respondents to be questioned by an
unfamiliar, more neutral person. In addition, NGT has ve regional vari-
ants, all of which are mutually understood by native signers. For edu-
cational reasons, a standard version of NGT’s most basic lexicon was
introduced in the 1990s and is rmly established today. The two signers
of the questionnaire came from different parts of the country. They were
both instructed to sign in as standard a manner as possible, but they did
so with a slightly different NGT accent.
Two-thirds of the participants nally chose the Deaf woman as the
sign model. This model also resulted in fewer respondent dropouts than
with the male interpreter as the sign model. This supports our hypothesis
that the current practice of using ad hoc, noncultural Deaf sign language
interpreters to administer nonculturally adapted tests may cause bias and
a false sense of feeling that the participants’ needs are suf ciently met.
Adaptations in Translating the Questionnaires
While translating the questionnaire, one must understand both the
underlying reason for the questions and the frame of reference of the
target group related to them. One of the issues that we encountered
during translation was that the items on some questionnaires are written
in the rst person, for example, “I feel sad.” This is done to encourage
the participant to internalize the item. Since sign language is a visual
language, an interviewer will always be present. If a question is written in
the rst person, the interviewer will always have to use a form of direct
: S .
speech (i.e., the interviewer will point to himself while asking the ques-
tion). It is arguable whether with the original purpose of rst-person use,
more internalization of the question is reached in this situation. We found
that second person (e.g., signer points at respondent [= viewer] while
asking the question [e.g., “Do you feel sad?”]) is a more direct and more
suitable form for sign languages. Therefore we adapted the rst-person
phrases, for example, in some of the questions in the American DAS ques-
tionnaire (see example 2).
EXAMPLE 2. (from DAS)
Original question Adapted question
I call myself Deaf (yes/no) -> Do you call yourself Deaf? (yes/no)
One of the questions that was adapted for person.
An example of cultural issues during translation was the question
“Are you a member of a club or society?” (yes/no). The purpose of this
question is to test social involvement, but it was placed between mainly
medically oriented questions. Within a hearing population this will not
cause any problems, but within the Deaf community, lobby groups and
associations of Deaf or hard of hearing persons are often also seen as
clubs or societies. We transferred the question to the section where other
social questions were asked and added the word “socially” to avoid “yes”
answers when the respondent was actually not socially active (“Are you
a member of a socially active club or society?”).
Technical Issues of Translation
Our questionnaires consisted of 151 questions. This meant that, includ-
ing formal introduction and instructions, we had to translate 170 items.
We estimated that it would take approximately 8hours to produce the
rst forward sign language translation (translationF) and 4hours to lm
the adaptations. We expected the backward translation (translationB)
to take 2 hours. For the production of the sign-supported version we
reserved 4hours. Our estimations for the sign language version turned
out to be very accurate.
Although no full translation had to be made, it turned out that it is
as time consuming to lm a sign-supported version as it is to produce
a sign language version. In our case we also used a signer who was not
Scienti c-Based Translation of Standardized Questionnaires :
accustomed to being lmed, which possibly caused some extra delay. It
took a total of 12hours to lm the rst sign-supported version.
For all of the recordings on different lming days, the signer has to wear
the same clothes, which should be neutral in color but contrasting with the
background. When the camera is positioned, one has to make sure that the
whole signing area (picture) is captured. The signer’s hands should not go
outside the lmed area, not even during breaks between contiguous sen-
tences. Moreover, the signer must always look directly into the camera.
Often a helper will be standing next to the camera or a text will be put up
next to the camera; alternatively, an autocue can be used. When the signer
looks at this person or the text during the lming, the signer’s eye direction
changes, which may have grammatical consequences in most sign languages.
Since “person” in NGT, like in most sign languages, is expressed by pointing
at a certain locus (localization1), the question will seem to be directed to a
third person rather than to the viewer/participant (second person).
Presentation
Since sign language is a visual language and deaf and hard of hearing
individuals are visually oriented, it is imperative to pay special attention
to the layout of the recordings and how the lms are presented online
(Baker, van den Bogaerde, & Woll, 2008). The combination of yellow and
blue is known to provide the best contrast and be the most comfortable to
read; therefore we chose a light yellow background and a dark blue font.
