Speech Sound Disorders Screening & Assessment PDF

SPEECH SOUND DISORDERS Screening and Assessment What is screening? ♦ is not comprehensive evaluation ♦ is a quick procedure ♦ provides an initial impression ♦ gives a pass/fail result ♦ does not permit a diagnosis ♦ identifies children ne...

SPEECH SOUND DISORDERS Screening and Assessment What is screening? ♦ is not comprehensive evaluation ♦ is a quick procedure ♦ provides an initial impression ♦ gives a pass/fail result ♦ does not permit a diagnosis ♦ identifies children needing further testing of speech ♦ identifies children needing referral to medical or other professional services ♦ is often used for large groups ♦ often done at sites such as baby clinics, crèches, preschools or primary schools ♦ can be done using standardised or non-standardised procedures ♦ the formal screening procedure often has norms and cut-off points ♦ the informal screening procedure is often devised by the clinician for a particular population or age level. ♦ A more in-depth and regular screening would be done on children where there is possible or identified risk. ♦ A rule of thumb is that if you have doubts, refer for further assessment. NOTE: Standardised screeners may be used on the population for whom they were normed, or used with caution for speakers of the same language and dialect e.g. British English screeners. Nonstandardised screeners can be used by applying clinical judgement and by finding out or knowing about normative expectations for that language and dialect (Bowen, 2007). These are then used with setting criteria and using these in comparison to the assessed child’s speech, using interviews with caregivers and via observing the child e.g. on the playground. What is an in-depth Speech Sound Disorders Assessment? ♦ Is comprehensive ♦ Is dynamic (Miccio, 2005) ♦ Consists of a battery of procedures ♦ Is a complete collection of information? ♦ Allows for a diagnosis ♦ Usually takes one to two hours (Williams, 2003) An in-depth SSD assessment includes the following areas:  Information gathering (case history)  Speech sound system assessment: o Word level (informal assessment procedure, standardised test or a combination of both) o Connected speech level Both speech samples (word level and connected speech level) need to recorded and analysed to determine:  phonetic inventory  processes operating and frequency of occurrence  speech intelligibility  variability  severity  Assessment of related or associated areas: o Oral peripheral examination (OPE) of the oro-facial mechanism o Diadochokinesis (DDK) o hearing mechanism o auditory discrimination o stimulability testing o (receptive and expressive language) Purposes of in-depth speech sound system assessment ♦ To determine if there is a speech problem ♦ to determine the client’s current level of and prognosis for future articulation and phonological development. ♦ to determine if the client’s problem is severe enough to warrant intervention” ♦ to provide information useful in planning treatment. ♦ To determine the nature and severity of the speech problem ♦ To differentiate between delayed and deviant SSS development (Williams, 2003) ♦ To determine the child’s strengths or capabilities ♦ To determine possible contributory or causal factors In the assessment of speech in bilingual children, the aim is to determine if the speech is within the normal range for that child’s language community (Goldstein & Fabiano, 2006). SPEECH SOUND SAMPLING Speech samples are collected to ascertain the nature and severity of the SSD. o Can be collected using elicitation techniques (picture and object naming, sentence completion, delayed imitation or modeling and imitation) or by spontaneous methods (conversation). o When collecting a speech sample, always audio and if possible video record. Recording of speech samples Suggestions for recording include: - Use high quality recorder. - Test equipment and electrical supply beforehand - Have batteries as a backup - The environment should be quiet, avoid noisy toys that may mask the client’s speech. - Try to conceal the recording device. - Write and record responses in writing while recording and verbally noting visible characteristics like dentalisation - If reduced intelligibility, unobtrusively repeat the child’s utterance e.g. Child: “Peep” Clinician: “Oh look, a sheep” Activities to elicit speech samples Consider the child’s age, developmental level and cognitive abilities. observe CG-child interaction gender and culturally appropriate toys, pictures and books. busy box, mirror, bubbles, picture books, telephone, dolls, cars. Games like: shopping, tea party, puppets picture sequence cards, play acting, explaining a game for connected speech and naming cards, aiming at and naming pictures, magic box Transcribe the sample. Use phonetic transcription. Practise your transcription! Transcribe the whole word. For school-aged children and residual problems, make notes above target word, or write just individual sounds. If there is a sound you are not sure of, transcribe according to the closest sounds, put a cross underneath, till you can work out what the sound is. Speech sampling at word level Criteria for an adequate speech sample at word level (Grunwell, 1985, p. 8) Size of sample should be: minimally 100 different words and preferably 200- 250 words  Type of sample should be spontaneous naming of objects and pictures and modeling avoided. Sample should be continuous, with focus being on errors and correct