Are = we meeting=20 the challenge? An evaluation of speech and language therapy services for = bilingual children in three centres of population

Ineke Mennen, = Sally=20 Johnston, and Jois Stansfield

1. =20 Introduction

This study investigates the current speech and language = therapy (SLT)=20 service to bilingual children within three cities in the United Kingdom = (UK),=20 i.e. two in Scotland (S1 and S2) and one in England (E1), and = specifically=20 addresses whether speech and language therapists (SLTs) believe that = their=20 service meets the professional guidelines of the Royal College of Speech = and=20 Language Therapists (RCSLT). Through a combination of interviews, postal = questionnaires, and Census data, information was gathered about the SLT=20 paediatric service provision in each of these cities. For each of the = three=20 centres we established: (i) the % of monolingual and bilingual children = in the=20 population, (ii) the % of bilingual and monolingual children on the SLT=20 caseload, (iii) the languages spoken by the bilingual children on the = caseload,=20 and (iv) whether SLTs in these cities believe that their service is = currently=20 meeting the RCSLT guidelines.

2. =20 Method 2.1 = Aims

Our = study aimed=20 to establish, whether (i) there is an over- or under-representation of = bilingual=20 or multilingual children on the SLTs paediatric caseloads in the cities = of=20 investigation; and (ii) = SLTs in=20 these cities believe that their service is currently meeting RCSLT=20 guidelines. The=20 choice of cities was primarily motivated by the fact that no published = data on=20 bilingual clients in Scotland are available, and we wanted to compare = service=20 provision in two Scottish cities to that in an English city which is = renown for=20 its multicultural population. Throughout this=20 paper, the term bilingual is used to refer to individuals who use two or = more=20 languages in any of its modalities (Mackey, 1968) and we do not = distinguish=20 between bilingualism, multilingualism, or biculturalism and=20 multiculturalism.

2.2 = Population=20 statistics, interviews and questionnaires

We = used Census=20 data to establish the number of bilinguals in the population in the UK. = Census=20 data on ethnic groups give some indication of the linguistic minorities = that are=20 present, but do not provide an unequivocal measure of bilingualism since = there=20 will be individuals in the minority groups who are monolingual just as = there may=20 be bilinguals in the self-categorised white group. Another problem with = Census=20 data is the fact that they are organised around geographical areas, = whereas data=20 on SLT services are organised around NHS Trusts. As a result, = information=20 obtained from census data is not necessarily fully compatible with = information=20 from Trusts (cf. Winter, 1999).

Interviews were=20 held with the managers of the NHS Trust providing paediatric services = within=20 each area, to identify service coordinators in each of the areas who = were=20 responsible for service delivery to paediatric bilingual clients. The = two large=20 English and Scottish cities (E1 and S2) had service coordinators, and = these were=20 each sent a questionnaire. As the smaller Scottish city (S1) did not = have a=20 designated paediatric service coordinator, questionnaires were sent to a = stratified random sample of 25 individual SLTs working with paediatric = clients.=20 The questionnaire addressed the following issues: Caseload and languages = represented; Policy and RCSLT guidelines; Accessibility of service; = Support=20 services; and Training/education.=20 This paper only reports on the first phase of the research, which = involved=20 only the former two of the above issues.

3. =20 Results 3.1 = Ethnicity and=20 population

As summarised in Table I, the total population of children = up to and=20 including 17 years of age at the 2001 Census was 82,529 in S1, 120,410 = in S2,=20 and unknown in E1. For E1, there were at the time of this research no = population=20 data available which were categorised by age. Therefore, the population = data of=20 E1 in Table I refer to the total population (children and adults) at the = 2001=20 Census and amounts to 977,087. = When=20 the figures for the non-white ethnic children are collated together, = they amount=20 to 9 % of the total child population of S1, whereas the white ethnic = children=20 amount to 91 %. For S2, the non-white = ethnic groups=20 represented=20 10 % of the child population,=20 whereas the white ethnic group represents 90 % of the child population. For E1 (for = which there=20 are no data for non-white children available), the non-white ethnic = groups (i.e.=20 adults and children) amount to 29.6 %, whereas the white ethnic group = represents=20 70.4 % of the total population.