We placed the lm clip with the NGT question in the center of the screen
because this is the most important item. To create a layout that would
be familiar we placed the written text below the clips on the screen, as
in subtitles. A bar at the top of the screen shows the participants what
percentage of the questionnaire they have lled out (Figure2).
When answers were only short phrases (e.g., yes/no/don’t know), these
could be provided only in writing because, when several clips are placed
on one screen, the overall view is reduced, and a great deal of viewing
time is required. Although using NGT instead of written text improves
Deaf respondents’ comprehension, it is also more time consuming to look
at movies than to read. When the answers are only short phrases, these
side effects compromise the positive effect of NGT on comprehensibility.
Other researchers have tried to compensate for the dif culty of display-
ing sign language answers by making the answer options more visual in
written text (Graybill etal., 2010; Hocker, 2010; Munro & Rodwell, 2009)
: S .
. Examples of layout.
(e.g., color-coding [green for a positive answer, red for a negative one],
the use of illustrations such as thumbs up, at hand, thumbs down for
good, moderate, poor, respectively], or smiley faces). However, because
colors or illustrations might impart a positive or negative connotation to
the answers and thereby possibly add a bias, we decided not to use these.
Scienti c-Based Translation of Standardized Questionnaires :
Some computer-based tests require an answer to every question before
the respondent can proceed to the next question. We used this restriction
only for the most crucial questions (e.g., “date of birth”). We gave the par-
ticipants the option of skipping questions because the written questionnaires
also have the option of leaving a blank in case the test participant does not
understand the question or does not know the answer or does not want to
answer the question. Filling out an answer just to be able to continue the
test would compromise the reliability. Although it was possible to do so,
none of our respondents left blanks while lling out the questionnaires.
Testing the Questionnaires
Pretesting can be done by many different methods. For the question-
naires we used a cognitive interviewing method based on paraphrasing as
a variant of the think-aloud method (KIDSCREEN Group Europe, 2006)
(asking the respondents to repeat the item in their own words immedi-
ately after answering the item). This technique permits the researcher to
nd out whether a respondent understands the question and interprets it
in the manner intended. It may also reveal weak wordings of items.
Additionally, using the general probing method (KIDSCREEN Group
Europe, 2006), the participants were asked whether the items were com-
prehensible and clear and whether they were easy or dif cult to answer.
During the translation process the review committee (a committee designed
to assess the entire forward-backward process in order to provide a nal
forward translation) noticed a difference in focus of the two forward trans-
lation teams. The forward translation team that was producing the Deaf
woman sign model was focusing more on comprehensibility and cultural
adaptation, whereas the team that was producing the male interpreter model
focused more on making the translation as literal as possible. After consulta-
tion, the former team was instructed to translate more strictly, and the latter
team was instructed to focus more on a conceptual than a literal translation.
During the testing phase differences also surfaced. Testers who were deaf-
ened early in life preferred the Deaf woman sign model because they felt
that the language used in that version was slightly more accessible. Testers
deafened at a later age tended to favor the more literal male interpreter
model. Six persons were asked to test all 170 items. They considered only
one item to be dif cult to understand because there is no proper translation
for the concept “leisure activities” in NGT (see example 3). This is because
the concept of leisure time seems to be unfamiliar in the Deaf community.
: S .
EXAMPLE 3. (from WHOQoL-BREF)
“To what extent do you have an opportunity for leisure activities?”
Conducting the Questionnaire
In contrast to written questionnaires, questionnaires in sign language
cannot be lled out using paper and pencil. Some sort of visual technical
support is needed. We placed our questionnaires in an online environ-
ment. Participants lled out our questionnaire at home on their own com-
puter. Several meetings at Deaf clubs and a center for elderly deaf persons
were organized. People who did not possess enough computer skills to
ll out the questionnaire at home could receive help at these meetings.
Assistance was given by three members of our team who were trained to
provide only technical assistance; none with regard to content was given.
