The use of dynamic assessment to evaluate narrative language learning in children with hearing loss: Three case studies

Abstract

The primary objectives of this research were to establish whether dynamic assessment could be implemented in children with hearing loss with a range of language abilities and to obtain pilot data to support the use of dynamic assessment for determining narrative language learning difficulties in children with hearing loss. Participants were three children aged seven to 12 years with moderate to profound hearing loss. The Test of Narrative Language (TNL) was used to measure narrative skills before and after two Mediated Learning Experience (MLE) sessions. Responsiveness to mediation in MLE sessions was compared to ability to learn based on a comparison of pre- versus post-test TNL scores. Two participants showed substantial improvements in their Oral Narrative subtest scores, and showed good responsiveness and modifiability in MLE sessions. One participant showed no change in TNL results, and low responsiveness and low modifiability in MLE sessions. Children with greater response to mediation had more improved TNL scores. Differential diagnosis of language learning difficulties in children with hearing loss may be possible using ‘dynamic assessment’ (DA).

Keywords

Hearing impairment language disorder narrative

I Introduction

1 The significance of narrative language skills in children

Narrative is a widely used indicator of language development and disorders in children (Botting, 2002). Narrative ability develops through childhood (Applebee, 1978) and represents the coming together of a number of linguistic skills to produce a coherent text, combining several ideas and sentences (Hughes et al., 1997). Narrative ability is related to success in school as it correlates with the acquisition of literacy; for example, Crais and Lorch’s (1994) review concluded that narrative ability predicted academic performance for children who were typically developing and those with language disorder, and Griffin et al.’s (2004) longitudinal study of 32 children found a correlation between early oral narrative in kindergarten and reading comprehension at the age of eight. Narrative has also been found to be a useful method for uncovering language difficulties. Crais and Lorch (1994) concluded that academic performance for those with language disorder is predicted by narrative skills. This is supported by Boudreau (2008), who looked at major findings in studies investigating the narrative abilities of children with and without language impairment and concluded that ‘narrative difficulties are a persistent characteristic of language impairment’ (2008: 101).

2 Narrative language skills in children with hearing loss

Children with hearing loss typically experience a range of language difficulties, including delays in semantic, syntactical, phonological, morphological, and pragmatic aspects of language (Remine et al., 2003; Schönweiler et al., 1998). Most studies of children with hearing loss have concentrated on phonology, syntax, and vocabulary development (e.g. Briscoe et al., 2001; Curtiss et al., 1979; Pressnell, 1973); only a few studies have examined their narrative skills. Crosson and Geers (2001) investigated the narrative abilities of 87 eight and nine year olds with hearing loss (four years post cochlear implantation) and 28 eight and nine year olds with normal hearing using an eight picture sequence story. The narrative stories were analysed using narrative structure based on high-point analysis, conjunctions, and referents that assisted in recognizing and differentiating characters in the narrative. Children with normal hearing produced cohesive stories that used conjunctions and included a high point, a resolution, and evaluation statements. Children receiving cochlear implants before age five whose speech perception scores were above average produced narratives that were similar in structure to those of children with normal hearing but they used fewer orientation and more evaluation statements. Young et al. (1997) studied four boys with unilateral hearing loss (UHL) aged 7.2–10.7. Narrative abilities of children with UHL were compared to findings in previous studies of children with normal hearing and language disorder. The children with UHL did not reach what Young et al. refer as a ‘true narrative’ stage, which is typically acquired by age 5–6 years in children with normal hearing. They used more episodes and lacked goals in their narratives, similar to children with language disorder. Similarly, Yoshinaga-Itano and Snyder (1984) studied 49 children with severe to profound hearing loss and 49 hearing controls aged 10–15 years and showed that more than 50% of the children with severe to profound hearing loss produced poor narratives compared to children with normal hearing, based on story grammar analysis.

Children with no hearing difficulties typically acquire stories as they can hear personal and other types of narratives in different social settings (Crosson and Geers, 2001). On the other hand, children with profound hearing loss may not benefit from the incidental learning from hearing stories in their environment, and hence for these children a greater emphasis on teaching story-telling in therapy and in the school setting is recommended (Swanwick and Watson, 2005). It might be overly simplistic, however, to say that the only reason children with hearing loss have a problem with learning stories is simply because they cannot hear; further exploration of their ability to learn narrative skills is therefore warranted.

The study of deaf children who use sign language to communicate may provide some answers to the question of whether children with hearing loss have language difficulties that are not simply related to the degree of hearing impairment. Morgan et al. (2006) reported a single case study of a deaf child with signing parents who had significant delay in both the production and comprehension of certain sign language grammatical constructions, consistent with specific language impairment. Quinto-Pozos et al. (2011) conducted a qualitative study involving four focus groups with language professionals and general educators and one-on-one interviews to investigate the presence of ‘developmental communication disorder’ in children who are native users of American Sign Language (children exposed to sign language from birth). Participants reported that ‘native-signing deaf children’ can have atypical language development. They also commented on other possible causes of language disorder in children with hearing loss, including: (1) parental input, (2) co-morbid attention deficits, and (3) social and emotional problems. Interestingly, Quinto-Pozos et al. found that the incidence of language disorder in children who are native signers appears to be lower than published prevalence estimates for specific language impairment (Law et al., 2000; Tomblin et al., 1997); however, larger scale studies would be needed to confirm this.

3 Narrative and narrative analysis

Personal and fictional narratives are different types of narrative that are useful for assessing narrative skills (Hughes et al., 1997). Personal narrative is a recount of factual past events, while fictional narrative is creating or recalling a story from what was heard or read before (McCabe et al., 2008). The first analysis of narrative to become influential in the field of modern linguistics was that of Labov and Waletzky (1967) who produced ‘high-point analysis’, whereby they saw the point of a story as being the moment of climax (the ‘high point’), and the structure of the story as being about what was needed to lead up to that point, and back down from it. In high-point analysis, narrative clauses are ordered in a temporal sequence, and there is an overall structure of narratives that serves to organize narrative around high points that occur in the following order:

The ‘abstract’ occurs at the beginning and summarizes the whole story so the speaker gets the listener’s attention.

‘Orientation’ provides information about characters, time, and place, which gives an overall view about what will happen next.

‘Complication’ is considered the main ingredient of narrative clauses as this contains a series of events with a complication.

‘Evaluation’ forms give emotional comments in order to make the high point clear.

The ‘resolution’ represents how the story complications were resolved.

‘Coda’ clauses close the story and bridge the end of the story to the present moment.

High-point analysis has been used by researchers as the coding scheme to describe developmental changes in both personal and fictional narratives (Crosson and Geers, 2001).

Peterson and McCabe (1983) used high-point analysis with children aged four to nine years in order to categorize the structural patterns of their personal narratives. They categorized the structural patterns of the children’s narratives as follows: leap frog, end at high point, classic, chronological, impoverished, disoriented, and miscellaneous. McCabe and Rollins (1994) added another pattern – two-events narrative – based on their study of the narrative development of typically developing preschool children; they also suggested methods for assessing narrative for children with language impairment. This combines Labov and Waletzky’s ideas of (adult) narrative structure with those of the development of narrative in children.