Table=20 I. Total population = and=20 ethnicity of residents in S1, S2 and E1. Source: 2001 Census, = Neighbourhood=20 Statistics

Ethnic=20 Group

S1

S2

E1

White

90.99 %

90.35 %

70.4%

Black

0.49 %

0.38 %

6.1%

Indian

0.66 %

0.92 %

5.7%

Pakistani

1.72 %

4.83 %

10.7%

Bangladeshi

0.27 %

0.05 %

2.1%

Chinese

0.93 %

0.65 %

1.2%

Other-Asian

0.37 %

0.54 %

1.0%

Mixed

1.50 %

0.68 %

2.9 %

Other=20 ethnic

0.47 %

0.44 %

Other=20 white

2.60 %

1.16 %

Total

100.00 %

100.01 %[1]

Total=20 Population

82,529 =20

120,410

977,087

3.2 = Caseload

From the sample of paediatric speech = and=20 language therapists working in S1, 77% (14/18) had bilingual clients on = their=20 caseloads. These therapists all reported that bilingual clients = represented=20 between 1% and 25% of their total caseloads, with a range of 1-26 = bilingual=20 clients. The total sample caseload in S1 was estimated at 1162 children, = with an=20 estimated 75 (6.45 %) bilingual or multilingual children.

Eighteen = languages other=20 than English were spoken by the children on the SLT caseloads in City = S1.=20

City=20 S2 has a specialist bilingual team, which is responsible for = providing SLT=20 to bilingual children. This means that all staff (currently three SLTs) = have=20 bilingual children on their caseloads. The total paediatric = caseload for=20 S2 was 3613, and the number of bilingual paediatric clients was 207,=20 representing 5 % of the total paediatric caseload in S2. However, the figures for the = number of=20 bilingual children on the paediatric caseload may be a little on the low = side,=20 as it appears that some bilingual children =96 especially those with = motor=20 problems or structural abnormalities =96 are being seen elsewhere in the = service.=20 A total of eighteen different languages was = recorded in=20 statistical records and case notes as the languages spoken by the = bilingual=20 clients on therapists=92 caseloads.

The number of SLTs working with = paediatric clients in E1 was estimated to be 70, and it was reported = that all=20 therapists are expected to work with bilingual clients. The total = paediatric=20 caseload in E1 was estimated at 8000 children. Of these, an estimated = 3335=20 clients (41.69%) are bilingual or multilingual. A total of twenty one = languages=20 other than English was represented on the caseload. Table II summarises = the=20 information about the paediatric caseload in the three centres of=20 investigation.

Table II. The paediatric caseload in S1, S2 = and E1,=20 with the total number of clients, the number and percentage of = monolingual and=20 bilingual clients on the caseload, the total number of languages and the = five=20 most prevalent languages other than English (OTE) spoken by the clients, = as well=20 as the % of the total caseload.

Paediatric Caseload

Ethnic=20 Group

No. of clients

% of total caseload

No. of clients

% of total caseload

No. of clients

% of total = caseload

Total no of clients

1162

100 %

3613

100 %

8000

100 %

Monolingual clients

1087

93.55 %

3406

95 %

4665

58.31 %

Bilingual clients

6.45 %

207

5 %

3335

41.69 %

Most prevalent languages

Panjabi

1.46 %

Panjabi

3.0 %

Panjabi

25.00 %

Hindi-Urdu

1.20 %

Hindi-Urdu

0.83 %

Bengali

7.05 %

Bengali

0.52 %

Farsi

0.28 %

Pashto[2]

6.25 %

Arabic

0.43 %

Swahili

0.28 %

Kokni[3]

0.50 %

Gaelic

0.43 %

Cantonese

0.25 %

Cantonese

0.50 %

No of languages (OTE) spoken by=20 clients

3.3 = Under- or overrepresentation

When we compare the proportion of non-white ethnic children = in the=20 population to the proportion of bilingual children on the SLT caseload = in City=20 S1, it appears that there are less children on the caseload (6.45) than = one=20 would expect from the population data (9 %). This suggests that there is = an=20 under-representation of bilingual children on therapists=92 caseloads, = which=20 possibly indicates that some of them may be falling through the net. As = in City=20 S1, there appears to be an under-representation of bilingual children on = therapists=92 caseloads in City S2, with less bilingual children on the = caseload=20 (5 %) than one would expect from the population data (10 %). However, it = must be=20 noted here that the caseload data provided only refer to the specialist=20 bilingual service, and some bilingual children may be seen elsewhere in = the=20 service. Although this may increase the total number of bilingual = clients on the=20 caseload, it is unlikely that it will increase with a full 5%, and this = would=20 not alter our conclusion of an under-representation. In City E1 there = appear to=20 be considerably more bilingual children on the caseload (41.69 %) than = one would=20 expect from the population data (29.60 %), suggesting an = over-representation of=20 bilingual clients on the SLT caseload. The population data, however, = were not=20 categorised by age, and the figure for non-white ethnic children aged 0 = to 17 is=20 likely to be larger than the population figure presented here. Therefore, the apparent=20 over-representation may be due to a larger number of children in the = non-white=20 ethnic group.

