INT J LANG COMMUN DISORD, NOVEMBERDECEMBER 2011,

VOL. 46, NO. 6, 613627

Review
Evaluation of speech and language assessment approaches
with bilingual children
Caroline De Lamo White and Lixian Jin
Speech and Language Therapy, Leicester Partnership Trust, Leicester, UK
Speech and Language Therapy, De Montfort University, Leicester, UK

(Received 8 June 2010; accepted 4 December 2010)

Abstract
Background: British society is multicultural and multilingual, thus for many children English is not their main or
only language. Speech and language therapists are required to assess accurately the speech and language skills of
bilingual children if they are suspected of having a disorder. Cultural and linguistic diversity means that a more
complex assessment procedure is needed and research suggests that bilingual children are at risk of misdiagnosis.
Clinicians have identified a lack of suitable assessment instruments for use with this client group.
Aims: This paper highlights the challenges of assessing bilingual children and reviews available speech and language
assessment procedures and approaches for use with this client group. It evaluates different approaches for assessing
bilingual children to identify approaches that may be more appropriate for carrying out assessments effectively.
Methods & Procedures: This review discusses and evaluates the efficacy of norm-referenced standardized measures,
criterion-referenced measures, language-processing measures, dynamic assessment and a sociocultural approach.
Outcomes & Results: When all named procedures and approaches are compared, the sociocultural approach appears
to hold the most promise for accurate assessment of bilingual children. Research suggests that language-processing
measures are not effective indicators for identifying speech and language disorders in bilingual children, but further
research is warranted. The sociocultural approach encompasses some of the other approaches discussed, including
norm-referenced measures, criterion-referenced measures and dynamic assessment.
Conclusions & Implications: The sociocultural approach enables the clinician to interpret results in the light of the
childs linguistic and cultural background. In addition, combining approaches mitigates the weaknesses inherent
in each approach.

Keywords: speech and language assessment approaches, bilingual children, sociocultural approach

What this paper adds

What is already known on this subject
Many procedures and approaches for assessing bilingual clients have been trialled within different clinical studies,
however clinicians may not have the time or the opportunity to become familiar with all procedures and approaches
available to themlet alone make a decision about which approaches should be employed with their bilingual
clients.

What this study adds

This study reviews major approaches used for assessing the speech and language needs of bilingual children with
the aim of informing clinical practice. It evaluates these procedures and approaches with examples and evidence
to present arguments for using a sociocultural approach as a more appropriate method, whilst incorporating the
strengths of other approaches. This study makes a case for the provision of training to equip students and speech
and language therapists with the knowledge and skills to use these approaches with bilingual children.

Address correspondence to: Dr. Lixian Jin, Speech & Language Therapy, De Montfort University, H0.19b Hawthorn Building, Leicester, LE1
9BH, UK; e-mail: jin@dmu.ac.uk
International Journal of Language & Communication Disorders
ISSN 1368-2822 print/ISSN 1460-6984 online  c 2011 Royal College of Speech and Language Therapists
DOI: 10.1111/j.1460-6984.2011.00049.x
614 Caroline De Lamo White and Lixian Jin
Introduction associated with bilingualism; it implies participation in
two or more cultures (Martin 2009), but biculturalism
Accurate assessment of a childs communicative ability
is not necessarily a given result of being bilingual or vice
plays an essential role in diagnosis and informs
versa.
therapy so that intervention strategies meet the childs
Britain is a multicultural and multilingual society
needs. The Royal College of Speech and Language
and in principle each person is perceived to have the
Therapists (RCSLT) (1996: 268) defines bilingual-
same rights and access to services. SLTs frequently
ism as people who acquire communicative skills in
come into contact with clients who speak more
more than one language. Bilingualism does not cause
than one language and increasingly with those whose
language disorders and is seen by the RCSLT as a
primary language is not English. The UK Office of
distinct communicative advantage. The sociocultural
National Statistics (2001) states that the minority ethnic
and linguistic differences associated with bilingual-
population accounts for 7.9% of the total British
ism result in a more complex assessment procedure.
population (in a city like Leicester it is over 45%),
This, combined with a perceived lack of appropriate
while minority ethnic groups account for 73% of
assessment tools (Roseberry-McKibbin et al. 2005),
Britains overall population growth (Kings Fund 2006).
has placed bilingual children at risk of misdiag-
This ethnic diversity implies bilingualism in practice
nosis (Crutchley et al. 1997). Bilingual children
though not necessarily for some communities. School
are consequently under and over-represented on the
pupils in London between them speak well over 300
caseloads of speech and language therapy (SLT) services
different languages, plus many more dialects (Baker and
in the UK (Winter 1999, 2001).
Eversley 2000) and about 30% of Londons children
Although this paper does not focus on bilingual-
use English as an Additional Language (EAL). In some
ism per se, it is important to position the context
schools, over 90% of pupils are considered to be within
within the key concepts of bilingualism and different
the EAL group. As this diversity looks set to increase
types of bilinguals to set up a theoretical framework
through migration and minority ethnic birth rates,
to review speech and language assessment approaches
clinicians will more frequently come into contact with
with bilingual children. Current concepts of bilingual-
bilingual children. Despite the fact that the majority of
ism recognize its diversity and complexity, thus the
children nationally are monolingual English speakers,
term bilingualism has various connotations (Martin
Winter (1999) estimates that 59% of therapists will
2009). For some people the term implies complete oral
encounter at least one bilingual child and that 11% work
fluency in two languages; for others it refers to any
with a minimum of 20 bilingual children; however, as
degree of competence in two languages. Baker (2006)
the figures cited above show, these percentages will be
has presented the complexity of defining bilingual-
far higher in some cities. Clinicians have an ethical, legal
ism by giving different dimensions of bilingualism,
and professional duty (RCSLT 1996) to meet the needs
rather than defining bilingualism in one sentence or
of these children.
even a paragraph. Li (2007: 7) presents 37 types of
Professional standards (Health Professionals Council
bilinguals, covering many social, economic and linguis-
2007, RCSLT 1996) require clinicians to offer a non-
tic dimensions and the degrees of competence of a
discriminatory service and to be competent in carrying
bilingual person.
out appropriate and sensitive assessment procedures.
However, the RCSLT (1996) defines bilingualism as
It is imperative that the profession addresses any
an ability to communicate in two or more languages,
inequality of care being provided to bilingual children.
in a variety of modalities and to varying levels of
Consequently, the purpose of this paper is to answer the
proficiency. Currently, bilingualism generally refers to
following questions.
people who use more than one language for interac-
tion on a day-to-day basis (Martin 2009). Further, the
papers pertinent to this review use a variety of terminol- What assessment techniques are available to
ogy to describe their target populations including: clinicians for use with bilingual children?
minority children, culturally and linguistically diverse Which assessment techniques are the most
(CLD) children, English language learners (ELL), second- appropriate and effective at diagnosing language
language learners and bilingual or bicultural children. impairments in bilingual children?
Aspinall (2002) acknowledges the ambiguous nature of
terminology surrounding minority ethnic populations. This review will start with the examination of
For clarity, this review defines bilingual children as those assessment challenges; it then moves to identify
who speak or are regularly exposed to two or more assessment techniques discussed in research literature. It
languages. The term minority ethnic (Multiverse 2004) will systematically evaluate the scope and efficacy of each
is used in all other instances. Biculturalism is commonly technique so that clinicians are better informed about
SLT assessment approaches with bilingual children
assessment options for use with bilingual children. It will
make recommendations for improving future practice
and identify areas for further research in the hope of
reducing occurrences of misdiagnosis and inequalities
in client care.
Assessment challenges
Bilingual children present SLTs with a unique challenge;
the clinician must differentiate between a fundamen-
tal speech and language disorder and perceived difficul-
ties resulting from speech and language differences.