During the rst phase of the study the questionnaire was made avail-
able at a secure Internet site. After signing a written consent form partici-
pants received a personal login for the questionnaire. During the second
phase of the study this was altered because the procedure seemed to hinder
both Deaf and hearing people in their study participation. The Dutch
Deaf community is a small, close-knit community. Some of its members
reported to us that they had doubts about the anonymity because they
had to write their name on the informed consent, while some Deaf com-
munity members were team participants. In addition, some of the pos-
sible candidates for our hearing control group reported that the written
informed consent procedure was too time consuming.
During the second phase of our study we tried to overcome these barriers
by placing the questionnaire in a secure environment without login authori-
zation, enabling people to give online consent instead of written consent.
Participant Recruitment
We generated much publicity about the project with articles and
announcements in patient group newsletters, magazines, national and local
newspapers, and websites of Deaf clubs and/or organizations for people with
hearing impairment. General information about the study was provided
at gatherings of the Deaf community, symposia for people with hearing
impairment, and medical symposia. In addition, participants were recruited
through snowball sampling and newsletters of hearing aid manufacturers.
Scienti c-Based Translation of Standardized Questionnaires :
Informed consent was obtained by providing information brochures
and consent forms both in written Dutch and NGT. All of the informa-
tion was customized for people with weak language skills.
A total of 274 deaf and hard of hearing individuals lled out our
questionnaires. The audiological characteristics based on self-reports are
shown in Figure3. Twenty-eight percent of our participants described
themselves as members of the Deaf community; 19.7% had at least one
deaf parent; and 15% had a cochlear implant. Of the 76 participants who
were deafened at an early age and described themselves members of the
Deaf community, 37% lled out one of the sign language versions of the
questionnaire; 4% lled out the sign-supported version; and 59% chose
the written Dutch questionnaire.
GUIDELINES
The most important differences between guidelines for spoken lan-
guage translations and translating for deaf and hard of hearing individu-
als are the following:
• Current guidelines always advise translating the original question-
naire using one native speaker of both languages. However, often
no native speakers of both the target sign language and the source
language are available. Therefore a full forward and backward
FPO
Berekend gehoor
Deafened at brith
Sudden deafness/ lat
e
deafened
SHoH at brith
SHoH at later age
MHoH at later age
MHoH at brith
Born HoH, deafened
after age 5
. Hearing status of our participants based on self-report (SHOH=
severe hard of hearing, MHOH= mild hard of hearing).
: S .
translation to the written form of the local spoken language
(intermediate) has to be made rst, in accordance with interna-
tional guidelines, before starting the forward translation into the
sign language of choice.
• We advise setting up a multidisciplinary team instead of making
the translation with one person alone. These teams should include
at least the following:
a) a professionally trained sign language interpreter
b) a native user of the target sign language, preferably a Deaf
communication specialist or Deaf sign language teacher
If possible, the team should also include the following:
c) a linguistic specialist in sign language (sign linguist)
d) a psychologist/psychotherapist with experience in the
development of psychological tests
All of the team members must be familiar with the target sign language
and the cultures of both the region of the source (written) language and
the target (sign) language and have experience with psychological testing.
• As prescribed in international guidelines, at least two separate
translations should be made, differences discussed, and adaptations
made. International guidelines advise proceeding with one version
after performing the rst forward and backward translation. The
experiences of both translators and translation groups are used to
improve what is considered to be the stronger translation. Ideally,
when adapting to and translating into a sign language, one should
start with four translations (two signed by a sign language inter-
preter, and two by native Deaf signers), choosing the best version of
both groups to continue with. Depending on possible local cultural
issues, more versions might be necessary. Since resources for carry-
ing out this kind of project are often limited, it will frequently be
unfeasible to start with four or more translations. In this context
we advise performing two translations using the experiences of
the two translation groups to improve both versions. Continuing
with at least two versions also serves to minimize the interviewer
bias. At least one of the versions should be signed by a Deaf native
signer of the target sign language. Other personal characteristics,
depending on local culture, should be taken into account.