II Diagnosis of language impairment

Children are generally diagnosed as language disordered based on results of norm-referenced tests, whereby their scores are compared to an age-matched normative sample. Although best practice suggests that children are diagnosed with language impairment based on standardized tests and language sample analysis which establishes baseline data, results of norm referenced tests may not represent the child’s ability to respond to language learning experiences in children whose language exposure has been limited by hearing loss. Standardized assessments may also fail to provide speech language pathologists (SLPs) with information about how a child approaches a task and the difficulties the child encounters, and thus can have limited value for planning intervention. Furthermore, standardized tests do not enable the examiner to differentiate between children with normal language learning ability but limited experience who are expected to learn new language rules with adequate targeted tasks from those have difficulty learning a new language rule due to language learning difficulties.

There is a need for assessment tools that help the SLP to know more about the difficulties that the child encounters while doing a specific task in order to help build an appropriate intervention plan. One approach that has been used to achieve this is ‘dynamic assessment’ (DA), developed as an educational assessment tool by Feuerstein et al. (1981; see also Vygotsky, 1978).

Dynamic assessment is designed to assess not only what a child’s skills are at a point in time (which most assessments do) but also how well they learn. This means the assessment involves teaching them an appropriate target using a specific teaching method (mediated learning), and assessing their learning, both in terms of what they learn and the strategies they show in the process (Gutiérrez-Clellen and Peña, 2001; Gutiérrez-Clellen et al., 1998; Lidz, 2003). The DA model includes a test–teach–retest format in which the assessor manipulates specific materials, activities and training procedures in order to modify the child’s performance during the assessment procedure (Gutiérrez-Clellen and Peña, 2001; Gutiérrez-Clellen et al., 1998). Interest in DA first emerged based on the ‘limitations of static tests’ (Hasson and Joffe, 2007: 10) that neither show the challenges that children may have while doing a task nor do they anticipate language change as they do not inspect children’s ‘responsiveness to intervention’ (Dockrell, 2001; Gutiérrez-Clellen et al., 1998; Hasson and Joffe, 2007). DA also has the capacity to differentiate between children who have not yet learned something (e.g. due to limited exposure) from those who, presenting with the same language level, show real difficulty in learning. Two children could perform similarly on a typical language test, but perform very differently in a DA of language. DA may be particularly useful for differentiating between children with a language disorder from those with culturally and linguistically diverse backgrounds (CLD) whose home and school language experiences differ (Gutiérrez-Clellen and Peña, 2001; Gutiérrez-Clellen and Quinn, 1993; Laing and Kamhi, 2003; Peña et al., 2001). Children with hearing loss may have language impairment due to late identification and treatment of their hearing loss or they may have additional language learning problems (Yoshinaga-Itano et al., 1998). DA may have the potential to differentiate between children with hearing loss who do not have a difficulty with language learning, from those who do.

Even though DA was developed by Feuerstein (1979), the concept is based on Vygotsky’s (1978) views of learning, whereby children are seen to learn through interaction with more knowledgeable adults (Gutiérrez-Clellen and Peña, 2001; Lidz, 1991). Vygotsky (1978) highlighted that learning occurs in the ‘zone of proximal development’ (ZPD), in which the child’s most effective learning would be in an area close to that which he or she is capable of achieving independently. Feuerstein (1979) developed the concept of a Mediated Learning Experience (MLE). Mediated learning refers to the interaction between mediator and child which allows the child to reach a higher level of their learning potential. Peña et al. (2001) described four of the critical components for MLE, according to Feuerstein (1979), as (1) intentionality, (2) meaning, (3) transcendence, and (4) competence (see Table 1).

Table 1.

Four of the critical components for MLE from Feuerstein (1979), as described by Peña et al. (2001).

Intentionality	Mediator’s goal to introduce a statement about the goals of the sessions; the mediator organizes the session with the aim of teaching and helping the child to know the purpose of the interaction.
Meaning	Explaining to the child why the task is important so that the child’s attention is drawn to the critical features of the task.
Transcendence	Help the child to link the instant task to events in the child’s life experience, so the mediator asks the child ‘What if?’ during activities in order to encourage the child to reach independent, hypothetical thinking and learning.
Competence	To manoeuvre a task in order to help the child to plan, and to fulfil the aimed strategies for approaching a task.

The DA approach determines the child’s modifiability (ability to learn). Modifiability is an important aspect of the practice of DA, in order to observe the child’s ability to change during MLE sessions (Peña, 2000). Peña (2000) found out that modifiability differentiates between typically developing children and low language ability from culturally diverse background, and it is a valuable method in offering a less biased assessment. Modifiability has three components: (1) ability to transfer, (2) responsiveness to mediation, and (3) examiner effort and intensity. Gutiérrez-Clellen and colleagues (1998) developed a Modifiability Scale that was adapted from Lidz (1991) and Peña (1993).

It appears that there are only two published studies of DA in children with hearing loss, but neither study investigated narrative. Tzuriel and Caspi (1992) used DA with 26 deaf preschool children and 26 control group children with normal hearing to evaluate cognitive modifiability using the Children’s Analogical Thinking Modifiability Scale (CATM). Children with hearing loss scored better on the post-test than was expected based on standardized tests, suggesting that the standardized test was underestimating their abilities. Lidz (2004) also used DA to evaluate cognitive abilities in children who were deaf and found that behaviour rated during the mediation phase was better than in the static pre-test phase.

The aims of the study were to investigate whether DA is a useful alternative approach to traditional assessment of fictional narrative skills in children with moderate-severe to profound bilateral hearing loss, and to investigate whether a DA procedure can differentiate between normal versus poor language learning ability in children with hearing loss.

III Methodology

1 Participants

A case-study series design was used to examine DA of narrative skills in three children (see Table 2). All participants were diagnosed with bilateral hearing loss when they were pre-lingual and ranged in age from one to two years of age. None of the children had congenital or neurological impairments other than deafness, and all had normal cognition. Sylvia (all names are pseudonyms) received Auditory Verbal Therapy (AVT) for approximately four years. Brendan had a cochlear implant fitted at two years but did not wear it consistently until the age of five. He had been in an AVT intervention program for two years and was just learning to speak. Because of Brendan’s reliance on sign language, his usual therapist attended all his sessions to facilitate understanding of the instructions. Brendan did not use any signs during the sessions except when he communicated with his mother. Clara received her cochlear implant one year prior to the study and had received no post-implantation speech or language therapy. She had teacher aide assistance at school periodically over the previous five years.

Table 2.

Characteristics of the three participants.