productions. Dialectal variations are noted. The sample must be representative of the sounds and combinations of the language of the client; the sample must be representative of that child’s SSS (e.g. all phonemes sampled in all positions in a wide range of CV structures). Each word should be repeated at least once to test for variability. Recording of sample should be audio and /or video recorded with detailed transcription and the use of ‘M’ for any modeled word. Word level samples have the following advantages: they are quicker to administer and analyse. you know what the child is trying to say. the target words and picture stimuli for the samples can easily be made locally relevant. Word level samples have the following disadvantages: they are not representative of how we speak naturally the exact target has to be elicited they can be problematic in a child with a poor vocabulary many do not allow for repetition thereby ignoring variability. Speech sampling at connected speech level Any word level sample must be accompanied by a connected speech sample Any utterances during assessment must be recorded, even if correct or unintelligible. The sample should be at least 200 words in length, taking 10-15 minutes to record. Connected speech samples are considered one of the most socially valid measures of speech and language (Flipsen, 2006; Kwiatkowski & Shriberg, 1992) and have long been used to assess both phonology and language. Connected speech samples have the ff. advantages: they are representative of how that child usually talks the effects of co-articulation can be noted the interaction between the speech and other problems e.g. syntax, may be noted some words will be said a number of times, so variability can be assessed Connected speech samples have the ff. disadvantages: they are time consuming to elicit and to analyse a child with severe speech or language problems may be unwilling to attempt sentences  the clinician may not know what a child is trying to say if intelligibility is poor, thereby making analysis of the sample difficult a child who is aware of their speech problem may avoid talking. Assessing the speech sounds of bilingual children using speech sampling: Take speech samples at word and conversation levels in both languages, remembering that it is unlikely that the child’s language development including phonology will be the same for both languages. Do an analysis separately for each language, determining all the parameters for each language e.g. consonant and cluster inventory, process operating, developmental level, intelligibility, etc... Do an analysis comparing speech in both languages. Determine the accuracy of consonants and vowels in each language, examine the accuracy of segments (consonants and vowels) which the languages share and for segments they do not share, expecting to see greater accuracy for the shared segments. Analyse the phonological processes occurring in each language, taking into consideration that the frequency and type of phonological patterns is different across different languages. For example, ST may occur in Zulu and English, but also affects clicks in Zulu as well as having many opportunities to occur in the prefixes of Zulu. On the other hand, CR typically occurs in English, but not in Zulu, and Nasal Feature Deletion (NFD) occurs in Zulu, but not in English. Do an error analysis, after working out which patterns are cross-linguistic effects and dialectal features (these are features or differences, NOT errors). Analyse the phonemes omitted, the inventory constraints and the substitutions. Apply what we know about speech development in those languages and about the typical pattern of speech development in bilinguals. Speech sample analysis Speech samples can be analysed in two ways: Independent analysis where the child’s speech system is looked at independently, and not in relation to the expected SSS or to developmental norms. That child’s SSS is described, especially the phonemes in the phonetic inventory and the range of syllable structures used. This method is often used when developmental norms are not available for a language. Relational analysis of the sample looks at the child’s SSS in relation or compared to developmental norms. Thus, accuracy and errors are looked at, including level of development of the phonetic inventory and phonological processes used. Speech parameters Phonetic inventory The phonetic inventory for consonants i.e. the sounds the child produces. This gives information about the child’s articulatory skills, independent and relational. For very young children assessment should include sound play and sound imitation. (Williams, 2003) Sounds are regarded as present if correctly produced on at least 50% of opportunities, and are circled. Marginal sounds were produced correctly but only on 1-49% of opportunities and are placed in brackets e.g. (s). Absent sounds or inventory constraints were given opportunity to be used, but were not correctly produced and are indicated by a cross over the sound. The cluster inventory is as above but applies to clusters if present in the language. The syllable structures and shapes the child is able to produce e.g. CV; CV, CV; CCCVCC; V, CV, CVCC. This gives us the phontactic or syllable structure inventory. Processes operating and frequency This measure looks at how frequently a process occurred. Other considerations are whether the processes are usual or unusual, age-appropriate or not, multiple process use