Assessing the speech and language skills of children is obviously an integral part of our work as speech-language pathologists. However, with an estimated 5.2 million bilingual children enrolled in schools in the United States 1), the challenge of assessing children whose first language is not English is compounded by the lack of adequate knowledge of speech and language development of bilingual children living in the United States, the lack of appropriately developed and normed assessment tools and tests, the lack of personnel who speak the language of the child to be assessed, etc. The purpose of this website is to give the speech-language pathologist general principles and information of the steps to follow when performing assessment with children whose language is not English.

According to Kohnert2), a bilingual assessment (as in any speech-language assessment) is performed to identify potential impairment, to describe the individual’s communicative system, to plan a course of action, predict long-term outcomes of the plan and evaluate the effects of the implemented plan of action over time. The earlier a child is identified, the earlier effective therapy can be provided and better outcomes can be achieved. The bilingual child uses two language systems that need to be assessed to appropriately identify if the child truly has a disorder, or the language difficulties are due to the normal process of second language learning.

When the speech pathologist receives a referral (usually from the class teacher) to evaluate a bilingual child, the referral reasons usually deal with academics, language comprehension and language expression. In a study by Kayser,3) over 70% of referrals of bilingual children by teachers dealt with academics and comprehension and approximately 22% with expressive language. The referrals usually mentioned that the child was “not keeping up with the rest,” “does not follow directions,” “unintelligible,” “has trouble with English,” or “the speech unclear.” In the same study Kayser4) also reported that referrals from kindergarten or first grade usually related to speech, but when they came from second grade or above, the concerns were usually academics or comprehension in the classroom. Also, she found common characteristics among Mexican American children being referred due to language impairment. Some of the characteristics were: low socio-economic level, Spanish-speaking parents, poor conversational skills, English-only speaking teachers and classrooms. Click on this link for a complete list of demographic characteristics of Mexican American children. The clinician should always keep in mind these characteristics of referrals and make sure to seek additional information from teachers, parents and other professionals involved with the student before proceeding with the assessment.

Roseberry-McKibbin 5)suggests that there are characteristics that might signal language learning disabilities in students such as, non-verbal aspects of language are not culturally appropriate, the student fails to express basic needs or rarely initiates verbal interactions with peers. Other signs might be that the student uses gestures when vocalizations are expected and his/her peers have difficulty understanding responses or are confused because the student does not give enough information or the information given is disorganized. Follow this link for a complete list of characteristics of students with possible language learning disability.

The language skills of bilingual children are similar, but different to the ones of monolingual children 6). These language skills are not equally distributed across both languages. Bilingual children may show certain skills in one language but not in the other 7). Therefore, comprehensive assessment of bilingual children has to be performed in both languages. A bilingual child is not two monolinguals in one. The rate of acquisition of language milestones is similar to monolingual children provided that both languages are taken into account. The rate of acquisition of the L1 and L2 varies among different learners. Some considerations8) to take into account are: length of exposure to each language, time of exposure to L2, reasons for learning the L2, academic experience with L1 and L2, ability to use each language, linguistic structure of the two languages, and individual variation. No two children are equal. Any bilingual child might have relatively equal facility with both languages or the language skills in one language might be superior than in the other.

For more information on bilingualism and bilingual speech and language development see: bilingualism.

This is the question at the core of a bilingual assessment. The speech-language pathologist has to determine from the general description of the referral, if the child really has a language impairment or not. According to Kohnert9), the following characteristics should be present in order for a child to be identified as speech or language impaired:

* There should be difficulty in both languages. The language impairment manifests itself in both languages because the nature of the difficulty is due to a problem with the processing of language input, which would affect both languages. However, the impairments might look different on the surface because each language has different features and there are different demands for L1 and L2. For example, the nature and frequency of speech errors may be different according to the different sounds and sound patterns of each language or dialect 10).

* Late achievement of milestones. Typically developing bilinguals achieve developmental milestones at approximately the same rate as monolingual peers, but the skills might be distributed across both languages. A comprehensive and detailed case history would enable the clinician to determine how the child's development has progressed over the years.

* Bilingual peers at different level. The clinician during the assessment might compare the child to other bilingual peers of the same age or siblings who are considered to be typically developing. If there are consistent differences between the performance of same age level peers, a disorder might be suspected.

* Parent report. During the interview with parents, the clinician needs to inquire about the parents' views on the child's speech language proficiency and perceptions about the language difficulties he or she is experiencing. In a study conducted by Restrepo 11), there was a high correlation between the parents' report of language input and performance to testing results done by the researchers. Therefore, the information provided by parents is invaluable in considering the child's proficiency in their first language and how they view their child's language.