Child (pseudonyms are used)	Age	Type of hearing loss	Hearing instruments	Age of receiving hearing aid or cochlear implant	School year and setting	Type of communication
Sylvia	7;4	Moderate to severe bilateral	Bilateral Hearing aids,	Hearing aids fitted at age 2;3	Year 3, mainstream classroom	Spoken English
Brendan	7;8	Severe to profound bilateral	Unilateral cochlear implant	Hearing aids fitted at age 19 months; cochlear implant at 2 years	Year 3, Signing unit in mainstream school, integration for sports only	New Zealand Sign Language and Spoken English
Clara	12;3	Severe to profound bilateral	Unilateral cochlear implant	Hearing aids fitted at age 2;3; cochlear implant at 11 years	Year 8, mainstream classroom	Spoken English and lip-reading

Participants were recruited from the University of Auckland Listening and Language Clinic after the study received ethics approval from the University of Auckland Human Participants Ethics Committee. Participants were contacted by the clinic SLP and invited to participate in the study, and were not known to the principal researcher prior to agreeing to consider participating. Parents provided the researcher with recent hearing test results for the child.

Previous SLP reports from the Listening and Language Clinic indicated that participants had varying degrees of speech perception ability, language delay, and phonological disorder. Speech perception rated using the revised Categories of Auditory Performance (CAP-II; Gilmour, 2010) showed that Sylvia scored 8 (‘Follows group conversation in a reverberant room or where there is some interfering noise, such as a classroom or restaurant’), Clara scored 4 (‘Discrimination of speech sounds without lip reading’), and Brendan scored 5 (‘Understanding of common phrases without lip reading’). Sylvia and Clara have mild speech disorder but their speech was intelligible most of the time. Brendan’s speech intelligibility showed moderate–severe impairment. It is difficult for an unfamiliar listener to understand the language being produced by Brendan. Because the researchers were familiar with the phonological patterns used by Brendan, they were able to understand the phrases and sentences that he constructed. In addition, no vowel distortion occurred. Consonant error patterns such as ‘car’→ /tar/ were readily interpreted. Language scores for standardized tests Clinical Evaluation of Language Fundamentals: 4th edition (CELF-4; Semel et al., 2003) and South Tyneside Test of Syntactic Structure (STASS; Armstrong and Ainley, 1988) indicated slight language impairment for Sylvia and severe language impairment for Brendan and Clara. All were New Zealanders born to families who were monolingual speakers of English.

2 Procedure

Each child had a pre-test session, two mediated learning sessions, and one post-test session. The pre-test established baseline levels so that the MLE sessions could be designed. Each session took approximately an hour in a speech and language therapy clinic room, and was videotaped and transcribed by the first researcher. Some sessions were later transcribed by a second researcher to check reliability. The pre-test occurred 2–5 weeks before the first MLE. The two MLE sessions and the post-test occurred over approximately one week. The timetable varied because of family circumstances, as shown in Table 3.

Table 3.

Timing of MLE and post-test for the three participants.

Participants	MLE 1	MLE 2	Post-test
Sylvia	Day 1	Day 3	Day 4
Brendan	Day 1	Day 3	Day 5
Clara	Day 1	Day 1	Day 7

3 Test instrument

The Test of Narrative Language (TNL) manual indicates that the normative sample for standardization of the test consists of 1,059 children from 20 states. The TNL measures comprehension and production of stories and includes six tasks, all of which were given to the children. The authors report that TNL scores have good item reliability (alpha coefficients of 0.70–0.90) and test–retest reliability (0.81–0.85). TNL validity was confirmed by comparing scores with the results of language sample analyses and TOLD-P:3 (Test of Language Development – Primary: 3rd edition; Newcomer and Hammill, 1997); these comparisons show moderate–large statistically significant agreement. The inter-rater point to point agreement of scores for narrative comprehension was 94% and 90% for oral narrative. The TNL (Gillam and Pearson, 2004) was used as the pre- and post-test. Several words were changed to New Zealand English. The words Lisa, Raymond, clerk, Samantha, and trail were changed to Liz, John, waitress, Susan, and path, respectively. In addition, three adjustments to the test administration were made because of the children’s hearing loss. The researcher repeated questions two to three times, whenever children gave a verbal or non-verbal indication of not being able to hear the question, and the researcher changed the test’s abbreviated words to unabbreviated ones for conjunctions, such as we’re to we are and won’t to will not. In addition, the researcher used acoustic highlighting (Easterbrooks and Estes, 2007) in all sessions, including the test and MLE sessions. Acoustic highlighting is an AVT technique developed for children with hearing loss in which the clinician draws the child’s attention to specific words or phrases by adding extra emphasis using repetition, rate of speech, pitch and rhythm, and acoustic contrast (Easterbrooks and Estes, 2007). Consistency of acoustic highlighting between test and retest was controlled for by underlining the written words in the narrative script in the TNL by the principal researcher. All sessions and all assessments were conducted by the same person.

4 Construction of MLE sessions

After the pre-test the researcher analysed three of the TNL tasks. The first was a narrative retell task (McDonalds Story), the second required children to generate a narrative from a picture sequence (Late for School Story), and in the third task children generated a narrative from a single stimulus picture (Aliens Story). These subtests were analysed using Labov’s high-point analysis (Labov and Waletzky, 1967) in order to establish baseline functioning for the MLE.

During the MLE sessions the mediator (principal researcher) taught principles of fictional narrative production to the participants based on areas of difficulty found during pre-testing, in order to evaluate their ability as a narrative learner. The mediator developed teaching scripts to follow during the MLE sessions, which were reviewed by a second investigator who had previously used DA. The structure for the scripts was based on Miller et al.’s (2001) Dynamic Assessment and Intervention programme. The sessions used concrete stimuli and sequenced pictures, and each script was different. There were four goals in total; for example, teaching the child to include a complication (exciting, funny, and sad part), name people and places in the story, describe characters, and describe what happened next and what finally happened. Sylvia and Clara were given four goals, and there were two goals per MLE session for Brendan as his severe receptive and expressive language delay made four goals too difficult. The teaching scripts followed the MLE principles of intentionality, meaning, transcendence, and competence (Lidz, 1991).

All activities had the general goal of teaching children about telling a story that included all of the areas outlined by high-point analysis. Materials used during each MLE session included toys, stimulus pictures, sequenced pictures, and a maze. The mediator used a Christmas tree to visually reinforce goals, with the ‘who?’ and ‘where?’ at the top of the tree, the ‘exciting and sad part of the story’ at the middle, and ‘what happened next?’ between the start and the exciting part. At the bottom of the tree was ‘what happened at the end?’ The tree represented the idea that a story consists of a beginning, middle, and an ending. When the children reached the ending part of the story in the second MLE session, they were given small ‘Christmas presents’.

As the MLE approach requires the mediator to respond to each child based on his or her responses in order to modify his or her learning, the MLE sessions were not totally scripted, even though materials and session content stayed constant. In each session, the mediator introduced the goals by using different activities in order to represent the mediation of intentionality. Here is an example (where M = Mediator and C = Child):

We are going to have some fun today. See I have got my games here and we will play and we will tell some stories.

This is a boy whose mother asked him to buy some milk from the shop but look at him what is he doing?

(C response: Playing soccer)

Oh! What is the funny part in this story?