and interaction of processes. How to decide if a process really is a process? sound change should have a minimum of four opportunities to occur process should be applied at least 20% of the time a class of sounds should be affected, so cannot be a residual error at least two members of that class are affected, but not in same word Calculation for process frequency of occurrence: Number of times the process occurred X 100 Number of times the process could have occurred Speech intelligibility Intelligibility or the degree to which the child’s speech can be understood, or the amount of the child’s speech that a listener can understand readily (Bowen, 2006). This is a subjective measure. This means it is based on the therapist’s discretion. It is measured at word and connected speech levels, using the speech samples. Bowen (2006) suggested a quick calculation for children in the 1 to four year range where the child’s age in years divided by four yields a percentage. This correlates to the percentage of words expected to be intelligible/understood in conversation with an unfamiliar adult. One year – 25%; two years – 50%; three years – 75% and four years 100%. The lower the intelligibility, the greater the need for therapy and a child of three years or older who is not intelligible is regarded as needing therapy. Calculation for speech intelligibility: No. of intelligible words X 100 No. of total words in the sample According to ASHA (n.d.); Bankson & Bernthal, ch 5 of Bernthal & Bankson, 2004, p. 234; Bowen (n.d.) 1, speech intelligibility can vary depending on a number of factors, including: the number, type, and frequency of speech sound errors (when present); the speaker's rate, inflection, stress patterns, pauses, voice quality, loudness, and fluency; linguistic factors (e.g., word choice and grammar); complexity of utterance (e.g., single words vs. conversational or connected speech); the listener's familiarity with the speaker's speech pattern; communication environment (e.g., familiar vs. unfamiliar communication partners, one-on-one vs. group conversation); communication cues for listener (e.g., nonverbal cues from the speaker, including gestures and facial expressions); and signal-to-noise ratio (i.e., amount of background noise). Severity Severity of the SSD 2 is measured in two ways: using percentage of consonants correct (PCC) (Shriberg & Kwiatkowski, 1982; see Lowe, 1994, pp. 123-139). The formula is the number of consonants produced correctly divided by the total number of consonants in the continuous sample, multiplied by 100. This is a positive measure, and the higher the score the less severe the problem. The lower the score, the more severe the problem and the greater the need for therapy. The PCC score can be used as a baseline from which to monitor progress and therapy effectiveness. It is interpreted with the following: 85-99 mild 65-85 mild-moderate 50-65 moderate-severe 40% suggests Inconsistent Speech Disorder. A particular therapy approach is required for children with the latter type of phonological disorder. Developmental level/appropriateness of SSS Compare the phonetic inventory and use of processes, as well as the interaction of these to that of normally developing children, and determining the age level for speech the child is functioning at (See Grunwell, 1985, PACS, Developmental assessment; Bankson & Bernthal, 2004, pp. 244-251)). 3 See Dodd (1995). Differential diagnosis and treatment of children with speech disorder (2nd ed.). London: Whurr 4 Holm, A., Crosbie, S. & Dodd, B. (2007). Differentiation normal variability from inconsistency in children’s speech: Normative data. International Journal of Language and Communication Disorders, 42, 4, pp. 467 - 486 ASSESSMENT OF RELATED OR ASSOCIATED AREAS Oral facial examination/ oral peripheral examination It is done to identify structural and/or functional factors, which may contribute to or cause a SSD, particularly an articulation disorder. Equipment needed includes a penlight torch, tongue depressor, cotton bud, mirror and gloves for infection control. Substances such as water or peanut butter may be used after explanation and obtaining the permission of the caregiver. Some children find this invasive or medical, so the procedure should be conducted professionally, with respect, and in a “fun” manner. Select which procedure you will use and determine if additional areas will be included e.g. for oromyofunctional disorders. Interpretation of findings is complex. Examine the face, jaw and teeth, lips, tongue, pharynx and hard and soft palates. Shipley & McAfee (2006) recommend looking for aspects such as the following: ♦ Colour of the pharynx, palate or tongue which is abnormal ♦ Abnormal width or height of the palatal arch ♦ Any asymmetry of the palate or face ♦ The uvula or tongue deviating to the side ♦ Tonsils which are enlarged ♦ Teeth which are missing ♦ Mouth breathing ♦ Poor maintenance of intraoral pressure ♦ Shortened lingual frenulum ♦ Absent or weak gag reflex ♦ Weakness of the jaw, tongue or lips. Tests available include: Oral Speech Mechanism Screening Examination – Third Edition (OSMSE-3) by St. Louis & Ruscello, 1987 Shipley & McAfee Testing of diadochokinetic rates or non speech oral repetition (Howell & Dean, 1994) (DDK) Done to assess a child’s ability to make rapid alternating movements for speech. This could affect articulatory programming and tactile and kinesthetic feedback. Children with phonological disorders may show some