3.4 = Records, = policy and RCSLT=20 guidelines

The two cities with a larger number of non-white ethnic = residents=20 (City E1 and S2) have designated service coordinators who are = responsible for=20 service delivery to paediatric bilingual clients. The paediatric SLT = service in=20 these two cities also have a published operational policy for working = with=20 bilingual clients, and information about the languages spoken by clients = is=20 stored in statistical records and case notes (E1 even stores information = about=20 the languages spoken by carers/parents in statistical records). = Conversely, the=20 smaller Scottish city (S1) with the smallest number of individuals from=20 non-white ethnic groups does not have a specialist service coordinator, = the=20 service does not appear to have a published policy for working with = bilingual=20 clients, and information about language use of parents/carers and = clients is=20 either not recorded or only recorded in case notes.

Interestingly, only the respondent = from the=20 English city (E1) believed that their service was currently able to meet = all=20 RCSLT guidelines. Respondents for the Scottish cities believed that = their=20 service was only partially able to meet RCSLT guidelines (S2), or that = it was=20 unable to meet any of the guidelines (S1). These views reflect our = findings of=20 an under-representation of bilingual clients on the SLT caseloads in = both=20 Scottish cities.

4. =20 Conclusion

This study showed that only the SLTs in E1 are confident = that they=20 are meeting the challenge of providing an equal service to bilingual and = monolingual clients. SLTs in the two Scottish cities are unable to = (fully) meet=20 RSLT guidelines, and bilingual children are under-represented on the paediatric SLT=20 caseloads. = The study highlights that service planning is hindered by a = lack of=20 easily retrievable statistical information about the number of bilingual = and=20 monolingual individuals in the population, the number of clients on SLT=20 caseloads, and the languages spoken by clients. If we want to provide an = equal=20 service for bilingual and monolingual individuals, there needs to be a = change in=20 how data on linguistic diversity in society is collected and = disseminated so=20 that informed decisions can influence the future of quality services to = minority=20 groups. It is a recommendation of this study that all SLT services = should=20 collect information about the languages their clients use and store this = as an=20 easily retrievable statistic. Such data collection will not only help = service=20 planning but it will also be a move towards recognising the importance = of social=20 inclusion, community cohesion and good community and race relations.=20

References

Ethnologue: Languages of the World=20 (14^th Edition, 2002), at www.ethnologue.com.

Hall, D. (1995). Assessing the needs of bilingual = pupils:=20 Living in two languages. London. David Fulton Publishers = Limited.

Mackey, W.F. (1968). The = description of=20 bilingualism. In J.A. Fishman (Ed.) Readings in the sociology of = language.=20 The Hague: Mouton, pp. 554-584.

Royal College of Speech and = Language Therapists=20 (1996). Communicating Quality2: = Professional Standards for Speech and Language Therapists London: = Royal=20 College of Speech and Language Therapists.

Winter, K. (1999). Speech and = language therapy=20 provision for bilingual children: aspects of the current service. International Journal of Language = and=20 Communication Disorders 34 (1),=20 85-98.

Young, T. =20 & Westernoff, F. (1999). Reflections of Speech and Language=20 Pathologists and Audiologists on Practices in a Multicultural, = Multilingual=20 Society. Journal of Speech- = Language=20 Pathology and Audiology, 23 (1),=20 24-30.

[1] This information is taken from the Neighbourhood Statistics = database.=20 Due to rounding of percentages, the addition of categories may not sum = to=20 exactly 100 % in all cases. Figures quoted should therefore only be used = for=20 guidance.

[2] Pashto (sometimes called Mahsudi, Pakhto, or Pushto) is a = language=20 mostly spoken in Pakistan, but also in Iran and the United Arab Emirates = (Ethnologue, 2002)

[3] Kokni = (also=20 called Kanari or Kokna) is an Indo-Aryan language spoken in Western=20 India.