Scienti c-Based Translation of Standardized Questionnaires :
CONCLUSIONS
Specialized mental health care for deaf and hard of hearing clients has
emerged during the last 50 years in many Western countries. The deaf
and hard of hearing client group poses additional linguistic and social
challenges to (mental) health care providers. This makes the process of
diagnosing and treating deaf and hard of hearing clients more challeng-
ing than that for hearing clients. Various diagnostic tests have been used
for these clients over the years. However, these tests were not speci cally
developed for this group or were translated only ad hoc. The reliability
of such testing is questionable.
We advise the use of internationally validated written questionnaires
in this population. However, questionnaires have to be carefully selected,
translated, and modi ed both linguistically and culturally. If this is not
properly done, bias can occur, potentially resulting in the misdiagnoses
of many clients. With this chapter we hope to raise the awareness of this
group’s special needs, and we propose a set of guidelines (see Figure1) for
practitioners and researchers who would like to use standardized tests for
deaf and hard of hearing sign language users.
NOTES
. Localization refers to pointing to a speci c place in space, whereby the point-
ing acquires semantic meaning (e.g., “ rst person” is pointing at yourself,
while “second person” is pointing at the interlocutor or at the space directly
in front of the signer).
REFERENCES
Baker, A., van den Bogaerde, B., & Woll, B. (2008). Methods and procedures in
sign language acquisition studies. In A. Baker & B. Woll (Eds.), Sign Language
Acquisition (Special issue). Sign Language and Linguistics, 1–49. Amsterdam:
Benjamins.
Barnett, S. (1999). Clinical and cultural issues in caring for deaf people. Family
Medicine 31, 17–22.
Crowe Mason, T. (2005). Cross-cultural instrument translation: Assessment,
translation, and statistical applications. American annuals of the deaf, 150(1),
67–72.
: S .
Fellinger, J., Holzinger, D., Dobner, U., Gerich, J., Lehner, R., Lenz, G., & Gold-
berg, D. (2005a). An innovative and reliable way of measuring health-related
quality of life and mental distress in the Deaf community. Social Psychiatry
and Psychiatric Epidemiology, 40, 245–250.
Fellinger, J., Holzinger, D., Dobner, U., Gerich, J., Lehner, L., Lenz, G., & Gold-
berg, D. (2005b). Mental distress and quality of life in a deaf population.
Social Psychiatry and Psychiatric Epidemiology, 40, 737–742.
Fellinger, J., Holzinger, D., Dobner, U., Gerich, J., Lehner, L., Lenz, G., & Gold-
berg, D. (2007). Mental distress and quality of life in the hard of hearing. Acta
Psychiatrica Scandinavica, 115, 243–245.
Goldberg, D. P., & Williams, P. A. (1988). User’s Guide to the GHQ. Windsor:
NFER Nelson.
Graybill, P., Aggas, J., Dean, R. K., Demers, S., Finigan, E. G., & Pollard, R. Q, Jr.
(2010). A community-participatory approach to adapting survey items for deaf
individuals and American Sign Language. Fields Methods, 22(4), 429–448.
Guillemin, F., Bombardier, C., & Beaton, D. (1993). Cross-cultural adaptation of
health-related quality of life measures: Literature review and proposed guide-
lines. Journal of Clinical Epidemiology, 46(12), 1417–1432.
Hocker, T. J. (2010). Sozialmedizinische Aspekte der medizinischen Versorgung
gehörloser Menschen in Deutschland Entwicklung und Durchführung einer
internetbasierten Umfrage mit Gebärdensprachvideos. Inauguraldissertation
zur Erlangung des Doktorgrades der Medizin der Universitätsmedizin der
Johannes Gutenberg-Universität Mainz.
Hoem Kvam, M. (2004, March). Sexual abuse of deaf children: A retrospective
analysis of the prevalence and characteristics of childhood sexual abuse among
deaf adults in Norway. Child Abuse and Neglect 28(3), 241–251. http://www.
unipark.de
Jones, E. G., Renger, R., & Firestone, R. (2005). Deaf community analysis for
health education priorities. Public Health Nursing 22, 27–35.