(C response)

When people tell stories, they talk about what happened to start the action. This is the exciting, funny, sad, and interesting part.

The session continued by doing another activity (e.g. pick a card and talk about the exciting part) to represent the ‘mediation of meaning’:

M holds four cards and asks C to pick up one.

M asks ‘What’s happened here, to the boy?’

(C response)

That is right. The bird pops the balloon.

Take another one and tell me the funny or interesting part. We are learning to say ‘what happened?’. We are talking about the interesting, sad, and funny part of the story. And if you don’t tell people, what is the interesting part in your story; you will not draw their attention and they won’t be interested to hear it.

Later on the mediator continued by relating the planned activities to school and home activities in order to represent ‘mediation of transcendence’:

You tell your mother what happens at school, don’t you?

(C response)

If you said ‘A boy fell from a tree’ your mother would say: ‘Oh! Poor boy’ so you told your mother the sad part of the story and you drew her attention to your story.

Finally the mediator helped the child to master the goal of the session by applying ‘mediation of competence’ as a new picture was presented and asking the child to describe it. Also the mediator provided the child with the needed encouragement about their achievements which were observed during the session:

What is this number for? Let’s find out. Yeah. Here it is. It is a picture, can you tell me what is the interesting part?

(C response.)

That’s great. At the beginning you couldn’t tell me the exciting part of the story but now you did it all by yourself. Well done. Today we learned a new thing. Can you tell me, what have we learned today?

5 Family involvement

For two of the participants (Clara and Brendan), their mothers attended the sessions as they wanted to know more about their own child’s ability to learn. The researcher answered parents’ questions at the end of the session. Parents asked if they could focus on the session goals at home and the researcher informed them that they should not target the goals addressed in the sessions until after the project, but rather encouraged them to increase social interaction with their children while reading stories, talking about past events and focusing on the achieved goals of each session only when the child drew their attention to one of the goals.

6 Data analysis

Test and retest scores for the TNL (Gillam and Pearson, 2004) were scored according to the manual. The child’s responses to the questions about the narrative comprehension are scored as correct, partly correct, or incorrect to generate a comprehension score. For the McDonald story subtest the child retells the stories and the child’s responses are scored as correct, partly correct, or incorrect. In the Late for School and Aliens subtests the child creates a fictional story that is scored based on setting, characters, story elements, vocabulary, grammar, and whether the story is complete and creative and makes sense. Thus, the oral narrative score of the TNL is based on a combination of skills: content, coherence, grammar, and overall completeness of the story. Based on Labov’s high-point analysis the McDonalds story does not contain evaluation and coda. Hence, the structural analysis for this subtest ends at end high point, which may not be a very appropriate model for fictional story but it is useful in case children are able to narrate a story that ends at end high point and differentiate this type of narrative from classic narrative. Late for School and Aliens stories represent classic narrative based on structural analysis. The principal researcher conducted and videotaped the MLE sessions and transcribed them. Half of the audio recorded assessment sessions were transcribed by another researcher so that inter-transcriber reliability could be determined. When there was a disagreement between the two transcribers (for approximately 10% of the children’s utterances), the two raters listened together to the audio recordings until they reached a consensus. In addition to the TNL, children were evaluated using a variety of measures (see Table 4). These outcome measures are:

the MLE Rating Scale, which measures the mediator quality of mediation during sessions;

the Modifiability Scale, which measures the child’s ability to learn; and

the Response to Mediation Scale, in which a child is scored based on observation of the joint activity between the mediator and the child.

Table 4.

Description of instruments and scales used to analyse the MLE and the children’s narrative.

Data	Instrument
Test versus retest narrative scores	Test of Narrative Language narrative production and comprehension subtests. Source. Gillam and Pearson (2004). This test generates standard scores and percentile ranks for a hearing population.
Quality of mediation during the sessions	MLE Rating Scale Source. Adapted from Lidz (1991); see Appendix 1. This has four components: intentionality, meaning, transcendence and competence rated on a 4-point scale (0–3). 0 = The component is not in evidence. 1 = the component is inconsistently in evidence. 2 = The component is consistently in evidence. 3 = The mediator included an elaborated statement to the child of the principle being taught during the session.
Child’s modifiability during the sessions	Modifiability Scale Source. Gutiérrez-Clellen et al. (1998), adapted from Lidz (1991) and Peña (1993); see Appendix 2. This measures three areas using a 4- or 5-point scale (1–4 or 1–5): 1. ABILITY to transfer (apply new strategies to a new task); 2. RESPONSIVENESS; 3. EXAMINER effort (the amount of effort that the mediator had to apply to induce change during mediation).
Child’s responsiveness to mediation during the sessions	Response to Mediation Scale. Source. from Lidz (2003); see Appendix 3. Six elements from the original scale were included, rated using a 5-point scale (1–5), ranging from failure to achieve behavior (= 1) to maximum achievement (= 5) and a ‘does not apply section’ option: 1. Self regulation of attention. 2. Interactivity with the mediator. 3. Responsiveness to initiations of mediator. 4. Response to challenge. 5. Use of adult as a resource when child needs help. 6. Interest in activity materials.
Narrative analysis	1. High-point analysis (Source. Labov and Waletzky, 1967). Six parts of personal narrative: abstract, orientation, complication, resolution, evaluation, and coda. MacDonald’s Retell, Late for School Story, Aliens Story are the oral narrative subtests of TNL that analysed according to high-point analysis over test and retest. 2. Structural Patterns (Source. McCabe and Rollins, 1994). Narratives were analysed to determine which of the following eight structural patterns they matched: two-event, leap frog, end-at-high-point, classic, chronological, impoverished, disoriented, and miscellaneous.

Thus, the scale measures the child’s learning behaviour that reflects responsiveness to the mediation experience. These scales were developed in consultation with Lidz (personal communication, February 2010). The original Response to Mediation Scale (Lidz, 2003) had 11 sections. Only six sections were included because the other five (self-regulation of motor activity, self-regulation of emotions, strategic problem solving, evidence of self-talk when working on challenging task, comprehension of the task) were not applicable. In addition to these outcome measures, a high-point analysis and a structural pattern analysis were performed.

7 Inter-rater reliability

Both the principal researcher and a second SLP scored the TNL. For each rating scale (see Appendices 1, 2 and 3), two trained, experienced SLPs performed independent ratings. Training on use of the scales was provided by the principal researcher, using examples from Gutiérrez-Clellen et al. (1998) and Lidz (2003). Inter-rater reliability was evaluated by determining percentage agreement between the two raters for each scale. The principal researcher was the mediator for the MLE sessions and hence was not able to rate her own performance using the MLE scale. The mediator was rated on the MLE scale by two blinded SLPs, one of whom was the third investigator. The SLPs were blinded to the goals and the planned structure of the MLE session. SLPs were provided with the transcribed MLE sessions and rated the mediator. Thus, the principal researcher was one of the raters for all measures, except for the MLE scale. MLE ratings were based on the transcripts only; for the Modifiability Scale and Response to Mediation Scale the second rater also had access to video recordings of the sessions. Percentage agreement ranged from 83% to 100% across measures (average 92%). When ratings differed the values were always within one scale point. Given the broadness of the definitions and the potential for differing evaluations, this is a good result for reliability.