Differences may be expressed in sentence structure,
speech sound production, vocabulary and pragmatics
(Roseberry-McKibbin 2002). For bilinguals who are
learning English there may be a linguistic influence
or interference on any of these language levels from a
first language onto the second (Odlin 1989) or from
a second language to the first (Cook 2003). In the
case of specific languages, such influences often have
predictable effects on a bilinguals use of English as a
second language (Swan and Smith 2001), which means
that SLTs can ascertain some influences. As Hamayan
and Damico (1991) point out, these phenomena are
superficially similar to those exhibited by children who
have a fundamental speech and language disorder. The
following short discussion highlights other features
that make assessment of bilingual children particularly
complex.
An important part of assessment is knowledge of the
familys culture and language use, including dialectal
variation (Laing and Kamhi 2003). Figure 1 is one
model of cultural identity (London SIG Bilingualism
2007a) which takes clients socio-cultural factors into
consideration while assessing them. The four boxes
loosely represent the components of cultural identity
which include: ethnicity, culture, race and nationality,
along with relevant sub-factors (London SIG Bilingual-
ism 2009). Without considering these factors, a clinician
Values & Beliefs Diet
Traditions Dress
Language Lifestyle
Religion Family Dynamics
Country of Origin Class
Physical Appearance / Socioeconomic Status
Features Political Climate
Gender Education
Age Employment
Sexuality Experience
Disability
Figure 1. Factors that constitute cultural identity (London SIG
Bilingualism 2007b).
615
runs the risk of carrying out a culturally and linguis-
tically biased assessment procedure and is unlikely to
obtain a genuine profile of the childs communicative
competence.
Further, a socio-cultural approach should be seen as
the essential approach underpinning other assessment
methods for assessing bilingual clients; rather than
simply being an alternative to other approaches, it
should be seen as the core context for them. Yet,
the socio-cultural approach may not in itself solve the
totality of problems associated with other approaches.
However, Mennen and Stansfield (2006) comment
that many SLTs lack knowledge about languages and
cultures other than their own. The majority of SLTs in
the UK are white, middle-class and women (Cortazzi
and Jin 2004). One possible solution would be to
encourage more people from minority groups and
speakers of minority languages into the profession.
However, Papoutsis Kritikos (2003) points out that SLTs
who speak minority languages will inevitably come into
contact with children who speak yet other languages and
are from unfamiliar cultures. It is unrealistic to suggest
that recruitment would be a solution in itself.
Sanchez (2006) proposes that another of the
challenges facing clinicians is the variability of language
proficiency amongst bilingual children. Language
proficiency is dependent upon a number of factors
including the childs age and intensity of exposure to
a language, as well as the opportunity and motivation
which the child has to use that language. Similar variabil-
ity applies to cultural factors (Cheng 1997). To a large
extent, the childs cultural identity will be determined
by the values and beliefs held by their carers, and as
argued by Jin and Cortazzi (1993: 84), these may be
inexplicit: Culture is necessarily subjective and involves
assumptions, ideas and beliefs which are often not
articulated. The challenge is that the clinician must
fully consider the childs language history and cultural
background in order to make an informed diagnosis
but clients and their families cannot always be expected
to make explicit the relevant linguistic and cultural
information
Chavda et al. (2003) stress that insufficient
knowledge on the part of health and educational
professionals about language acquisition in bilingual
children often results in misdiagnosis. The clinician
must be aware of developmental processes and patterns
of language use which are used by bilingual individuals.
These include language mixing, code switching, interfer-
ence, language loss, fossilization and a silent period
(Hamayan and Damico 1991, Nicoladis and Genesee
1997, Roseberry-McKibbin 2002). Well-intentioned
professionals unfamiliar with these processes may
erroneously refer bilingual children for SLT when it is
not required.
616
Given the complex interplay between the above
variables, it is unsurprising that clinicians report finding
it difficult to make sound clinical judgements about
the communicative competence of bilingual children
(Crutchley et al. 1997, Winter 2001). Many clinicians
do not believe that they have the necessary skills or
training to do so (Papoutsis Kritikos 2003).
Despite language differences, any battery of
speech and language assessments used must accurately
differentiate between children with genuine language
impairment (LI) and children with normally develop-
ing language (LN)or languages.
Caroline De Lamo White and Lixian Jin
are perceived as an efficient testing method which
provides a good indication of a childs position amongst
his/her peers in terms of a percentile rank. It is a
description which many other professionals understand.
However, Caesar and Kohler (2007) comment that
norm-referenced assessments have proven to be an
inadequate diagnostic tool for bilingual children. This
inadequacy is attributed to content bias, linguistic
bias and a disproportionately small representation of
bilingual children in normative samples (Laing and
Kamhi 2003, Stow and Dodd 2003, Martin 2009).

Assessment proceduresrelevant research
There is an emerging body of research into the
assessment of bilingual children (Rodekohr and
Haynes 2001, Irvine Saenz and Blake Huer 2003,
Stokes et al. 2006, Gould 2008). There is a
variety of assessment approaches discussed, which
mainly include five types: norm-referenced standardized
measures, criterion-referenced (CR) measures, language-
processing measures, dynamic assessment (DA) and a
sociocultural approach. Table 1 provides an overview of
each type of assessment measure.
These approaches will be discussed in the order
outlined in Table 1. The research will be presented
and evaluated for each approach before considering the
implications for practice and drawing conclusions.
Norm-referenced standardized measures
Norm-referenced, standardized speech and language
assessments have been described as the tools of the
profession (Stow and Dodd 2003: 363), and the
cornerstone of diagnosis in the field (Pena and
Spaulding 2006: 247.) This is perhaps because they
Content bias
Whilst norm-referenced testing enables clinicians to
evaluate a childs existing knowledge base and compare
it with normally developing children of the same age,
it works on the assumption that all children have been
through similar life experiences and have knowledge of
the same concepts and vocabulary. However, this can
be an erroneous assumption when it does not take into
account the differences in experience associated with
cultural diversity. For example, Sanchez (2006) found
that the use of traditional reading and writing tasks
for evaluating Peruvian childrens language proficiency
placed bilingual children from rural communities at a
distinct disadvantage, partly because of the predomi-
nance of oral communication over written forms within
their home communities. Some languages even today
do not have a written form (Stow and Dodd 2003)
and in others with different scriptssuch as Chinese
or Japanesethe nature of literacy can be different.
Significantly, the assessments in Peru were carried
out predominantly in Spanish, thus discounting the
childrens proficiency in their home language. Sanchez
(2006) found that the pictures used for assessment
needed to reflect the childrens own rural environment in

Table 1. Assessment measures summarized

Assessment approach What it measures Example

Norm-referenced standardized measures
Criterion-referenced measures
Language-processing measures
Dynamic assessment
Sociocultural approach
A measure of a childs performance determined
by direct comparison with the performance of
a specified population
A measure of a childs level of performance on a
specific skill, e.g. grammatical structure
A measure of a childs ability to process language
An evaluation of a childs language-learning
potential and the mechanisms by which they
learn
A holistic evaluation of a childs communicative
abilities within their wider environment
Published formal assessment, e.g. Pre-School
CELF (Wiig et al. 2006) in which a
standard score and percentile rank is
derived
Language sampling for linguistic analysis, e.g.
Using mean length of utterance to evaluate
the individuals functional ability
Memory tasks such as non-word repetition
(NWR) and competing language
processing tasks (CLPT)
Mediation. Testteachretest. Graduated
prompting
Ethnographic assessment such as RIOT
(review, interview, observe, test) approach
SLT assessment approaches with bilingual children
order for them to make accurate interpretation. Other
cross-cultural researchers have found that interpreting
pictures is culturally specific and closely linked to the
use of picture books for story-telling. For example,
Carter et al. (2005) found that Kenyan children had
difficulties interpreting picture stimuli; children were
found to have difficulty with perspective; particularly
when interpreting pictorial representations of preposi-
tions. They confused items of a similar shape, or had no
experience of the item depicted. This is to be expected
in a multicultural context, given that many items used
on a regular basis by one community may never be
used by another (e.g. cooking utensils and food items).
Cultural identity also subsumes differences in beliefs
and everyday practices, in religious celebrations and,
less obviously, in culture-specific language styles of, say,
storytelling (Cortazzi 1993, Minami 2002), which could
affect language assessments using narratives.