Kennedy, C., & McCann, D. (2004). Universal neonatal hearing screening moving
from evidence to practice. Archives of Disease in Childhood: Fetal and Neona-
tal Edition, 89, 378–383.
KIDSCREEN Group Europe. (2006). The KIDSCREEN Questionnaires: Qual-
ity of life questionnaires for children and adolescents. Handbook. Lengerich:
Pabst Science.
Ladd, P. (2003). Understanding Deaf culture: In search of Deafhood. Clevedon:
Multilingual Matters.
Maxwell-McCaw, D., & Zea, M. C. (2011, January). The Deaf Acculturation
Scale (DAS): Development and validation of a 58-item measure. Journal of
Deaf Studies and Deaf Education. Advance Access published. DOI:10.1093/
deafed/enq061
Scienti c-Based Translation of Standardized Questionnaires :
Munro, L., & Rodwell, J. (2009). Validation of an Australian Sign Language
instrument of outcome measurement for adults in mental health settings.
Australian and New Zealand Journal of Psychiatry, 43, 332–339.
Musselman, C. (2005). How do children who can’t hear learn to read an alpha-
betic script? A review of the literature on reading and deafness. Journal of
Deaf Studies and Deaf Education, 5(1), 9–31.
Padden, C., & Humphries, T. (1988). Deaf in America: Voices from a culture.
Cambridge, MA: Harvard University Press.
Padden, C., & Humphries, T. (2005). Inside Deaf culture. Cambridge, MA: Har-
vard University Press.
Pollard, R. (1994). Public mental health services and diagnostic trends regarding
individuals who are deaf or hard of hearing. Rehabilitation Psychology, 39(3),
147–160.
Ravens-Sieberer, U., Gosch, A., Rajmil, L., Erhart, M., Bruil, J., Duer, W., Auquier,
P., Power, M., Abel, T., Czemy, L., Mazur, J., Czimbalmos, A., Tountas, Y.,
Hagquist, C., Kilroe, J., & European KIDSCREEN Group. (2005). KID-
SCREEN-52 quality-of-life measure for children and adolescents. Expert
Review of Pharmacoeconomics and Outcomes Research, 5(3), 353–364.
Smeijers, A. S., & Pfau, R. (2009). Towards a treatment for treatment: The com-
munication between general practitioners and their deaf patients. Translator
and Interpreter, 3(1), 1–14.
Stein, L. K., Mindel, E. D., & Jabaley, T. (1981). Deafness and mental health. New
York: Grune and Stratton.
Van Wieringen, J. C., Harmsen, J. A., & Bruijnzeels, M. A. (2002, March). Intra-
cultural communication in general practice. European Journal of Public
Health, 12(1), 63–68.
Vernon, M., & Andrews, J. F. (1990). The psychology of deafness: Understanding
deaf and hard-of-hearing people. New York: Longman.
WHOQoL Group. (1998). Development of the World Health Organization
WHOQoL-BREF quality of life assessment. Psychological Medicine 28,
551–558.
: S .
Appendix 1.
Translation Methodology for Translating
Questionnaires into Sign Language
1. INTRODUCTION
This guideline is based on the WHOQoL (WHOQoL Group, 1998)
and KIDSCREEN (Ravens-Sieberer et al., 2005; KIDSCREEN Group
Europe, 2006) guidelines for translating questionnaires between written
languages. Our adaptations conform to these international guidelines for
translating into sign languages.
2.1. Overview
The translation process should focus on achieving conceptual equiva-
lence of all versions rather than linguistic/literal equivalence. With regard
to the translation methodology the forward-backward-forward tech-
nique should be applied.
In Step 1, the translation of the original questionnaire into the local
written language (called the “intermediate language”) should conform to
the international forward-backward translation guidelines.
Step 2: Two independently working translation teams should translate
the questionnaire into the local target sign language (TranslationF-1 and
TranslationF-2) (cf. Figure1, step 2).
In Step 3 (the following reconciliation), the two forward translation
teams and one member of the research group review the two forward
translations into the target sign language in order to create the reconciled
forward translation. They identify problematic items, discuss the origi-
nal formulation and intention of these items in the source language, and
consider possible ways of translating them into the target language while
meeting all of the demands of conceptual equivalence with the original
questionnaire in the source language (cf. Figure1, steps 3a and 3b).