IV Results

1 Test of Narrative Language (TNL)

Pre-test to post-test gains on the TNL, Narrative Comprehension, and Oral Narration subtests are shown in Figure 1. Subtest standard scores of 10 indicate performance at the 50th percentile. The standard deviation is 3 and hence standard scores of 7–13 represent typical performance (16th to 84th percentile). For both pre- and post-test Brendan scored below the 1st percentile for both subtests with raw scores changed from 0 to 3 for the Narrative Comprehension subtest and from 2 to 4 for the Oral Narrative, indicating low levels of narrative performance for both comprehension and expression. Standard scores for the Narrative Comprehension subtest, which was not targeted in the MLE sessions, show no change for Clara, with raw scores changed from 35 to 36, and minimal change for Sylvia, whose raw scores changed from 32 to 35. These scores are all above the norm, falling within the 84th to 98th percentiles. In contrast, standard scores for Oral Narrative (expressive), which was targeted in the MLE sessions, show substantial change for both Clara and Sylvia. Clara’s raw scores changed from 37 to 68 (improving from 3rd to 84th percentile), and Sylvia’s raw scores changed from 55 to 76 (improving from 84th to >99th percentile). Brendan did not show change between the pre-test and post-test measures. Standard scores corresponding to these raw and percentile scores are shown in Figure 1.

Figure 1.

Pre-test versus post-test Test of Narrative Language (TNL) standard scores for the Narrative Comprehension and Oral Narration subtests for the three participants.

2 MLE

The mediator (principal researcher) was assigned a rating of 2 or 3 on the MLE scale by both raters for each session. Rater 1 was the more experienced therapist; her ratings are shown in Table 5. Raters 1 and 2 each assigned a rating of ‘3’, indicating the highest level of mediation, to 83% of the sessions. A point-by-point comparison of the two raters showed 83% agreement between the MLE scale values assigned to the sessions. MLE Rating Scale results are shown in Table 5 for Rater 1. These results indicate that the SLP working with the children was using the mediated teaching techniques appropriately and consistently.

Table 5.

Rater 1 results for the MLE Rating Scale.

Participant	MLE Session	MLE Rating Scale (0–3)
Participant	MLE Session	Intentionality	Meaning	Transcendence	Competence
Sylvia	1	3	3	3	3
	2	3	3	2	3
Brendan	1	3	3	2	3
	2	2	2	3	3
Clara	1	3	3	3	3
	2	3	3	3	3

Note. Scale from 0 to 3, where 0 = component not evident in the session, and 3 = mediator made elaborated statements about the component in the session.

Source: Lidz (1991).

3 Modifiability

Modifiability Scale ratings show how much the participants responded to the MLE techniques. Figure 2 shows the results for the three participants averaged across the three scale items for the two MLE sessions. Sylvia was highly modifiable as she scored ‘4’ for both transfer and responsiveness in the two MLE sessions, and she required low effort from the mediator. Clara scored ‘3’ for transfer in both MLE sessions. Clara was able to apply the learned goals to improve her narrative skills, but with more prompting than Sylvia. Clara needed more effort from the mediator in the first session than the second, as she became more engaged with the mediator during the second session. Sylvia’s and Clara’s modifiability scores were consistent with the improvement in their standard scores for the Oral Narrative post-test. In the two MLE sessions, Brendan scored ‘2’ for transfer, indicating a minimal ability to use the newly learned goal. His overall responsiveness to the two MLE sessions was rated as ‘2’, which is low. Brendan required high effort from the mediator.

Figure 2.

Modifiability Scale ratings for the three scale items (transfer, response, effort). Transfer was rated using a 5-point scale (1-5); response and effort were rated using 4-point scale (1-4).

4 Response to Mediation

The Response to Mediation Scale ratings are another aspect of the child’s responsiveness to the MLE. ‘Response to mediation’ scale scores are about ongoing behaviours within the session, hence ratings are of such things as:

self-regulation of attention, ranging from needing to maintain attention to task by adult input to no input;

interactivity with the mediator, ranging from turn taking with mediator to initiating;

responsiveness to initiations of mediator, ranging from the child’s behaviour being resistive to the mediator to being responsive;

response to challenge; child’s behaviour ranging from refusing to persist to enjoying it;

use of adults as a resource when needing help; child’s behaviour ranging from not seeking adult help to valuing help provided; and

interest in activity; child’s behaviour ranging from not being interested in the materials to strong interest.

The Response to Mediation Scale ratings averaged across the six scale items, for the two MLE sessions for the three participants, are presented in Figure 3. Sylvia was highly responsive to mediation as she reached self-regulation with low input from the mediator, which was in evidence in her post-test TNL results. Figure 3 shows that Sylvia was slightly more responsive in the second session than the first. Clara showed more responsiveness to mediation in the second session than the first session. However, Brendan’s responsiveness to mediation was reduced in the second session compared to the first. The goal of the first session was ‘What is the exciting or sad part of the story?’ and Brendan was able to locate this. For the first session his responses included ‘Fall off bike, sad’, ‘He sad’, ‘He parked’. In the second session, Brendan commented ‘this is hard’ during the second goal ‘Who is he?’ as he was required to give a name to people in the picture and then make a sentence, but he was able to say ‘This is a boy called Sam. He fell off the bike and hurt his knee. He is sad.’ This was a challenging task relative to Brendan’s language level; however, his overall scores for the Response to Mediation Scale are comparable to Clara’s, indicating that he had behaviours that allowed learning. His Modifiability Scale scores, however, indicate that he required a lot of input from the mediator.

Figure 3.

Response to Mediation Scale ratings averaged across the six scale items.

5 Narrative analysis

Pre- and post-test Oral Narrative subtests were analysed using high-point analysis (Labov and Waletzky, 1967), and structural patterns (McCabe and Rollins, 1994) were determined. Sylvia was referred to the study with slight language impairment. Even though the pre-test standard scores of the TNL put her within the normal range for the Oral Narrative (expressive) subtest, the high-point analysis shows that she used general orientation (e.g. a boy) instead of using specific person, place, and behavioural situations. The high-point analysis for MacDonald’s Retell Story in pre-test indicates that she used orientation, complication, action, and resolution, while in the post-test she used more actions. For MacDonald’s Retell Story the structural analysis indicated no change, as she had reached ‘end high point’ on both occasions. The pre-test high-point analysis for Late for School Story shows that she used orientation, complication, action, and resolution (structural pattern ‘end high point’), but in the post-test Sylvia produced a more complex story that additionally included evaluation and coda (structural pattern ‘classic narrative’). The pre-test pattern is characteristic of the narrative of typically developing five-year-olds, and she was 7;4 years old; however, at post-test the structural pattern is age appropriate. High-point analysis for the pre-test Aliens Story showed that Sylvia’s narrative included orientation, complication, evaluation, resolution, and coda, which indicates classic narrative (e.g. ‘Well, one day, a boy and girl, they were going up to a park, and then they saw this sky ship, with aliens walking out, because they were going to the park too’). In the post-test she included more complications, resolutions, and evaluations (e.g. ‘One Sunny morning, Jessica and Michel were going to the park. They saw a big round orange and yellow, spaceship land. Some aliens were coming out; the mother alien said “Look this a wonderful place to have a holiday”). Although the post-test story was more well-structured and cohesive, the structural pattern classification was unchanged. Thus, the structural pattern analysis was too broad to capture some of the changes in Sylvia’s narrative. At post-test Sylvia used more complex language as she gave detailed description for characters, used adjectives to describe objects, used adverbial clauses to link actions, included dialogue of two characters, stated a detailed ending, and produced a complete and more organized story.