Many formal language tests require children to label
pictures of objects; however, Lidz and Pena (1996)
raise the crucial point that in different cultures parents
use different strategies to direct their childs attention.
They cite research in which Puerto Rican and African
American mothers were found to comment predomi-
nantly on object function rather than object labelling.
Consequently, for cultural reasons, some children from
diverse backgrounds may not perform well on object
labelling in tests of vocabulary, thus providing a negative
view of their linguistic competence.
Linguistic bias
The strict marking criteria of many standard assessments
do not make allowances for linguistic or dialectal
variation. Cheng (1997) maintains that linguistic
variation can affect all parameters including phonology,
syntax and pragmatics. Laing and Kamhi (2003) suggest
that this can be made worse as a result of disparity
between the language of the clinician and the child.
Perceived poor performance may be the result of
linguistic difference rather than impairment; conversely,
the clinician may wrongly attribute genuine errors
to linguistic variation and over estimate the childs
competence.
Disproportionate representation in normative data
Pring (2005) notes that standardized tests allow
clinicians to compare a childs score with a specified
population. If a test has been standardized among
white, monolingual children living in England, then
comparisons with children from other culturally or
linguistically diverse backgrounds will not yield scores
which can be reliably compared. London SIG Bilingual-
ism (2007a) guidelines recommend that children should
617
be assessed in all the languages which they use.
Translation might seem a solution, however experts
strongly advise against simply translating tests into other
languages due to notable differences in semantics, syntax
and phonology (Roseberry-McKibbin 2002, London
SIG Bilingualism 2007a, Bedore and Pena 2008, Martin
2009). Bedore and Pena (2008: 17) point out that
children speaking other languages do not necessar-
ily follow the same developmental trajectories as
English. The majority of language assessments which
have been comprehensively and appropriately adapted
from English to another language are also normed
on monolingual groups (Bedore and Pena 2008). Yet
tests designed for use with monolingual children are
unlikely to provide an accurate picture of a bilingual
childs communicative competence; at best they provide
a reasonable estimation. Research shows that normally
developing bilingual children tend to perform lower on
norm-referenced tests than their monolingual peers even
when one or both languages are assessed independently
(Thordatottir et al. 2006). It is this which can lead to
over-identification of speech and language disorders in
bilingual children. Grosjean (1989) emphasizes that a
bilingual speaker is not the sum of two monolinguals,
but an individual with a unique, integrated, linguistic
profile: a bilinguals linguistic repertoire generally has a
unique profile of the uses of languages across domains
and topics with particular speakers in specific contexts.
Removing biasevaluating some solutions
Modifying assessment content and format
So is it possible to remove the bias from standard-
ized assessments? Irvine Saenz and Blake Huer (2003)
conducted an investigation into the effects of test
modification using the CELF-3 (Semel et al. 1995).
First an unmodified version, then a modified version of
the test was administered to 25 LN bilingual Latino
children. Test modifications included rewording or
expanding instructions, providing additional response
time, providing credit for the use of dialect and using real
items and objects if the participants lacked experience
of books and pictures. Although the modifications did
improve raw test scores, the difference between the
unmodified and modified scores was not statistically
significant. On both versions the participants performed
below the mean. In contrast, they are reported as
performing within the standard range on familiar
academic tests. Test performance has been shown to
be affected by experience and task familiarity (Pena
and Quinn 1997). Irvine Saenz and Blake Huer (2003)
observed that despite the lack of statistical significance,
repeated testing provided students with the opportunity
to develop strategies which improved their raw scores.
618
They concluded that there might be a case for use of
DA which includes a test, teach, retest format (discussed
below).
Re-standardization and inclusion of minority groups
Attempts have been made to re-standardize assessments
by modifying the sample population to include a larger
number of children from ethnic minority backgrounds.
Washington and Craig (1992) issued the Peabody
Picture Vocabulary TestRevised (PPVT-R; Dunn and
Dunn 1981) to African American preschool children
but found that even with scoring allowances the test
was racially biased, indicated by a lack of performance
spread, with 91% performing significantly below the
mean. A third version of the PPVT was developed
(PPVT-III; Dunn and Dunn 1997) with modified
assessment questions following a pilot with three
major American groups (African American, Hispanic
and Native American). These ethnic groups were
then included within the normative sample in line
with their percentage representation amongst the total
population. Washington and Craig (1999) examined
the performance of 59 African American preschool
children on the PPVT-III and found that that they
performed within normal limits: there was no statisti-
cal difference from the normal standard distribution.
They concluded that modifying the questions and re-
standardizing the test resulted in a less biased assessment.
However, Stockman (2000) points out that reduced bias
for some groups may have come at the expense of being
able to make clinical judgements about others. In a
re-examination, Stockman found that the raw scores
resulted in much higher standard scores in the PPVT-
III than the PPVT-R suggesting that the modified test
content resulted in the PPVT-III being easier. She
argues that this could result in failure to identify correctly
LI in children from other ethnic groups, particularly
Caucasian ones. Stockman concludes that test bias
would need to be investigated for all major groups of
children.
Critical evaluationnorm-referenced measures
Given the ethnic diversity in Britain and many other
countries, one could speculate whether it would be
possible to devise a one-size-fits-all norm-referenced
standardized assessment. Mennen and Stansfields
(2006) study of SLT service delivery in three UK cities
identified in excess of 60 languages spoken in the
smallest city. Whilst some minority populations may
be relatively static (such as Native American populations
within the USA), the UK is undergoing massive shifts in
its social diversity as a result of immigration (Kings Fund
2006). Re-standardizing any assessment to take account
Caroline De Lamo White and Lixian Jin
of all multiethnic variation could pose an insurmount-
able challenge. Nevertheless, in areas where there is a
high proportion of people from one particular ethnic
group with a similar linguistic and cultural background,
it is worth re-standardizing an assessment on that
specific population. Chavda and Jin (2003) have redevel-
oped and re-standardized the STASS (South Tyneside
Assessment of Syntactic Structures; Armstrong and
Ainley 1988) for use with Gujarati-speaking children
living in the UK. However, Irvine Saenz and Blake
Huer (2003) comment that test renorming requires
significant time and resource and that tests which
have been modified and/or renormed should have their
reliability and validity evaluated. Moreover, to focus on
bilingual children, norm-referenced measures require a
relatively homogenous group: yet almost by definition,
a bilingual population embraces diversity and extensive
heterogeneity (Thordatottir et al. 2006: 4). Second
language acquisition is by its very dynamic nature, with
children learning at different times and rates. Even if
children from the same ethnic origin were placed into
one group, these children may be at different stages
of language acquisition as a result of differing life
experiences. This also raises the question of whether
the nature and dimensions of bilingualism should be
considered (e.g. sequential or simultaneous), but this
crucial aspect was not mentioned in any of the above
test-modification research papers. Mattes and Omark
(1984) argue that the development of a standardized
test suitable for all bilingual children is unrealistic due
to the heterogeneity of the client group.
Criterion-referenced (CR) measures
CR measures also evaluate a childs existing knowledge
base, but within a specific domain. In contrast to norm-
referenced tests which are interpreted in relation to the
performance of others, CR measures are interpreted in
relation to predetermined performance criteria; that is,
a set of behaviours deemed to constitute an acceptable
response (Terrell et al. 1992). For the purpose of
identifying speech and language disorders, these may
be standards of performance based on developmen-
tal data or language descriptions. McCauley (1996)
suggests that CR measures can be used when norm-
referenced measures are unavailable or inappropriate,
which as shown is often the case for bilingual children.
The advantage of using CR measures is that the
clinician can use materials and interaction patterns
which will be familiar to the child and thus try to reduce
cultural/linguistic bias. Therefore, this assessment tends
to be informal and includes techniques such as language
sampling and probing.