According to Goldstein,12)a complete and comprehensive assessment of a bilingual child needs to be performed in both languages and the clinician needs to consider socio-linguistic variables when examining the interaction of the child's skills in both languages. The purpose of this section is to give the clinician suggestions of steps to follow and considerations for the assessment. This section is subdivided in: considerations before the assessment, assessment steps, assessment tools and a special section for articulation and phonological assessment of bilingual children.

According to Kohnert,13)the clinician who needs to do an assessment with a bilingual child needs to always keep in mind that the task of assessing this child is more complex because it is not only that the clinician is dealing with a child with two languages, but also the way the assessment is conducted will affect the results. The lack of adequate testing materials and other resources makes the task more difficult. Therefore, the clinician needs to keep in mind certain principles that will guide him or her in performing a valid assessment with a bilingual child.14)

Identify and Reduce Sources of Bias
Bias in assessment will result in a failure to identify a child who has a disorder or identify a child with a disorder who is actually typically developing, or developing a flawed plan of action for therapy. The information collected would not reflect a child’s real areas of strengths and needs.

Types of bias:

• Content bias—-the child’s knowledge is measured with tools and methods unfamiliar to his or her experience.
  
• Linguistic bias–the language or dialect used in the assessment is inconsistent with the child’s experience.
  
• Data interpretation bias–measurement standards are inconsistent with child’s experience.
  
• Cultural bias–There is an unrecognized mismatch between the culture of the child and his or her family and the assessment team.
  

How to prevent bias:

• Review literature of bilingual speech and language development.
  
• Review features of L1 and L2.
  
• Take into account child’s age.
  
• Take into account child’s cultural and language acquisition history.
  
• Identify peers with similar experiences (siblings, classmates).
  
• Consider the expectations from the child's family, community, culture and the academic team.
  
• Compare the child’s language change across time.
  
• Observe the child’s needs across a range of settings to assess needs.
  
• Understand the cultural context in which the child’s languages develop and the assessment and team members’ cultural context.
  

Individualize Assessment Timing
One of the most common mistakes that specialists make in regards to bilingual children is to set rigid time standards for the child to reach certain language and developmental milestones. For example, the team might decide to wait until a bilingual child has had 2 years of experience with English before assessing using the commonly known BICS/CALPS guidelines of second language acquisition. Although these guidelines are useful in the sense that they emphasize that second language learning is a process that takes time, every child’s experience with language acquisition is different. The child's rate of L1 and L2 development is going to be affected by the environment, the uses and demands for each language, sociolinguistic factors and individual differences. It is, therefore, important to watch carefully how decisions are made regarding the best timing for an assessment.

Consider L1 and L2 Abilities and Needs
Bilingual children need both languages to communicate in their different family, social and academic environments. Their bilingualism is not a choice. The speech-language needs to look at the child’s abilities in both languages and its past, present and future needs. Inventories and questionnaires of language use can be done to evaluate present and future needs. Parents, family members, peers, teachers and other people who interact with the child would be important sources of information. The goal is to consider the child as a whole. The clinician needs to assess the child's total communication system.

Look Beyond the Obvious
This principle again underscores the importance of considering both languages to assess the abilities and needs of the bilingual child. As mentioned before, a bilingual child's language abilities are distributed across the two languages. Assessing only one language (usually, the majority language), could lead to inappropriate identification or misidentification of disorder. Also, according to Kayser (1995), clinicians tend to decide to assess in the majority language assuming that because the child speaks some English, then it is appropriate to test in that language, perhaps ignoring that L1 might be the dominant language of the child.

Gather Data Using Multiple Measures at Different Points in Time
This principle stresses the importance of taking in consideration various sources of data to make decisions regarding a bilingual child. The goal is to evaluate the child's performance at different points in time. These considerations are:15)

• Reviewing developmental and educational reports;
  
• Interviewing the family, child, cultural representatives, classroom and special education teachers to understand the cultural and educational environments and how language is used in each;
  
• Observing the child at home, in the classroom, in the playground and in the clinic room which will provide information on the communicative functioning of the child;
  
• Testing to gather data of the child’s receptive and expressive skills using picture identification, following instructions, listening to stories and answering questions, telling stories, etc.
  

Before deciding to conduct an assessment, it is important to verify that the referral of the child for speech-language services is an appropriate one.

According to Olson,16) in order to reduce the inappropriate referrals to special education of minority children, the professional needs to identify the child's problem, the source of the problems and the steps to resolve the difficulties first within the classroom setting. Kayser17) describes a pre-referral process based on modifying the curriculum and teaching strategies to help students learn. These steps relate to the characteristics of the curriculum and the child's cultural experience to access it, the efforts made to identify the source of the problems and any alternative solutions that have been tried. Click here for a complete list and explanation of the Eight-Step Prereferral Process.

• Research and learn about the child’s cultural, social and linguistic community.
  
• Research and learn about the child’s language characteristics and dialectal differences.
  