motor immaturity or difficulty. Children who have difficulties at the level of motor programming need to develop motor speech skills in therapy (Howell & Dean, 1994). It is important to give clear instructions and to model the responses desired, as well as to ensure that the child understands what is expected. Equipment needed is a stopwatch. Recording responses can be used for cross checking. Select which method and whose format and norms will be used beforehand. In addition, qualitative comments can be made about other aspects such as articulatory accuracy, sequencing and rhythm. Assess one, two and three point diadochokinesis 5. Two methods are available: see how many repetitions can be made in a predetermined time see how long it takes the child to produce a predetermined number of syllables. Shipley & McAfee, 2016, including Fletcher’s norms or for children under 6 years use the norms of Robbins & Klee (1987). Further information on diadochokinesis norms can be obtained from Williams and Stackhouse (2000). Stimulability testing Stimulability - that is if the child can produce an error correctly or can improve production or an error, especially a distortion when given a model, instruction, prompting or with manipulation of the articulators or the use of a context which may help improve production. For very young children stimulability testing should include sound imitation of clicks, raspberries, animal sounds, etc. (Williams, 2003) Look at the phonetic inventory and identify age-inappropriate inventory constraints. See if the child is stimulable for the absent sounds, beginning with as little support as possible. Use modeling the sound as a noise, modeling and imitation of the sound without highlighting, modeling and imitation with highlighting the sound, modeling and imitation with highlighting the particular features, instruction, facilitating context and finally production in a syllable or word with simple syllable structure. Look at the phonetic distribution and note age inappropriate positional constraints. These will need to be modeled and elicited at syllable or word level (using real and/or nonsense syllables and words). Use modeling and imitation of the sound in syllable or word without highlighting, modeling and imitation with highlighting the sound and facilitating context, Stimulability gives important information for selecting targets for therapy (stimulable sounds are thought to be good initial targets for therapy) and for the choice of therapy approach, and target sound and/or position of target sound. Shipley & McAfee, 2016 Auditory discrimination testing Auditory discrimination refers to the child’s ability to perceive differences between speech sounds presented auditorily. It is an aspect of speech perception. Lancaster and Pope (1989, pp. 67-69) note that “perception is essential to stable production … but that: since production may aid perception there is an interdependence between the two …”. Although general auditory discrimination testing may be done, research indicates that it is better to test specific aspects of auditory discrimination. 5 See Williams, P. & Stackhouse, J. (2000). Rate, accuracy and consistency: Diadochokinetic performance of young, normally developing children. Clinical Linguistics and Phonetics, 14, 4, pp. 267 – 293 for additional norms  Bear in mind that our testing may not be subtle enough (Howell & Dean, 1994) Test auditory discrimination by presenting word pairs containing the child’s error and target in minimal pairs where the child would say these as homophones. Such testing usually necessitates the clinician developing their own materials, applicable to a specific child’s dialect and error patterns, Testing could also be done at sentence level. The testing should be done in the position(s) in which the errors are made. Each pair should be tested more than once in random order, and each error or contrast should be assessed in more than one set of words. Modify the child’s response to their cognitive level e.g. picture pointing after explanation, “Is that right?”, “Are they the same? Yes/No”, same/different, or odd-one- out. Allow for variables like attention or memory. The nature of the task (same/different judgement or picture pointing) influences what you are looking at. Thus short term memory probably plays a role in the former, while the child needs to access the internal representation of the words to do the latter task (Howell & Dean, 1994). Ensure that the child is familiar with the words used and understands the instructions. Many of the published tests are thus not suitable for testing, and you will need to use informal procedures devised for that child and his pattern of speech difficulty. Examples of testing (ST) are: Identify target sound in nonsense or real word or phrase e.g. “Can you hear a snakey sound in bis, bus, I see mom”. Same/different e.g. “Are these the same or different? Sea sea; sea tea, etc. Odd one out e.g. “Which is the odd one out? Sea, tea, sea” Picture pointing e.g. “Show me she, sea, tea” Answering yes/no to pictures of minimal pairs e.g. “Is it a she?”; “Is it a sea?”; “Is it a tea?” Howell & Dean (1994) advise against routinely including auditory discrimination work in therapy for phonology problems. Auditory discrimination tests available include: Wepman CELF subtest CTOPP Morgan-Barry Test of Auditory Attention and Discrimination. These are all tests of discrimination for English. Screening for any hearing problems via otoscopic