In the absence of a suitable standardized test for a
NigerianAmerican child exposed to English and Ibo
SLT assessment approaches with bilingual children
(a Nigerian language), Terrell et al. (1992) used the
fathers responses as the basis for evaluating the childs
responses. They termed this method of assessment
ParentChild Comparative Analysis (PCCA) (Terrell
et al. (1992: 34). Essentially this involved adminis-
tering the same standardized speech and language
tests to both the parent and the child and obtaining
spontaneous speech samples. When the childs response
deviated from Standard English it was compared with
the parental responses and compared with developmen-
tal norms for the childs age. Any child response failing
to match that of the parent and falling outside the
scope of normal (monolingual) language development
was deemed disordered.
Further research provides evidence that CR measures
may provide an alternative method of assessment.
Restrepo (1998) found that SLTs using a combination
of language sampling (analysing the mean length and
number of errors per T-unit) combined with a parental
report and family history were able to discriminate
between bilingual children with and without LI with
a high level of accuracy (> 90%). However, a recent
study (Paradis 2005) found that analysis of morphosyn-
tax in isolation did not accurately differentiate between
children with and without LI.
Critical evaluationcriterion-referenced measures
Whilst research provides evidence that CR measures
can be used to identify LI in bilingual children,
studies (Restrepo 1998) suggest that language analysis
needs to be supplemented by obtaining background
information from family/carers in order to differenti-
ate reliably between children with and without LI. This
provides evidence that the sociocultural approach should
underpin this assessment procedure.
Laing and Kamhi (2003) note that CR measures
are only as good as the data on which they are based.
CR measures may seek to draw upon developmen-
tal data regarding the process of language acquisi-
tion. Progress is being made within this field, but
Thordatottir (2005) suggests that research into patterns
of language acquisition in languages other than
English is comparatively lacking. Whilst Martin (2009)
acknowledges that difficulties with morphology are
symptomatic of LI, language acquisition patterns and
syntax have been found to differ across languages.
For example, Thordatottir (2005) found that French-
speaking children had a higher mean length of utterance
but smaller vocabulary size than English-speaking
children. This asymmetry makes comparing a childs
language skills across two or more languages addition-
ally complex.
One advantage of using CR measures is that their
application within a specified domain enables the
619
clinician to test definite clinical hypotheses and in this
way define the exact nature of the difficulty. Another
advantage of using an informal assessment technique
is that cultural differences can be embraced. However
Terrell et al. (1992) suggest that CR measures can be just
as biased as standardized measures if the clinician lacks
knowledge and awareness about the clients culture. One
way to overcome this is to involve a bicultural co-worker
and for clinicians to become familiar with the clients
languages and cultures through additional reading and
research.
Terrell et al. (1992) describe a single case study
of ParentChild Comparative Analysis. No further
research appears to have been carried out into the validity
of using this particular technique, which weakens
the reliability of its application with other bilingual
children.
Language-processing measures
In contrast to norm-referenced and CR measures,
language-processing tasks are independent of the childs
existing knowledge base so are considered to be a
less biased, potential alternative to the traditional
knowledge dependent tests discussed. Language-
processing tasks (LPTs) indirectly assess a childs
underlying processing functions and include memory
tasks such as non-word repetition (NWR; Gathercole
et al. 1994); competing language processing tasks
(CLPT; Gaulin and Campbell 1994); and perceptual
tasks such as the Revised Token Test (RTT; McNeil and
Prescott 1978). LPTs are therefore said to be process-
ing dependent. Gathercole and Baddeley (1990) found
that children with specific language impairment (SLI)
had poor language-processing abilities and performed
substantially below their age-matched controls in NWR
tasks.
With this in mind, Campbell et al. (1997) compared
the performance of Caucasian and minority ethnic
children on three LPTs (NWR, CLPT, RTT) and
one norm-referenced assessment. They found that the
minority ethnic children scored significantly lower on
the standardized test but did not differ on the LPTs.
Further research by Dollaghan and Campbell (1998)
provided evidence that using a NWR task had a high
degree of accuracy in distinguishing between children
with and without LI independent of ethnicity. Rodekohr
and Haynes (2001) research supports these findings:
they found that language-processing measures were able
to differentiate between LI and LN for both Caucasian
and African American groups.
However, children from minority ethnic groups
may not be necessarily bilingual, though their language
repertoires may well include culturally diverse features.
Kohnert et al. (2006) conducted research to determine
620
whether NWR and CLPT could be used to differenti-
ate between LN monolingual, LI monolingual and LN
bilingual children. The bilingual children are described
as proficient SpanishEnglish speakers. The researchers
hypothesized that if language-processing measures were
without sociolinguistic bias and sensitive enough to
detect LI, then bilingual and monolingual LN children
would perform similarly. In contrast, the LI children
would be expected to perform significantly below the
other two groups. However, results showed that the
LN monolingual children performed best, the bilingual
children performed significantly below their monolin-
gual peers, and the LI children performed signifi-
cantly below both the other two groups. Whilst there
was significant differentiation between the groups, the
researchers (Kohnert et al. 2006) suggest that NWR
and CLPT are not independent of previous language
experience and so cannot be considered to be free
from linguistic bias. Children who were less proficient
bilinguals than those in their study may not be expected
to perform as well. They concluded that NWR and
CLPT would not be good indicators of LI in bilingual
children.
By extension, Stokes et al. (2006) found that NWR
was unable to distinguish between groups of LI and
LN monolingual Cantonese children as there was
no significant difference between their performances.
They concluded that NWR is affected by existing
language knowledge and that the underlying processes
required to carry out this task are not impaired in
Cantonese-speaking children with LI. On the other
hand, these researchers also tested sentence repetition
(SR) and found that there was a significant difference
in performance between the LI and LN Cantonese-
speakers, which suggests that this may be an avenue
for further investigation.
Critical evaluationlanguage-processing measures
Whilst NWR is reported as easy and quick to administer
(Laing and Kamhi 2003), it cannot be used alone for
diagnostic assessment. NWR and CLPT do not inform
the clinician about the precise nature of the underly-
ing difficulty, they simply aim to differentiate between
children with and without LI.
NWR may hold promise for distinguishing between
LN and LI minority ethnic English-speaking children,
but research results from Kohnert et al. (2006) showed
that NWR and CLPT were not able to distinguish
between LN and LI bilingual children. This raises a
question of whether NWR can be truly free from
linguistic bias or independent of previous language
experience given that all the worlds languages are subject
to language specific phonotactic rules (Roach 2000).
Caroline De Lamo White and Lixian Jin
The work of Stokes et al. (2006) provides compelling
evidence against the assumption that NWR is indepen-
dent of language experience. Existing research into the
efficacy of NWR does not provide sufficient evidence
to warrant its use as an assessment tool for use with
bilingual children. On the other hand, LI affected
Cantonese childrens performance on the SR task, which
suggests that more research is needed to determine the
value of this task as a diagnostic tool. Research to
evaluate the diagnostic value of SR for children who
speak languages other than Cantonese is warranted; the
diagnostic value of SR for use with bilingual children
needs to be established. Martin (2009) concedes that
language-processing measures have yet to realize their
full potential. Kohnert et al. (2006) suggest that more
research is needed into the underlying cognitive systems
involved in language-processing.
Dynamic assessment (DA)
Similar to languageprocessing measures, DA does
not rely solely upon a childs existing knowledge or
life experience, instead it attempts to measure their
potential for learning (Hasson 2009). DA is based
on the work of Vygotsky (1978, cited in Gutierrez-
Clellen and Pena 2001: 212) and the zone of proximal
development (ibid), which encapsulates how a childs
cognitive development takes place as a result of social
interactions with others, especially within tasks which a
child is as yet unable to perform aloneadult supported
learning in this zone is also known as mediation and
scaffolding. DA commonly follows a testteachretest
format and in doing so, provides information about
current levels of performance, the effect of interven-
tion upon performance and highlights the best strategies
for supporting further learning. Evaluation is also made
about a childs learning behaviour during the mediation
phase. The term modifiability describes to what extent
a child is capable of change in response to interven-
tion (Lidz and Pena 1996: 368). A modifiability score
is based on the level of examiners effort required, the
childs responsiveness/motivation and demonstration of
a transfer of learning. This technique can be described
as diagnostic therapy because assessment and therapy
occur simultaneously. Other DA methods discussed
in the literature include task/stimulus variability and
graduated prompting. However, there was less attention
given to these techniques in the literature reviewed, so
they are not discussed in detail here.