• Research and learn about the child's cultural and social conventions. So, when meeting the child's parents you are knowledgeable of appropriate greetings, gestures and other basic formal conventions.
  
• Review case history if available.
  
• Determine if you will need an interpreter for the assessment. Get together with the interpreter before the assessment to review your plan. Follow this link for more on working with interpreters.
  
• Select appropriate assessment tools you are going to use.
  
• Make a plan.
  

During the assessment process the clinician can follow the steps outlined below:18)

• Step 1: Perform a Detailed Case history
  

This step involves not only gathering previous medical and educational information, but also interviewing with the parents and other family members, teachers and other important persons in the child’s life. This step is particularly important when assessing a bilingual child because the information gathered during this step will be crucial in deciding if the child has a speech-language disorder or not.

Of particular interest to the speech-language pathologist is to ask the parents and family members involved with the child information about:

• The child’s acquisition of L1 and L2.
  
• Language(s) spoken at home.
  
• Time and length of exposure to L2.
  
• Amount of input and output in both languages.
  
• Country of birth (immigrant vs. native).
  
• Urban vs. rural background.
  
• Generational membership.
  
• Length of residence in the United States and degree of acculturation.
  
• Family structure and child rearing practices.
  
• Play routines.
  
• Sociolinguistic factors such as, turn-taking, silence and style of questioning.
  

• Step 2: Perform additional routine assessments
  
• Oral-peripheral exam.
  
• Hearing screening to rule out a hearing impairment.
  

• Step 3: Administer assessment tools in both languages
  

The clinician has several options. According to Kohnert,19) in bilingual assessments the clinician cannot rely only on standardized assessments. The clinician will need to use a wide range of tools in order to be able to fully assess the child's bilingual skills. Kohnert20) classifies the assessment tools between product measures and process measures. Both types of measures have their purposes and need to be part of a comprehensive bilingual assessment.

• Product Measures: Measure the performance of the child in a specific language. There are two types of product measures: language samples and standardized tests.
  
• Process Measures: These tools attempt to reduce biases and limitations typical of the product-based measures. Examples of process measures are language learning measures, criterion-based measures, dynamic assessment and portfolio assessment.
  

Following is a discussion of product and process assessment tools:
\\

• Language samples: If you are using language samples, a sample needs to be taken for each language. They can be either spontaneous or elicited. They can be obtained during structured or unstructured interactions with parents, siblings, peers, teachers, etc. For each sample, you need to assess and analyze grammatical complexity, vocabulary and pragmatics. If you do not speak the L1 of the child, the interpreter or other support person from the child's community working with you needs to help in analyzing the sample. Also, analyze if proficiency in the language being used responds to the demands of the communicative situation. According to Gutierrez-Clellen et. al. (2000), language samples need to be described using measures of morphosyntax complexity, utterance length in words and a detailed analysis of the quality of the child's utterances, such as number and type of grammatical errors. Children might exhibit strengths in certain areas and not in others. The size of the language sample and its complexity will vary according to the situation, the topic and the familiarity with the interaction partner.
  

Follow this link for more suggestions for language sample elicitation.21)

• Standardized formal tests. If there are tests available in the child’s first language, and the clinician decides to use them, there are several considerations to keep in mind when using these tests. It is important to be careful how the results are used. Following are some precautions:
  
• Translated tests:

Many tests are translations of their English versions. According to Stubbe-Kester and Pena,22) this invalidates any results from the test because the hierarchy of difficulty of test items or complexity of language forms would be different in English and other languages. Thus, the test would not reflect the developmental order of items of the target language.

• Test Normative Data: Before using a test, review its normative data and make sure the test was normed with a population with the same linguistic and cultural characteristics of the child you are going to assess.23) The clinician needs to look in the test manual to see if the language proficiency of the children in the normative group ranged from balanced in both languages to greater proficiency in one language over the other.24)
  
• Test Sensitivity and Specificity: According to Plante & Vance,25) the recommended minimal acceptable criterion is 80% for specificity and sensitivity in a test when it is evaluated with a specific group of children similar to the child being assessed.
  

Please follow the link for suggestions when using standardized tests.26)

• Language Learning Measures or Limited Training Tasks:27) In these measures, the child is given new information (invented words or grammar rule) through modeling and imitation in structured contexts. After a familiarization and practice period, the child is tested to determine how successful they are. The goal is to measures the amount of gain before and after the test.
  

• Criterion-Based Measures: In contrast with norm-referenced standardized testing, criterion-referenced tests measure the student's mastery of specific objectives defined by predetermined standards. It compares a child's performance on a specific skill, a grammatical structure or a linguistic concept. This allows for consideration of the social context where communication and how the language is used in a culture. The clinician can design the measure,thus adapting it to the specific student in a specific context and can be linked to the class curriculum.28)
  

• Dynamic Assessment (DA): This type of assessment is based on Vigotzky's concept of Zone of Proximal Development (ZPD). The child brings to the learning situation, his or her interests, motivation and knowledge. The adult is responsible for maintaining the child's interest. In assessment, DA helps the clinician determine how a child learns, identify the child’s potential for change when given guided support by the examiner, induce self-regulated learning and inform intervention. Potential for change is measured by 3 factors: child responsiveness, examainer's effort and transfer.
  