Brendan was referred to this study with severe language impairment, and did not improve his post-test scores on the Oral Narrative subtests. He had the poorest modifiability and responsiveness to mediation ratings during sessions. His oral fictional narrative consisted of short descriptive sentences in Late for School Story in the pre-test (e.g. ‘Wake up and this time’, ‘shoes on’), but in the post-test his narrative included orientation, complication, and coda, which indicates a group of structures. Brendan included orientation, resolution and evaluation in the pre-test and post-test of the Aliens Story, but in the pre-test he used confused word order to convey his message (e.g. ‘scared mum’, ‘Scaring’, ‘surprised monster biting’) in comparison to the post-test, in which he used the syntactic structure to tell about the agent, actions, or reactions (‘I’m scared’, ‘Mum not scared’, ‘He bite’). Overall, the post-test narrative production suggests that he may have had a slight gain in narrative skills from the two MLE sessions. According to the structural patterns analysis his oral fictional narrative was impoverished at pre- and post-test for Late for School Story and Aliens Story (he did not retell the McDonald Retell Story); this broad classification does not show the small changes that he appeared to make after the MLE sessions.

Clara was referred to the study with severe language impairment. Clara scored in the 3rd percentile for the Oral Narrative subtest of the TNL in the pre-test and 84th percentile in the post-test after the two MLE sessions. According to high-point analysis for MacDonald’s Retell Story, at the pre-test she used orientation, complication, and coda, while in the post-test she used resolution and more orientations, complications, and actions. The pre-test high-point analysis for Late for School Story shows that she used orientation, complication, resolution, and evaluation, but in the post-test her narrative contained orientation, coda and more complications, evaluations, and resolutions. Clara’s structural pattern analysis was at the end of high point for Late for School Story in pre-test, which is typical for five-year-old children. In the post-test the structural pattern shows that she had reached classic narrative, which is age appropriate. For Aliens Story, high-point analysis of the pre-test revealed that Clara’s fictional narrative consisted of orientation, complication, evaluation, resolution, and coda which indicates classic narrative (e.g. ‘The boy went home, to tell his mom, and the girl ran after him. When he got home, he told his mom and she didn’t believe him’). In the post-test, she used more resolutions and actions, which raised her fictional narrative to a greater level of sophistication (e.g. ‘They made good friends and um, it was getting dark; and the aliens didn’t have anywhere to stay, um, so Sarah and Jack offered them to come and stay at their house, and um, Jack and Sarah introduced the alien to their parents.’). The structural patterns for the pre- and post-test for the MacDonald’s Retell story reveals that she reached end high point whereas in the Aliens Story she reached classic pattern in pre- and post-tests. This broad classification of structural analysis did not show the changes she achieved in her post-test stories. Additionally, Clara’s post-test of the TNL showed more complex language as she identified story characters, used temporal and causal relationships between actions (e.g. ‘and’, ‘so’), stated a detailed ending, used the same tense through the story, and produced a complete organized story.

V Discussion

The purposes of this study were to examine whether DA is a useful alternative approach to traditional assessment of fictional narrative skills in children with moderate severe to profound bilateral hearing loss and to investigate whether a DA procedure can differentiate between normal versus poor language learning ability in children with hearing loss. DA allows the examiner to assess the child’s ‘potential for language change’ (Gutiérrez-Clellen, 2000). The incorporation of the MLE approach into DA allows for peer measurement of the mediator’s effectiveness in meeting objectives (measured using the MLE Rating Scale), the child’s responsiveness during sessions (measured using the Response to Mediation Scale), the ability of the child to carry over newly learned strategies, and the efforts of the mediator during the session (measured using the Modifiability Scale). These scales were useful because they provided a clear indication of what the child was capable of during a session. Overall, the results indicate that DA is useful for differentiating language learning difficulties in children with hearing impairments.

There were notable indications of language learning capability for both Sylvia and Clara. Sylvia and Clara both had narrative comprehension within normal limits on the TNL. Sylvia also scored within the normal range on the oral narrative production aspect of the TNL. It is possible that these children scored within normal limits because of the ‘acoustic highlighting’ that was used during the sessions; this may have given them an advantage. The pre- versus post-test analysis showed substantial differences in Oral Narrative subtest scores, and modifiability during MLE sessions was in evidence for these two participants. Prior to mediation, Sylvia had some strategies to build up a story, but her narratives were not highly structured and lacked the characteristic of a true story. Clara had inefficient strategies for narrating a fictional story.

Sylvia’s speech perception score was high, but her narrative in the pre-test was not highly structured. Clara and Brandon’s speech perception scores were lower (categories 4 and 5, respectively). Clara’s speech perception is slightly poorer than Brandon’s, but her narrative was better than his. Thus, there was not a clear relationship between narrative language and speech perception in these three children. It seems likely that narrative development in children with hearing loss depends on a combination of speech perception and other factors. This might be clarified based on Yoshinaga-Itano and Downey’s (1996) theory of the Colorado Progress analysis of written language (COPA). They explained that children are able to create narrative using life experience that is concentrated on scripts and sequenced actions in their daily life, and schemata that are based on sequenced events that are planned and stored in memory. Rumelhart (1980: 34) defined a schema as ‘a data structure for representing the generic concepts stored in memory’ and ‘there are schemata representing our knowledge about all concepts: those underlying objects, situations, events, sequence of events, actions and sequence of actions’. According to Yoshinaga-Itano and Downey, children acquire knowledge through an increased interaction with people and things in the environment; they are expected to gain more schemata, which help them to form information into ‘accessible pathways’. Children with hearing loss may not acquire enough schemata because of lack of interaction. They therefore need help from their parents and teachers to develop and acquire ‘missing schemata’. When children with hearing loss are provided with the missing information, they can combine it with the previous stored ones, and they will be able to produce extravagant language (Yoshinaga-Itano and Downey, 1996). In the current study, therapy focused on the language to encode experiences, rather than the schemata themselves, and it is presumably the language that was provided during the study sessions that accounts for both Sylvia and Clara producing better stories at the post-test. The changes observed suggest that these children presented with language difficulties because of lack of appropriate input rather than inability to learn language. Sylvia and Clara may have normal language-learning ability; however, both children have had limited language exposure due to their hearing loss impacting on the development of auditory skills, which are essential to comprehend heard stories. They have language and speech impairment that aggravates their problem as they have less interaction experience in retelling stories to their parents or peers than children with normal hearing. Hence they are expected to learn new narrative skills when given the opportunity to learn. This would be consistent with Budoff’s (1987) suggestion that children who have limited language experience and normal language-learning ability learn new language rules willingly, given the opportunity. It is also possible that, after the MLE sessions, Sylvia and Clara became aware of what a fictional oral narrative consists of according to Labov’s high-point analysis.