Testteachretest
Pena et al. (1992) carried out research into the effective-
ness of using testteachretest for assessing the language
abilities of 50 bilingual children (African American and
SLT assessment approaches with bilingual children
Puerto Rican children exposed to Spanish and English
teaching environments on a daily basis).
The children were split into two groups, LN and LI,
and tested using two formal standardized assessments;
a test of vocabulary (object labelling) and a test of
comprehension (object description) in order to obtain
baseline scores. Both groups of children performed
significantly better on the test of comprehension. So for
6 weeks both groups of children were taught about object
labelling. During this phase the childrens performance
was also rated in relation to their learning behaviour or
modifiability. The children were finally retested.
Whilst all of the childrens raw scores improved upon
retest, the LN children performed significantly better
on the vocabulary retest and had higher modifiability
scores than the LI children. The LI childrens modifia-
bility scores indicated that they required a higher level of
examiner effort, were less responsive to input and did not
readily transfer their learning. Therefore the testteach
retest format, combined with the modifiability rating
was shown to be effective at differentiating between the
children with and without LI, irrespective of cultural
identity or language experience.
Research by Ukrainetz et al. (2000) supports that
of Pena et al. (1992). They undertook a testteach
retest research study with Native American children
and divided the children into two groups; stronger
and weaker language-learners. The testteachretest
assessment method reliably differentiated between the
two groups.
Lidz and Pena (1996) undertook a small case study
with two bilingual Latino American preschool children,
A and B, who were performing significantly below the
mean on a standardized vocabulary test. The testteach
retest approach was used to evaluate their language
ability. B responded well in mediated sessions and
achieved a high modifiability score, whereas A needed
high levels of examiner effort and had difficulty learning
the new concepts. B achieved a significantly higher
retest score following mediation; A showed very little
improvement. B was diagnosed as LN with language
differences, while A was diagnosed as LI. The clinician
was able to provide the class teacher with an insight into
the nature of As difficulties.
Critical evaluationdynamic assessment
Research provides evidence that DA is effective at
identifying LI in bilingual children. Cheng (1997)
comments that a distinct advantage of DA, is
that it allows the clinician to assess a child over
time thereby providing opportunities to monitor the
childs performance and outcomes of intervention at
regular intervals. The dual, diagnostic and therapeu-
tic components of DA enable the clinician to adjust
621
intervention plans and mediational strategies to meet
the needs of the child. The mediation phase provides
opportunity to identify strategies that facilitate the
greatest change and enables the clinician to give
recommendations to others involved with the child.
Mediation can be used to predict a childs responsiveness
to intervention (Hasson and Joffe 2007).
However, Isaac (2002) argues that clinicians
must consider the childs cultural background before
making judgements about the childs responsiveness to
mediation. Carter et al. (2005) discuss the influence of
culture upon performance and note that culture can
influence variables such as eye contact and turn-taking.
Ukrainetz et al. (2000) concede that DA is most reliable
when a child is compared with several others from the
same cultural group, so the clinician may want to assess
others to obtain a frame of reference. When this is
not feasible, ethnographic information could be used to
bridge the knowledge gap. Ethnographic assessment is
discussed in more detail in the section entitled sociocul-
tural approach.
Diagnostic therapy is used informally by many SLTs,
but DA is not currently used as part of routine evaluation
(Caesar and Kohler 2007). DA is very time consuming
and requires intensive contact, particularly during the
mediation phase. Bray et al. (1999) recognize that in a
climate of stretched resources clinicians may have less
available direct contact time for their clients. Another
issue with DA is inter-rater reliability (Hasson and Joffe
2007) clinicians are free to modify their mediation and
assessment procedures to suit the child, but cannot be
sure of their validity or reliability unless they have access
to other clinicians DA with the same clients, which is
unlikely.
Sociocultural approach
The sociocultural approach is a technique advocated
by various experts (Westby et al. 1990, Cheng 1990,
1997, Isaac 2002, Roseberry-McKibbin 2002, Martin
2009) and is based on the assumption that an individ-
uals language use is inseparable from the influence of
their social and cultural environment. Martin (2009)
refers to Bronfenbrenners model, the ecology of human
development, in which the child is at the centre
of their ecosystem. In contrast to other methods
of assessment that attempt to decontextualize the
assessment procedure in a bid to reduce the influence
of external factors on performance; the sociocultural
approach takes a holistic view of the childs developing
speech and language skills within the context of their
social and cultural environment.
Figure 2, taken from Anderson and Van der Gaag
(2005: 91), shows the influence of each social subsystem
upon the next, and how jointly they shape the childs
622
Figure 2. Bronfenbrenners ecology of human development after Bronfenbrenners model (cited in Anderson and Van der Gaag 2005: 91).
experiences and development. The microsystem is the
childs immediate environment and includes those
who interact directly with the child. The mesosys-
tem denotes the interplay between the components of
the microsystem: significantly, when health profession-
als work alongside the parents and each other for the
good of the child. Components of the exosystem and
macrosystem tend to have a more indirectbut signifi-
cant impact upon the child. For example, governmental
decisions may affect the amount of support a child is
able to receive in school. The key to the left illustrates
the components of each system.
Cheng (1997) contends that the rationale for using
a sociocultural approach is that by collecting data and
viewing it through the perspective of the target culture,
the clinician is able to make a less biased and more valid
interpretation of findings: thus the clinician is less likely
to draw conclusions which lead to misdiagnosis.
In the course of any SLT assessment, the clinician
would take a case history, speak with parents and
carers and carry out observations of the child. However,
ethnographic assessment involves a more comprehensive
procedure. It is worth mentioning that ethnographic
interviewing has a long history of use within SLT,
e.g. getting information from family members and
understanding clients from the perspective of families
(Hammer 1998). Its use is advocated with all client
Caroline De Lamo White and Lixian Jin
groups, monolingual and bilingual alike. Westby et al.
(1990) suggest that ethnographic interviewing enables
the clinician to see the world from the clients perspec-
tive and in doing so places them at the heart of
intervention goals and strategies. They argue that this
client-centred perspective makes therapeutic interven-
tion more effective.
Cheng (1990) outlines the procedure as follows:
Observing the child in multiple settings with a
variety of different people to understand with
whom the child interacts and how, what they talk
about and what activities they enjoy.
Observation should include a setting which
facilitates a low level of anxiety and a high level
of motivation to give the child opportunity to
demonstrate their full communicative potential.
Interviewing members of the family to learn
about family dynamics, interactions, attitudes and
cultural differences, and about the childs language
history to find out if the child is a simultane-
ous or sequential bilingual. Westby et al. (1990:
102) outline the type of questions that should be
used to elicit ethnographic data and the principles
which should be adopted to gain an extended
and unambiguous response. This includes asking
SLT assessment approaches with bilingual children
descriptive, open-ended questions such as Tell me
about a typical day and limiting the use of closed
and leading questions.
Engaging in a childclinician session to learn
about the childs interests, conversational strategies
(maintenance, breakdown and repair) and interac-
tion style (pragmatic skills).
Consultation with teachers and teaching assistants
to understand their observations and perspective.
The clinician is recommended to reflect on their own
cultural perspectives, world views and biases. Increased
awareness of self and others helps to reduce barriers and
promote good cross-cultural communication.
Cheng (1990: 115) argues that the comprehen-
sive and detailed nature of ethnographic assessment
increases the representativeness, reliability and validity
of [collected] data. This approach also helps to create
a language profile for the child: which languages are
spoken in each setting and the means, reasons and
opportunities for communicating in each language. For
example, the child may speak English at school with
teachers and peers and with siblings at home, but use
Gujarati at home with parents and grandparents.