Follow this link to see the components for Dynamic Assessment29)

• Portfolio Assessment: According to Kayser,30)portfolio assessment is a collection of the student's work samples. These samples more accurately reflect the child's growth, achievements and efforts in meaningful tasks relevant to the curriculum and activities in the classroom. The samples could be selected from reading, writing, listening and speaking tasks. It provides an opportunity to see what the child CAN do, instead of what the child cannot do typical of product measures. The teacher and the child establish the evaluation parameters; thus, they become active participants in the process.
  

Follow this link for advantages of portfolio assessment.31)

• Step 4: Additional Measures for Articulatory and Phonological Disorders Assessment:
  

According to Goldstein & Gildersleeve-Neumann,32)the clinician will need to examine the child's skills “broadly” (e.g. segmental accuracy, phonological patterns) and “deeply” (consonant and vowel accuracy, sound errors, etc.).

According to Fabiano,33) the following are the measures needed when assessing articulation and phonological disorders in bilingual children:

• Assess in both languages: single word and connected speech (conversation or narrative) samples should be obtained in both languages. Phonological acquisition is not parallel. Each language has different trajectories.
  

• Perform an independent analysis. Phonetic inventory in both languages using single-word and connected speech samples and organize inventory.
  

• Perform a relational analysis. Overall consonant and vowel accuracy in each language and accuracy of shared elements. There is higher accuracy on shared phonological elements as compared with unshared. This aids in goal development.
  

• Perform an error analysis. Bilingual children sometimes use a language-specific phonological element in the production of the other language. Ex. Using the Spanish trill in English. This should not be counted as an error.
  

• Perform a phonological pattern analysis. Patterns vary across language. It is important to determine typical and atypical patterns for the language and age of the child. Ex. English allows 3-member onset clusters and Spanish allows only two. If this error is present in English, it is developmental, but if present in Spanish it might be considered “delayed.”
  

• Diagnosis. If a child has a fundamental language-learning problem, delay or disorder, it will be present in both languages. If only one language is assessed it is impossible to make a differential diagnosis between phonological disorder or low language proficiency.
  

After administering the different assessment tools and the results are analyzed. the clinician will be able to determine if the child is typically developing and needs other types of academic support or if the child has a speech or language disorder. Then, the process of planning for intervention begins. The results from the assessment, especially the dynamic assessment results will help determining goals for intervention and the most appropriate techniques for therapy. For more information on bilingual intervention, go to: Intervention.

Websites

Cross-cultural language and academic development
http://www.edtrust.org/sites/edtrust.org/files/publications/files/imprvgeduLatino_Espanol.pdf

“Colorin Colorado” A GREAT bilingual site for educators and families
http://www.colorincolorado.org/

Bilingual therapies
http://www.bilingualtherapies.com/

Technical reports for professionals and parents
http://www.nccrest.org/about.html - series of brief and technical reports for professionals and parents.

Bilingual speech and language resources
http://www.acadcom.com/acanews1/anmviewer.asp?a=14&z=6

Center for Positive Practices
http://www.positivepractices.com/BilingualEducation/BilingualEdResearchReport.html

Como mejorar la educación de los niños
http://www2.edtrust.org/NR/rdonlyres/CCF44357-B41F-4483-BB57-D2F3D25565B9/0/imprvgeduLatino_Espanol.pdf

La lectura es lo primero
http://www.nifl.gov/partnershipforreading/publications/pdf/PRF_espan_.pdf

CLASE: Center Latino Achievement and Success in Education
http://www.coe.uga.edu/clase/


Readings

Genesee, F., Paradis, J., Crago, M. (2004). Dual language development and disorders. A handbook on bilingualism & second language learning. Baltimore: Brookes Publishing.

Goldstein, B. (2004). Bilingual language development and disorders in Spanish-English speakers. Baltimore: Paul H Brookes Publishing.

Goldstein, B. (2000). Cultural and linguistic diversity resource guide for speech-language pathologists. San Diego, CA: Singular Publishing.

Kayser, H. (1998). Assessment and intervention resource for Hispanic children. San Diego: Singular Publishing.

Kayser, H. (1995). Bilingual speech-language pathology: an Hispanic focus. San Diego, CA: Plural Publishing.

Kohnert, K. (2008). Language disorders in bilingual children and adults. San Diego, CA: Plural Publishing.

Student Contributor for this page: Alexandra Guerra-Sundberg, Winter 2008