Brendan’s results indicated a language-learning problem. His post-test narrative language test scores showed no change, he scored low on modifiability measures, and he had difficulty learning new language rules for narration. These findings are consistent with Peña et al.’s (2001) report that children with high modifiability scores and good post-test gains in the targeted areas during MLE are expected to have typical language learning abilities, and children with slight modifiability and minimal post-test gains after MLE are expected to have limited language due to poor language learning ability.

A mismatch between standardized testing and DA was seen by Camilleri and Law (2007) who found that monolingual children and children with English as an additional language (EAL) scored similarly on DA assessments of receptive vocabulary, but the children with EAL performed much more poorly on standardized testing. Both groups had suspected language impairment. Hessels et al. (2008) found a significant but moderate correlation between DA assessment of cognitive abilities and standardized test scores, and attributed the achievement on DA to the child acquiring knowledge and learning strategies. It could be argued that Brendan’s results relate to his presenting with less language initially. The sentence-level language he demonstrated (2–3 words) would indicate that his text-level language (i.e. narrative) could not be expected to be very advanced, and the sentence-level reproduction problems in the sessions may have been an extra barrier to his learning. On the other hand, the low levels reached by Brendan may be a result of a language learning difficulty, as the DA results suggest. Brendan’s sentence structure in the TNL post-test showed some change, with more inclusion of the agent before the object or the verb in post-test narratives, suggesting that he has learning potential. Since some learning potential was detected he may benefit from both enhanced knowledge of language and better learning strategies, such as improved self-regulation of attention. More research is needed to clarify this point. This study needs to be replicated with studies of children with comparable language levels to Brendan, to see if DA can show differences in language learning capability within such children who present with low scores on standardized tests.

The high-point analysis was also useful for documenting improvements in the details and structure of the children’s narratives after two MLE sessions. High-point analysis might be a useful clinical method for assessment and for setting goals for narrative intervention. The primary advantage of high-point analysis over the TNL is that it provides useful information about narrative structure that can form the basis for session planning and DA. The structural pattern analysis provided broad classifications that did not show any distinction in the children’s stories between pre- and post-test. A different analysis tool may be needed to use in combination with Labov’s high-point analysis for examining structural patterns.

This study did raise other points about the use of DA. Examination of children’s learning behaviour is rarely done in speech language therapy contexts, but the use of these measures proved very interesting with these children. The results support the idea that if children do not match typical developmental trajectories for speech and language (which applies to children from culturally and linguistically diverse backgrounds, and those with hearing loss) then learning behaviour may tell us more information than point-in-time test scores. Judgements of intervention need could well flow from such information. For example, those most likely to learn rapidly with intensive input could be differentiated from those needing additional support, and therapy resources could be allocated accordingly. Other studies have shown the value of DA for typically developing bilingual children (Kapantzoglou et al., 2012). DA may be more flexible than point-in-time standardized assessments, with greater validity for culturally and linguistically diverse children, and hence the DA approach has considerable potential for assessment of children with both typical and atypical development.

The teaching approach, although intended for assessment and not for intervention, may also have some potential to be more widely applied. It is based on making the learning targets specific, and gives clear guidelines for teaching that could add to the speech language therapy repertoire. Scripts were developed for each DA session. They included examples of teacher instructions, teaching sequences, activities and required tools, and examples of stories that the mediator could target with the child. This was to ensure consistency between sessions and between children. Although this initial step was time-consuming, therapists could widely apply DA without the use of scripts once the elements of DA have been mastered.

Overall, the results for this small group of participants show that DA is a useful alternative approach to traditional assessment and that it is a useful approach for differentiating between children who appear to have poor language as a result of language learning difficulties and children who have poor language as a result of lack of input due to hearing loss. The present study establishes a baseline for the progression of research in the area of assessing narrative ability for children with hearing loss.

Footnotes

Appendix 1.

Appendix 2.

Appendix 3. Acknowledgements

We acknowledge the cooperation between the Discipline of Speech Science at the University of Auckland, New Zealand and the Center for Phonetics Research, University of Jordan, which provided the opportunity for the first researcher to conduct this work. We are also grateful for the comments of two anonymous reviewers.

Declaration of conflicting interests

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the article.

Funding

This research was supported by a University of Auckland Summer Research Scholarship.

References

Applebee

(1978) The child’s concept of story: Ages two to seventeen. Chicago, IL: University of Chicago Press.

Armstrong

Ainley

(1988) South Tyneside assessment of syntactic structures. Ponteland: STASS Publications.

Botting

(2002) Narrative as a tool for the assessment of linguistic and pragmatic impairments. Child Language Teaching and Therapy 18: 1–21.

Boudreau

(2008) Narrative abilities advances in research and implications for clinical practice. Topics in Language Disorders 28: 99–114.

Briscoe

Bishop

DVM

Frazier Norbury

(2001) Phonological processing, language, and literacy: A comparison of children with mild-to-moderate sensorineural hearing loss and those with specific language impairment. Journal of Child Psychology and Psychiatry 42: 329–40.

Budoff

(1987) Measures for assessing learning potential. In: Lidz

(ed.) Dynamic assessment: An interactional approach to evaluating learning potential. New York: The Guilford Press, 173–95.

Camilleri

Law

(2007) Assessing children referred to speech and language therapy: Static and dynamic assessment of receptive vocabulary. Advances in Speech–Language Pathology 9: 312–22.

Crais

Lorch

(1994) Oral narrative in school-age children. Topics in Language Disorders 14: 13–28.

Crosson

Geers

(2001) Analysis of narrative ability in children with cochlear implants. Ear and Hearing 22: 381–94.

10.

Curtiss

Prutting

Lowell

(1979) Pragmatic and semantic development in young children with impaired hearing. Journal of Speech, Language, and Hearing Research 22: 534–52.

11.

Dockrell

(2001) Assessing language skills in preschool children. Child Psychology and Psychiatry Review 6: 74–85.

12.

Easterbrooks

Estes

(2007) Helping deaf and hard of hearing students to use spoken language. Thousand Oaks, CA: Crowin Press.

13.

Feuerstein

(1979) Dynamic assessment of retarded performers: The learning potential, assessment device, theory, instruments and techniques. Baltimore, MD: University Park Press.

14.

Feuerstein

Rand

Hoffman

(1981) The dynamic assessment of retarded performers: The learning potential assessment device, theory, instruments and techniques. International Journal of Rehabilitation Research 4: 465–66.