RIOT procedure
Cheng (1997) is also an advocate of the review, interview,
observe and test (RIOT) procedure as a part of the
sociocultural approach. This procedure draws upon
other techniques discussed above, but modifies them
within the clinicians understanding of childs sociocul-
tural context and environment, culture and subsequent
development. Analysing all collected data in the light
of ethnographic data will contribute to the decision-
making process and reduce the chance of misdiagnosis.
Cheng (1997) describes the components of the RIOT
procedure as follows:
Review all documentation, client records, clinical
history and educational records.
Interviewconduct an ethnographic interview
with parents, carers, teachers and relevant others
to find out more about the childs language
history, family dynamics, interactions, attitudes
and cultural differences. For example, it would
be useful to build a profile of which languages are
spoken with each member of the family and which
languages are used at school.
Observe the child in multiple settings with a
variety of different people. This could include
observations at home with the immediate family
(parents and siblings) or with wider family
members (such as grandparents, cousins, etc.). In
conducting an ethnographic study, Komulainen
623
(2005) made contrastive observations in a nursery
and in a clinical setting (such as a therapy room).
She observed the childs interactions with peers
and adults (care staff and health professionals)
during individual and group activities and at
mealtimes. This meant observing interactions on
different days and at different times. In addition
to observing verbal communication, Komulainen
(2005) made a note of the situational context and
non-verbal interaction in order to interpret holisti-
cally a childs communicative ability. Westby et al.
(1990: 103) comment that every social situation
has multiple dimensions including the people
involved, places used, activities, routines, objects,
goals, time and feelings.
Test all languages which the child speaks
using (modified) formal assessments, informal
assessments including language sampling, and
alternative testing procedures such as DA.
In conducting research into bilingual childrens oral
proficiency, Sanchez (2006) developed a sociolinguis-
tic questionnaire to determine patterns of language use
amongst children. She was able to identify shifting
patterns of language use within different communi-
ties and consider the effect of language use upon test
data. The research carried out by Restrepo (1998)
which combined language sampling with parental
report also provides evidence that using the sociocul-
tural approach (in this instance the use of parental
interview) increases the accuracy of differentiating
between bilingual children with and without LI, which
could not be achieved by language sampling alone.
Critical evaluationsociocultural approach
The sociocultural approach is comprehensive and
encompasses aspects of the other approaches which
have been found to be effective indicators of LI in
bilingual children. The weaknesses inherent to each
approach are supplemented by ethnographic data which
jointly aids their implementation and interpretation. By
learning more about the childs cultural background,
the SLT can plan activities which are language and
culture specific. Cheng (1990, 1997) contends that only
a sociocultural approach truly accounts for differences
in cultural and linguistic background. Arguably, it
gives bilingual children opportunities to demonstrate
their full communicative competence because they are
observed in a variety of settings and their sociolinguistic
circumstances are better understood by the assessor.
However, the approach is not without its disadvan-
tages: it is an intensive, time-consuming and costly
procedure. It will often involve employing bilingual co-
workers/translators in order to be undertaken properly.
624 Caroline De Lamo White and Lixian Jin
Nevertheless, these limitations must be balanced by tic diversity that exists amongst bilingual clients cannot
significant benefits and the obligation to provide be represented in training courses, but carefully chosen
an equitable and diagnostically accurate service. The representative examples can be given, so that SLTs are
sociocultural approach may hold the key to reducing aware of the dimensions, scale and manner of variabil-
occurrences of misdiagnosis amongst bilingual children, ity as well as the range of cultural contexts, linguistic
thusin the long termit reduces costs by avoiding repertoires and cultural language styles. This should be
inappropriate spending. supplemented with opportunity to sample a good range
of assessment procedures and gain an understanding
of the associated difficulties. Continuous professional
Implications for assessment and the profession
training in this aspect is also a way forward to develop the
Isaac (2002) contends that standardized assessments can SLTs knowledge and competence in assessing bilingual
be used to gain a qualitative (descriptive) measure of a clients.
childs linguistic competence, and that clinicians should
not attempt to score the assessment for comparison with
Conclusion
normative data. She comments that clinicians should
use the test to analyse what skills the child has (in In conclusion, no single approach to identify language
the broadest sense) and consider ethnographic variables impairment (LI) in bilingual children is effective
which could explain poor performance. on its own. Criterion-referenced (CR) measures and
Cheng (1997) and Isaac (2002) maintain that dynamic assessment (DA) have been found to be the
standardized tests can be used as a component within a most effective identifiers of LI in bilingual children
broader ethnographic approach. Many researchers cited when integrated within the sociocultural approach.
here used standardized tests as the tool for implementing Norm-referenced, standardized measures can be used
other approaches such as CR measures and DA. In these to provide descriptive measures of a bilingual childs
instances the researchers adapted the scoring procedures language ability, but should not be formally scored and
and interpreted them differently. McCauley (1996: 124) compared with the monolingual norms. Interpretation
concurs with this view when she concludes that norm of assessments should be made in light of ethnographic
and criterion-referencing are probably best considered as information. A bilingual child should be assessed in all
modes of interpretation rather than mutually exclusive the languages they use and within each of their interac-
categories of tests. tion environments, involving a bilingual co-worker
Caesar and Kohler (2007) suggest that improve- where necessary. Figure 3 recommends that CR and DA
ments to practice will only be brought about when
undergraduate programmes have a greater focus on
alternative assessment procedures for bilingual children.
Roseberry-McKibbin et al. (2005) found that SLTs
whose undergraduate courses had a greater element of
assessment of bilingual children perceived a greater lack
of appropriate assessment procedures; at least this raised
awareness of the difficulties in assessing this client group.
SLTs are unlikely to modify practice (in line with Health
Professionals Council 2007: 2c.1) if they are unaware of
the need to do so. However, awareness raising alone is
unlikely to facilitate change. Caesar and Kohler (2007)
conducted a survey to investigate the type and frequency
of assessment approaches used by SLTs working with
bilingual children. Of the 400 plus SLTs who responded,
the majority stated that they used formal standardized
testing most frequently, thus providing evidence that
SLTs are over-reliant on standardized norm-referenced
measures. Reportedly, DA was not mentioned by any
of the clinicians responding to the survey. This implies
that undergraduate or postgraduate training needs to
give students opportunity to learn about and gain
practical experience of using alternative methods of
assessment within the context of the sociocultural Figure 3. Suggested framework for using a sociocultural approach
approach. Clearly, the full range of cultural and linguis- with bilingual clients.
SLT assessment approaches with bilingual children
contribute to a sociocultural approach; but this practice
with bilingual clients needs to be supported by updated
training programmes offered to speech and language
therapy (SLT) students and professionals.
Language-processing measures have not been found
to be good indicators of LI in bilingual children but
further research should be undertaken to determine the
diagnostic potential of sentence-repetition tasks.
Most of the relevant research has been carried out
in North America and it is important that studies
are carried out upon minority ethnic and bilingual
populations living in other countries where to be
bilingual is normal. More research is recommended
to analyse the inter-rater reliability and validity of DA
techniques so that clinicians have increased confidence
and understanding of their application.
Currently there may be an over-reliance on Western
standardized assessments (Gladstone et al. 2009) among
which more culturally appropriate assessment tools are
lacking in local multilingual and multicultural contexts.
This might explain the misrepresentation of bilingual
children on caseloads. However, with the constant
updating and improvement of SLT curricula and more
recently published research findings and outcomes on
bilingual assessments and SLI, e.g. over 50 papers
have been published or are in press since 2009 from
one European Union project, it is recommended that
SLT training and staff-development programmes should
contain updated elements on the assessment of bilingual
children and implementation of alternative assessment
procedures with the sociocultural approach, produced
within and outside Western countries, to enhance
the confidence of SLTs working with bilingual clients
(Harnett and OToole 2009). Such programmes should
include specific examples of linguistic and cultural
differences to equip SLTs with the knowledge and
skills to assess bilingual clients. They should encompass
relevant aspects of intercultural communication and
cultures of learning (Cortazzi and Jin 2002) to assist
SLTs in communicating with bilingual clients and
their carers within the sociocultural approach underpin-
ning their assessments. As a result, it is hoped that
bilingual children will receive a diagnostically accurate
assessment which is linguistically and culturally sensitive
and hence will only receive intervention when it is
actually merited.