In step 4, the reconciled forward translation should be translated back
into the intermediate language (translationB-1 is the backward translation
Scienti c-Based Translation of Standardized Questionnaires :
of translationF-1, and translationB-2 is the backward translation of trans-
lationF-2). This should be done by independent professional sign language
translators meeting the criteria as described later (cf. Figure1, step 4).
In step 5 the members of the research group (and, if available, external
experts with experience in instrument development and translation) and
the forward translation team compare TranslationB-1 and TranslationB-2
with the original, source language version, thus reviewing the reconciled
forward translation. If necessary adaptations are made, this version is con-
sidered the nal forward translation (cf. Figure1, steps 5a and 5b).
In step 6, subsequent to the generation of the nal forward translation
all reviews and translation data are sent to the developers of the original
questionnaire for documentation and consultation. The objective of the
consultation is to resolve both inadequate concepts of translation and all
discrepancies between alternative versions (cf. Figure1, step 6).
2.2. Pretranslation Phase
There are often no native speakers of both the target sign language
and the original language available. Therefore a full forward-backward
translation, conforming to international guidelines, must be made before
starting the forward translation into a sign language.
2.3. Forward Translation
All of the translators working on the project should not only be pro-
fessional and experienced translators but preferably also have experience
with psychological and/or psychiatric testing or even test development.
We recommend that the forward translation team and the review
committee be formed as a multidisciplinary team consisting at least of a
professionally trained sign language interpreter (often a nonnative sign-
er) and a native user of the target sign language. The team should also
include a deaf communication specialist or a deaf sign language teacher, a
linguistic specialist in sign linguistics, and a psychologist/psychotherapist
with experience in the development of psychological tests.
All of the team members need to be:
• Signers of the target sign language
• Knowledgeable about the source language, the intermediate lan-
guage, and the target sign language
: S .
• Familiar with the cultures of the region of the source language and
of the target language/community (e.g., local Deaf culture)
• Experienced in test development and/or psychological testing of
deaf and hard of hearing clients
Two translation teams should translate the questionnaire indepen-
dently from each other into the target sign language. The translators are
asked to use natural and acceptable language for the broadest audience
and to be simple, clear, and concise in their formulations.
The following guidelines may be given to the translators:
• The translators should always focus on conceptual equivalence
rather than on literal word-for-word translation. They should
always try to grasp the most relevant meanings of the original
terms and translate them accordingly.
• The translators should try to be simple, clear, and concise in their
formulations; long sentences with many clauses should be avoided.
• The translators should take into account what typical respondents
will understand when they see the items.
• The translators should take the age of the respondents into con-
sideration and thus not use any jargon or terms that would be
dif cult to understand. The translation has to be clear, simple,
and comprehensible. Double negatives should be avoided.
The two forward translators provide two forward translations: Trans-
lationF-1 and TranslationF-2. The two versions are then reconciled in the
next step.
2.4. Reconciliation of Items
Participants in the reconciliation procedure review the two forward
translations. These participants should include the following:
• The two forward translators
• One member of the research group with good knowledge both of
the source language and of the target language
To reconcile the two independent forward versions a reconciliation
meeting should be held so that the two translations can be compared and
assessed with regard to their conceptual equivalence, comprehensibility,
and clarity of signs relative to the original questionnaire.
Scienti c-Based Translation of Standardized Questionnaires :
Participants in the reconciliation procedure should document their as-
sessments item by item and, if neither is adequate, suggest another transla-
tion. They should focus on differences in culture and linguistics that may
cause dif culties when transforming the source version into the target
languages. The reconciled versions are to be derived by means of a sub-
sequent discussion between the participants. In contrast to most interna-
tional guidelines we advise that, after doing this, the participants proceed
with both versions to minimize the interrogator bias. At least one of the
versions has to be signed by a Deaf native user of the target sign language.