examination and pure tone screening Conduct an otoscopic examination or preferably a hearing screening. Need to determine this from the case history information. Need to consider referral for a full audiological evaluation. Estimation of receptive and expressive language abilities Receptive language can be informally assessed by observing the child’s responses to verbal instructions (see Shipley & McAfee, 2016) or by administration of a screener or test (Williams, 2003) e.g. Preschool Language Scale-3 (PLS-3) by Zimmerman (1992); Test of Early Language Development-3 (TELD-3) by Hresko, Reid & Hammill (1999) Note and record any language errors during speech assessment Use connected speech sample as language sample and work out size of lexicon, type/token ratio, Mean length of utterance (MLU) and semantic relations (Williams, 2003) HOW TO STRUCTURE YOUR ASSESSMENT SESSIONS In any assessment, you need to include: some time establishing rapport with the caregiver (often during the case history taking), explaining what and how you will be doing in the assessment (including why you are using play), some time establishing rapport with the client via short introductory type activities time explaining the results of your assessment ascertaining if the caregiver feels your assessment and findings were valid time explaining your therapy plan, , approach, and methods time explaining the time frame for therapy (expected length, and the University term system) time explaining the caregiver’s role in therapy (see PACT under therapy approaches), adapted to the circumstances of the caregiver and child the provision of home programmes mid and end of year. With regard to time, there are generally two ways of assessing. The first is to use a block of time (2x 45 minutes), assessment fully planned beforehand, and all analysis done once assessment is complete. You will be required to do this if assessing a Zulu- speaking child. The second system is to have a block of time, split into two sections (2x 45 minutes). In the first section, specific information is gathered, which is then analysed and used to form the basis for the planning of the second section or stage of the assessment. You will use this system for assessment of an English-speaking client in the second year clinic. Using these two ways of structuring SSSD assessment is done: - for student training purposes - so that you have exposure to both ways - to fit with the informal assessment procedures you will put together for your clinical module next year. Once out working, you will need to adapt to those circumstances. You may well not have two sessions available to you or may use a second session to do feedback to the caregiver. The one stage method of assessment: Zulu-speaking clients - Take/update case history information - assess speech using, and by taking speech samples at word and connected speech levels - assess associated and related areas of hearing, oral peripheral mechanism, oral motor abilities, stimulability and auditory discrimination skills - work out what aspects of auditory discrimination you need to assess and what inventory and positional constraints there are for you to assess stimulability after the first session; complete all analysis after session two. The two stage method of assessment: English-speaking clients Stage 1/Session 1: Case history taking/updating hearing screening (assessing receptive language) eliciting a conversational sample eliciting a brief word level speech sample doing an OPE.  Take the brief sample using the Metaphon Resource Pack (MRP), consisting of 44 words, focusing on those processes frequently found to occur in the speech of children with phonological disorders. These are VF, DP, labialisation, backing, ST of fricatives, ST of affricates, PV, FD, GL & LIQ REP, ICD incl. GR, FCD including GR, CR word initially & word finally, including deleting the entire cluster.  Analyse sample such that you can plan session 2  Analyse by process recording your analysis on the MRP and SLP forms.  Calculate process frequency of occurrence on the form and note any particular patterns of process use e.g. two interacting processes or processes specific to particular positions of the word. Note the effects of these processes on intelligibility.  Determine the phonetic inventory:  first independently i.e. only looking at what the child said  then dependently, i.e. in relation to the target or adult inventory.  List absent phones, marginal phones, present phones and those for which no opportunity was provided.  List the range of syllable structures used, first independently i.e. only looking at what the child said and then dependently, i.e. in relation to the target or adult syllable structures. References: Howell and Dean, 1994; Masterson, 1993.  Variability needs to be factored in by eliciting some of the words twice. Calculate variability figures. Stage 2/Session 2: completes assessment and looks at areas of difficulty in more depth, including: probing age inappropriate processes with a frequency of occurrence of ≥40%, establishing a baseline to measure progress, assessing auditory discrimination for errors and targets should be routinely done (Howell & Dean, 1994) assessing stimulability for inventory constraints/absent phonemes and positional constraints assessing oral motor abilities/characteristics including oral sensory and non speech oral repetition or diadochokinesis (DDK) abilities (Howell & Dean, 1994, Masterson, 1993). NOW WRITE THE ASSESSMENT REPORT