References
ANDERSON, C. and VAN DER GAAG, A., 2005, Speech and Language
TherapyIssues in Professional Practice (London: Whurr).
ARMSTRONG, S. and AINLEY, M., 1988, South Tyneside Assessment of
Syntactic Structures (Ponteland: STASS).
ASPINALL, P., 2002, Collective terminology to describe the
625
minority ethnic populationthe persistence of confusion and
ambiguity in usage. Sociology, 36(4), 803816.
BAKER, C., 2006, Foundations of Bilingual Education and Bilingual-
ism, 4th edn (Clevedon: Multilingual Matters).
BAKER, C. and EVERSLEY, J., 2000, Multilingual Capital (London:
Battlebridge).
BEDORE, L. and PENA, E., 2008, Assessment of bilingual children for
identification of language impairment: current findings and
implications for practice. International Journal of Bilingual
Education and Bilingualism, 11(1), 129.
BRAY, M., ROSS, A. and TODD, C., 1999, Speech, Language and
Clinical Process and Practice (Baltimore, MD: Whurr).
CAESAR, L. and KOHLER, P., 2007, The state of school-based bilingual
assessment: actual practice versus recommended guidelines.
Language, Speech and Hearing Services in Schools, 38, 190
200.
CAMPBELL, T., DOLLAGHAN, C., NEEDLEMAN, H. and JANOSKY,
J., 1997, Reducing bias in language assessment: processing-
dependent measures. Journal of Speech, Language and Hearing
Research, 40(3), 519525.
CARTER, J., LEES, J., MURIRA, M., GONA, J., NEVILLE, B. and
NEWTON, C., 2005, Issues in the development of cross-
cultural assessment of speech and language for children.
International Journal of Language and Communication
Disorders, 40(4), 385401.
CHAVDA, P. and JIN, L., 2003, Assessment of Gujarati Syntactic
Structures (Ponteland: STASS).
CHAVDA, U., KAO, R., SOLDATOU, A., GARDNER, A., KNUDSON,
P., HONGJUN, S. SCHOON EBERLY, S. and VAN DYKE,
D., 2003, Important issues in the care and evaluation
of bilingual/multilingual children. International Pediatrics,
18(1), 813.
CHENG, L., 1990, The identification of communicative disorders in
AsianPacific students. Journal of Childhood Communication
Disorders, 13(1), 113119.
CHENG, L., 1997, Diversity: challenges and implications for
assessment. Journal of Childrens Communication Develop-
ment, 19(1), 5562.
COOK, V. (ed.), 2003, Effects of the Second Language on the First
(Clevedon: Multilingual Matters).
CORTAZZI, M., 1993, Narrative Analysis (London: Falmer).
CORTAZZI, M. and JIN, L., 2002, Cultures of learning, the social
construction of educational identities, in D. C. S. Li
(ed.), Discourses in Search of Members (New York: American
University Press), pp. 4978.
CORTAZZI, M. and JIN, L., 2004, Reflection on speechlanguage
therapists talk: implications for clinical practice and
education. International Journal of Language and Communi-
cation Disorders, 39(4), 477480.
CRUTCHLEY, A., CONTI-RAMSDEN, G. and BOTTING, N., 1997,
Bilingualism and specific language impairment in children
attending language units. European Journal of Disorders of
Communication, 32(2), 267276.
DOLLAGHAN, C. and CAMPBELL, T., 1998, Nonword repetition and
child language impairment. Journal of Speech, Language and
Hearing Research, 41, 11361146.
DUNN, L. and DUNN, L., 1997, Peabody Picture Vocabulary Test
Third Edition (PPVT III) (San Antonio, TX: Pearson).
DUNN, L. M. and DUNN, L., 1981, Peabody Picture Vocabulary Test
Revised (PPVT-R) (Circle Pines, MN: American Guidance
Service).
GATHERCOLE, S. and BADDELEY, A., 1990, Phonological memory
deficits in language disordered children: is there a causal
connection? Journal of Memory and Language, 29, 336
360.
626
GATHERCOLE, S., WILLIS, C., BADDELEY, A. and EMSLIE, H., 1994,
The Childrens Test of Non-word Repetitiona test of
phonological working memory. Memory, 2, 103127.
GAULIN, C. and CAMPBELL, T., 1994, Procedures for assessing
verbal working memory in normal school-age children
some preliminary data. Perceptual and Motor Skills, 79, 55
64.
GLADSTONE, M., LANCASTER, G., UMAR, E., NYIRENDA, M., KAYIRA,
E., VAN DEN BROEK, N. and SMYTH, R. L., 2009, Perspectives
of normal child development in rural Malawia qualitative
analysis to create a more culturally appropriate assessment
tool. Care Health and Development, 36(6), 346353.
GOULD, J., 2008, Non-standard assessment practices in the
evaluation of communication in Australian Aboriginal
children. Clinical Linguistics & Phonetics, 22(8), 643657.
GROSJEAN, F., 1989, Neurologists beware! The bilingual is not two
monolinguals in one person, in K. Bolton and B. Kachru
(eds), Critical Concepts in Linguistics (Abingdon: Routledge),
pp. 202213.
GUTIERREZ-CLELLEN, V. and PENA, E., 2001, Dynamic assessment
of diverse children: a tutorial. Language, Speech and Hearing
Services in Schools, 32, 212224.
HAMAYAN, E. and DAMICO, J., 1991, Limiting Bias in the Assessment
of Bilingual Students (Austin, TX: PRO-ED).
HAMMER, C. S., 1998, Towards a thick description of families:
using ethnography to overcome the obstacles to providing
family-centred early intervention services. American Journal
of SpeechLanguage Pathology, 7(1), 522.
HARNETT, S. and OTOOLE, C., 2009, Speech and language
therapists training and confidence in serving bilingual clients.
Journal of Clinical Speech and Language Studies, 17, 5773.
HASSON, N., 2009, The use of dynamic assessment in speech and
language therapy, in Unpublished Proceedings of the Royal
College of Speech and Language Therapists National Student
Study Day, De Montfort University, Leicester, UK, April
2009.
HASSON, N. and JOFFE, V., 2007, The case for dynamic assessment
in speech and language therapy. Child Language Teaching and
Therapy, 23(1), 925.
HEALTH PROFESSIONALS COUNCIL, 2007, Standards of Proficiency
Speech and Language Therapists (available at: http://www.
hpc-uk.org/assets/documents/10000529Standards_of_
Proficiency_SLTs.pdf) (accessed on 6 April 2009).
IRVINE SAENZ, T. and BLAKE HUER, M., 2003, Testing strategies
involving least biased language assessment of bilingual
children. Communication Disorders Quarterly, 24(4), 184
193.
ISAAC, K., 2002, Speech Pathology in Cultural and Linguistic Diversity
(London: Whurr).
JIN, L. and CORTAZZI, M., 1993, Cultural orientation and academic
language use, in D. Graddol, L. Thompson and M. Byram
(eds), Language and Culture (Clevedon: Multilingual Matters)
pp. 8497.
KINGS FUND, 2006, Access to Health Care and Ethnic
Minority Groups (available at: http://www.kingsfund.org.uk/
publications/briefings/access_to_health.html) (accessed on
10 April 2009).
KOHNERT, K., WINDSOR, J. and YIM, D., 2006, Do language-based
processing tasks separate children with language impairment
from typical bilinguals? Learning Disabilities Research and
Practice, 21(1), 1929.
KOMULAINEN, S., 2005, The contextuality of childrens communica-
tion difficulties in specialist practice: a sociological account.
Child Care in Practice, 11(3), 357374.
LAING, S. and KAMHI, A., 2003, Alternative assessment of
Caroline De Lamo White and Lixian Jin
language and literacy in culturally and linguistically diverse
populations. Language, Speech and Hearing Services in Schools,
34, 4455.
LI, W., 2007, The Bilingualism Reader, 2nd edn (London:
Routledge).