The reconciliation procedure may also produce valuable clues to dif-
ferences in culture and/or linguistics that are relevant to the whole trans-
lation process.
2.5. Backward Translation
Designed to assess the conceptual equivalence of the reconciled for-
ward translation and the source questionnaire, the backward translation
serves as an instrument to measure the quality of the reconciled forward
translation. The backward translation and the intermediate language ver-
sion are supposed to be very similar, and if they are not, discrepancies
such as problematically translated items will thus become manifest and
can be corrected.
The backward-translator must be a professional sign language inter-
preter (if possible, a native signer) and should have familiarity with the
cultures, both of the original language/ country as well as of the target
signers (ecg local Deafculture), have experience in test development and/
or psychological testing of hearing impaired participants.
The reconciled forward translation is translated into the intermediate
language by the backward translator, who should not have worked with
the questionnaire before. The guidelines as described earlier (guidelines
for forward translators) may also be given to the backward translator.
The result of the backward-translation process is a translated version of
the reconciled forward translation in the intermediate language.
2.6. Review of the Forward and Backward Translation
The review is designed to assess the entire forward-backward process
in order to provide a nal forward translation. Participants in the review
procedure should include the following:
: S .
• Two members of the research group with good knowledge of both
the source and the target sign language
• One of the forward translators
• If available, external experts with experience in instrument devel-
opment and translation
Focusing on conceptual differences, the backward translation (in the
intermediate language) is to be compared with the original, source lan-
guage questionnaire. The participants review the translation process item
by item by comparing the TranslationB items to the original source lan-
guage items and suggesting a version for the nal forward translation in
the target sign language. This is done either by con rming the results of
the reconciliation process or by suggesting an alternative translation if
necessary. All changes in wording or meaning of the items are to be un-
dertaken while generating the nal forward translation.
In this nal process, the review board is expected to ensure that the
translation is simple, clear, and concise and, most important, that there
are no conceptual discrepancies between the original (source language)
and the nal forward translation (target sign language). The focus should
be on achieving conceptual equivalence and clarity as well as on using
colloquial language.
2.7. Assessment of Conceptual Equivalence and First
Harmonization of Problematic Items with the Developers
of the Original Questionnaire
International harmonization is intended to ensure the comparabil-
ity of the translated questionnaires. The reviews and translation data of
problematic items will be sent to the developers of the original question-
naire. The objective of the consultation is to resolve inadequate concepts
of translation as well as all discrepancies between alternative versions.
This is done to ensure and, if necessary, generate interconceptual
equivalence.
2.8. Pretest (Cognitive Interviews)
The pretest is expected to show whether all of the items are compre-
hensible and acceptable.
Scienti c-Based Translation of Standardized Questionnaires :
Test participants should be provided with a quiet place for testing.
With regard to their contribution to the test development procedure, test-
ed individuals are to be informed of the objective of the pretest.
Pretesting is critical for identifying questionnaire problems such as
misunderstandings about the intended meaning of items. Problems with
item content, including confusion about the overall meaning of items,
as well as misinterpretation of individual terms or concepts, can also be
identi ed. Pretesting incorporates many different methods. We propose
a cognitive interviewing method based on paraphrasing as a variant of
the “think-aloud method” (asking respondents to repeat the item in their
own words immediately after answering it).
This technique permits the researcher to determine whether the
respondent understands the question and interprets it in the manner
intended. It may also reveal better wordings for items.
Additionally, when using the “general probing method,” respondents
will be asked whether the items can be considered comprehensible and
clear and whether they were dif cult to answer.
2.9. International Harmonization
All translation reviews and translation data will be sent to the devel-
opers of the original questionnaire for documentation and consultation.
The objective of the consultation is to resolve inadequate concepts of
translation and all discrepancies between alternative versions.
The international harmonization procedure should be carried out by
one or two members of the research group, if possible those who have
already been involved in the review procedures. A telephone or Skype
conference with one member of the questionnaire group should serve as
a platform to discuss all questions about conceptual and cultural aspects
of the item translations.
The nal questionnaire versions are generated in the process of inter-
national harmonization and should preferably be subsequently tested in
a validation study.