15.

Gillam

Pearson

(2004) Test of narrative language. Austin, TX: PRO-ED.

16.

Gilmour

(2010) The inter-rater reliability of categories of auditory performance-II (CAP-II). Southampton: University of Southampton.

17.

Griffin

Hemphill

Camp

Wolf

(2004) Oral discourse in the preschool years and later literacy skills. First Language 24: 123–47.

18.

Gutiérrez-Clellen

(2000) Dynamic assessment: An approach to assessing children’s language-learning potential. Seminars in Speech and Language 21: 215–22.

19.

Gutiérrez-Clellen

Brown

Conboy

Robinson-Zañartu

(1998) Modifiability: A dynamic approach to assessing immediate language change. Journal of Children’s Communication Development 19: 31–42.

20.

Gutiérrez-Clellen

Peña

(2001) Dynamic assessment of diverse children: A tutorial. Language, Speech, and Hearing Services in Schools 32: 212–24.

21.

Gutiérrez-Clellen

Quinn

(1993) Assessing narratives of children from diverse cultural/linguistic groups. Language, Speech, and Hearing Services in Schools 24: 2–9.

22.

Hasson

Joffe

(2007) The case for dynamic assessment in speech and language therapy. Child Language Teaching and Therapy 23: 9–25.

23.

Hessels

MGP

Berger

J-L

Bosson

(2008) Group assessment of learning potential of pupils in mainstream primary education and special education classes. Journal of Cognitive Education and Psychology 7: 43–69.

24.

Hughes

McGillivray

Schmidek

(1997) Guide to narrative language: procedures for assessment. Eau Claire, WI: Thinking Publications.

25.

Kapantzoglou

Restrepo

Thompsona

(2012) Dynamic assessment of word learning skills: Identifying language impairment in bilingual children. Language, Speech, and Hearing Services In Schools 43: 81–96.

26.

Labov

Waletzky

(1967) Oral versions of personal experience. Seattle, WA: University of Washington Press.

27.

Laing

Kamhi

(2003) Alternative assessment of language and literacy in culturally and linguistically diverse populations. Language, Speech, and Hearing Services in Schools 34: 44–55.

28.

Law

Boyle

Harris

Harkness

Nye

(2000) Prevalence and natural history of primary speech and language delay: Findings from a systematic review of the literature. International Journal of Communication Disorders 35: 165–88.

29.

Lidz

(1991) Practitioner’s guide to dynamic assessment. New York: The Guilford Press.

30.

Lidz

(2003) Early childhood assessment. Hoboken, NJ: John Wiley & Sons.

31.

Lidz

(2004) Successful application of a dynamic assessment procedure with deaf students between the ages of four and eight years. Educational and Child Psychology 21: 59–73.

32.

McCabe

Bliss

Barra

Bennett

(2008) Comparison of personal versus fictional nrratives of children with language impairment. American Journal of Speech-Language Pathology 17: 194–206.

33.

McCabe

Rollins

(1994) Assessment of preschool narrative skills. American Journal of Speech-Language Pathology 3: 45–56.

34.

Miller

Gillam

Peña

(2001) Dynamic assessment and intervention: Improving children’s narrative abilities. Austin, TX: PRO-ED.

35.

Morgan

Herman

Woll

(2006) Language impairments in sign language: Breakthroughs and puzzles. International Journal of Communication Disorders 35: 95–116.

36.

Newcomer

Hammill

(1997) Test of language development: Primary: 3rd edition (TOLD-P:3). Austin, TX: PRO-ED.

37.

Peña

(1993) Dynamic assessment: A non-biased approach for assessing the language of young children. Unpublished doctoral dissertation, Temple University, Philadelphia, PA.

38.

Peña

(2000) Measurement of modifiability in children from culturally and linguistically diverse background. Communication Disorders Quarterly 21: 87–97.

39.

Peña

Iglesias

Lidz

(2001) Reducing test bias through dynamic assessment of children’s word learning ability. American Journal of Speech-Language Pathology 10: 138–54.

40.

Peterson

McCabe

(1983) Developmental psycholinguistics: Three ways of looking at a child’s narrative. New York: Plenum Press.

41.

Pressnell

(1973) Hearing-impaired children’s comprehension and production of syntax in oral language. Journal of Speech, Language, and Hearing Research 16: 12–21.

42.

Quinto-Pozos

Forber-Pratt

Singletonb

(2011) Do developmental communication disorders exist in the signed modality? Perspectives from professionals. language, Speech, and Hearing Services in Schools 42: 423–43.

43.

Remine

Brown

Cowan

(2003) Assessing children with profound hearing loss and severe language delay: Getting a broader picture. Cochlear Implants International 4: 73–84.

44.

Rumelhart

(1980) Schemata: The building blocks of cognition. In: Spiro

Bruce

Brewer

(eds) Theoretical issues in reading comprehension: Perspectives from cognitive psychology, linguistics, artificial intelligence, and education. Hillsdale, New Jersey: Lawrence Erlbaum Associates, 33-58.

45.

Schönweiler

Ptok

Radü

H-J

(1998) A cross-sectional study of speech- and language-abilities of children with normal hearing, mild fluctuating conductive hearing loss, or moderate to profound sensoneurinal hearing. International Journal of Pediatric Otorhinolaryngology 44: 251–58.

46.

Semel

Wiig

Secord

(2003) Clinical evaluation of language fundamentals: 4th edition (CELF-4). Chicago, IL: Harcourt assessment.

47.

Swanwick

Watson

(2005) Literacy in the homes of young deaf children: Common and distinct features of spoken language and sign bilingual environments. Journal of Early Childhood Literacy 5: 53–78.

48.

Tomblin

Records

Buckwalter

Zhang

Smith

O’Brien

(1997) Prevalence of specific language impairment in kindergarten children. Journal of Speech, Language and Hearing Research 40: 1245–60.

49.

Tzuriel

Caspi

(1992) Cognitive modifiability and cognitive performance of deaf and hearing preschool children. The Journal of Special Education 26: 235–52.

50.

Vygotsky

(1978) Mind in society: The development of higher psychological processes. (Edited by Cole

John-Steiner

Scribner

Souberman

.) Cambridge, MA: Harvard University Press.

51.

Yoshinaga-Itano

Downey

(1996) Analyzing deaf or hard-of-hearing students’ written metacognitive strategies and story grammar propositions. The Volta Review 98: 63–64.

52.

Yoshinaga-Itano

Snyder

(1984) Story grammar analysis of deaf and hearing children’s written compositions. Unpublished paper presented at the American Speech-Language-Hearing Association Convention, San Francisco, CA.

53.

Yoshinaga-Itano

Sedey

Coulter

Mehl

(1998) Language of early- and later-identified children with hearing loss. Pediatrics 102: 1161–71.

54.

Young

James

DGH

Brown

. (1997) The narrative skills of primary school children with a unilateral hearing impairment. Clinical Linguistics and Phonetics 11: 115–38.