LIDZ, C. and PENA, E., 1996, Dynamic assessment: the model, its
relevance as a nonbiased approach, and its application to
Latino American preschool children. Language, Speech and
Hearing Services in Schools, 27, 367371.
LONDON SIG BILINGUALISM, 2007a, Culture (available at: http://
www.grafixdesign.net/www.sig.co.uk/culture.html) (accessed
on 26 November 2010).
LONDON SIG BILINGUALISM, 2007b, Good Practice for Speech and
Language Therapists Working with Clients from Linguistic
Minority Communities (available at: http://www.rcslt.org/
docs/Linguistic Minority Communities good practice for
SLTs Oct 2007 Final.pdf ) (accessed on 9 April 2009).
LONDON SIG BILINGUALISM, 2009, Developing a Culturally
Competent Service (available at: http://www.grafixdesign.net/
www.sig.co.uk/pdf/devculture.pdf ) (accessed on 26
November 2010).
MARTIN, D., 2009, Language Disabilities in Cultural and Linguistic
Diversity (Bristol: Multilingual Matters).
MATTES, L. and OMARK, D., 1984, Speech and Language Assessment
for the Bilingual Handicapped (San Diego, CA: College-Hill
Press).
MCCAULEY, R., 1996, Familiar strangers: criterion-referenced
measures in communication disorders. Language, Speech and
Hearing Services in Schools, 27, 121131.
MCNEIL, M. and PRESCOTT, T., 1978, Revised Token Test (Baltimore,
MD: University Park Press).
MENNEN, I. and STANSFIELD, J., 2006, Speech and language therapy
service delivery for bilingual children: a survey of three
cities in Great Britain. International Journal of Language and
Communication Disorders, 41(6), 635652.
MINAMI, M., 2002, Culture-Specific Language Styles, The Develop-
ment of Oral Narrative and Literacy (Clevedon: Multilingual
Matters).
MULTIVERSE, 2004, Minority Ethnic (available at: http://www.
multiverse.ac.uk/viewarticle2.aspx?contentId=369) (accessed
on 18 December 2009).
NICOLADIS, E. and GENESEE, F., 1997, Language development
in preschool bilingual children. Journal of SpeechLanguage
Pathology and Audiology, 21(4), 258270.
ODLIN, T., 1989, LANGUAGE Transfer: Cross-linguistic Influence on
Language Learning (Cambridge: Cambridge University
Press).
OFFICE OF NATIONAL STATISTICS, 2001, Population Size (available
at: http://www.statistics.gov.uk/CCI/nugget.asp?ID=273)
(accessed on 10 April 2009).
PAPOUTSIS KRITIKOS, E., 2003, Speechlanguage pathologists beliefs
about language assessment of bilingual/bicultural individ-
uals. American Journal of SpeechLanguage Pathology, 12,
7391.
PARADIS, J., 2005, Grammatical morphology in children learning
English as a second language: implications of similarities with
specific language impairment. Language, Speech and Hearing
Services in Schools, 36, 172187.
PENA, E. and QUINN, R., 1997, Task familiarity: effects on the test
performance of Puerto Rican and African American children.
Language, Speech and Hearing Services in Schools, 28, 323
332.
PENA, E., QUINN, R. and IGLESIAS, A., 1992, The application of
dynamic assessment: a nonbiased procedure. Journal of Special
Education, 26(3), 269280.
SLT assessment approaches with bilingual children
PENA, E., SPAULDING, T. and PLANTE, E., 2006, The composition of
normative groups and diagnostic decision making: shooting
ourselves in the foot. American Journal of SpeechLanguage
Pathology, 15, 247254.
PRING, T., 2005, Research Methods in Communication Disorders
(London. Whurr).
ROYAL COLLEGE OF SPEECH AND LANGUAGE THERAPISTS (RCSLT),
1996, Communicating Quality 3 (London: RCSLT).
RESTREPO, M., 1998, Identifiers of predominantly Spanish-speaking
children with language impairment. Journal of Speech,
Language and Hearing Research, 41, 13981411.
ROACH, P., 2000, English Phonetics and Phonology, 3rd edn
(Cambridge: Cambridge University Press).
RODEKOHR, R. and HAYNES, W., 2001, Differentiating dialect
from disordera comparison of two processing tasks and
a standardized language test. Journal of Communication
Disorders, 34, 255272.
ROSEBERRY-MCKIBBIN, C., 2002, Multicultural Students with Special
NeedsPractical Strategies for Assessment and Intervention, 2nd
edn (Oceanside: Academic Communication Associates).
ROSEBERRY-MCKIBBIN, C., BRICE, A. and OHANLON, L., 2005,
Serving English language learners in public school settings:
a national survey. Language, Speech, and Hearing Services in
Schools, 36(1), 4861.
SANCHEZ, L., 2006, Bilingual/second-language research and the
assessment of oral proficiency in minority bilingual children.
Language Assessment Quarterly, 3(2), 117149.
SEMEL, E., WIIG, E. and SECORD, W., 1995, Clinical Evaluation
of Language Fundamentals (CELF-3) (San Antonio, TX:
Psychological Corporation).
STOCKMAN, I., 2000, The New Peabody Picture Vocabulary Test
III: an illusion of unbiased assessment? Language, Speech and
Hearing Services in Schools, 31, 340353.
STOKES, S., WONG, A., FLETCHER, P. and LEONARD, L., 2006,
Nonword repetition and sentence repetition as clinical
markers of specific language impairment: the case of
Cantonese. Journal of Speech, Language and Hearing Research,
43, 219236.
STOW, C. and DODD, B., 2003, Providing an equitable service to
bilingual children in the UK: a review. International Journal
of Language and Communication Disorders, 38(4), 351377.
SWAN, M. and SMITH, B., 2001, Learner English, A Teachers Guide
627
to Interference and Other Problems (Cambridge: Cambridge
University Press).
TERRELL, S., ARENSBERG, K. and ROSA, M., 1992, Parent
child analysis: a criterion-referenced method for the non-
discriminatory assessment of a child who spoke a relatively
uncommon dialect of English. Language, Speech and Hearing
Services in Schools, 23, 3442.
THORDATOTTIR, E., 2005, Early lexical and syntactic development in
Quebec French and English: implications for cross-linguistic
and bilingual assessment. International Journal of Language
and Communication Disorders, 40(3), 243278.
THORDARTOTTIR, E., ROTHENBERG, A., RIVARD, M. and NAVES,
R., 2006, Bilingual assessment: can overall proficiency be
estimated from separate measures of two languages? Journal
of Multilingual Communication Disorders, 4(1), 121.
UKRAINETZ, T., HARPELL, S., WALSH, C. and COYLE, C., 2000, A
preliminary investigation of dynamic assessment with Native
American kindergartners. Language, Speech and Hearing
Services in Schools, 31, 142154.
WASHINGTON, J. and CRAIG, H., 1992, Performance of low-income
African American preschool and kindergarten children on the
Peabody Picture Vocabulary Test Revised. Speech, Language
and Hearing Services in Schools, 23, 329333.
WASHINGTON, J. and CRAIG, H., 1999, Performances of at-
risk African American preschoolers on the Peabody Picture
Vocabulary TestIII. Language, Speech and Hearing Services
in Schools, 30, 7582.
WESTBY, C., BURDA, A. and MEHTA, Z., 1990, Asking the right
questions in the right waysstrategies for ethnographic
interviewing. Journal of Childhood Communication Disorders,
13(1), 101111.
WIIG, E., SECORD, W. and SEMEL, E., 2006, Clinical Evaluation
of Language FundamentalsPre-school Second UK Edition
(London: Pearson).
WINTER, K., 1999, Speech and language therapy provision for
bilingual children: aspects of the current service. International
Journal of Language and Communication Disorders, 34(1), 85
98.
WINTER, K., 2001, Numbers of bilingual children in speech and
language therapy: theory and practise of measuring their
representation. International Journal of Bilingualism, 5(4),
465495.
Copyright of International Journal of Language & Communication Disorders is the property of Wiley-
Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email articles for
individual use.