مهارت های تکرار غیر کلمه ای در کودکان جوان که لکنت زبان دارند/ندارند

The purpose of this study was to assess the nonword repetition skills of 24 children who do (CWS; n = 12) and do not stutter (CWNS; n = 12) between the ages of 3;0 and 5;2. Findings revealed that CWS produced significantly fewer correct two- and three-syllable nonword repetitions and made significantly more phoneme errors on three-syllable nonwords relative to CWNS. In addition, there was a significant relationship between performance on a test of expressive phonology and nonword repetition for CWS, but not CWNS. Findings further revealed no significant fluctuation in fluency as nonwords increased in length. Taken together, findings lend support to previous work, suggesting that nonword repetition skills differ for CWS compared with CWNS, and that these findings cannot be attributed to (a) weak language performance on the part of CWS, or (b) the occurrence of stuttering in the course of nonword production. Educational objectives: After reading this article, the learner will be able to: (a) describe one common means of assessing phonological working memory in children; (b) summarize the performance differences of children who stutter compared to peers on a nonword repetition task; (c) compare the results of the present study with previous work in this area.

As a group, children who stutter (CWS) tend to differ from their peers in a range of areas, including language (e.g., see Hall, 2004 and Weiss, 2004, for recent reviews). Perhaps because differences in the language performance of CWS tend to be subclinical (i.e., not constituting a language disorder), the literature has not focused on the many language-related areas that have been associated with language performance in other populations. In particular, one language-related area that has received considerable attention is the role of phonological working memory in the language performance of children with specific language impairment (SLI; e.g., Baddeley & Wilson, 1993; Botting & Conti-Ramsden, 2001; Conti-Ramsden, 2003 and Ellis Weismer et al., 2000; Gathercole & Baddeley, 1990; Gray, 2003; Marton & Schwartz, 2003; Montgomery, 1995a; Munson, Kurtz, & Windsor, 2005). As a group, these studies generally reveal that nonword repetition represents an area of weakness for children with SLI. For example, Montgomery (1995b) examined working memory and sentence comprehension in school-age children with SLI and language-matched peers with typically developing language using nonword repetition and sentence comprehension tasks. On the nonword repetition task, the typically developing children performed significantly better than the SLI children on three- and four-syllable nonwords. On the sentence comprehension task, the two groups differed only in comprehension of redundant sentences, with the SLI group performing more poorly on these. Findings further revealed a significant, positive relationship between children's nonword repetition and sentence comprehension scores. These findings were interpreted as evidence that limited phonological working memory capacity in children with SLI contributed to their poor auditory comprehension of sentences. More recently, Sahlen, Reuterskiold-Wagner, Nettelbladt, and Radeborg (1999) studied nonword repetition in 27 young children with language impairments, focusing on the relationship between nonword repetition and grammatical skills, and nonword repetition and phonological skills. They found significant relationships between nonword repetition performance and children's phonological stage and expressive grammar. However, when a subgroup of participants with the same phonological level was divided into two groups according to grammatical skills, the two groups did not differ in nonword repetition performance. This suggests that phonological skills, more so than expressive syntax, are related to nonword repetition performance. Munson et al. (2005) examined nonword repetition with items varying in phonotactic probability. Children with SLI were less accurate in nonword repetition than age-matched peers, with phonotactic probability impacting the SLI group's performance more than it impacted the performance of peers. These and other recent studies have focused not only on phonological working memory skills in children with language difficulties, but also on what nonword repetition, as a construct, measures in these children. Baddeley's model (Baddeley, 1986; Gathercole & Baddeley, 1993) is widely cited as the basis for research in phonological working memory (cf. Cowan, 1996; Just & Carpenter, 1992). The model accounts for ways in which information is held in memory until it is needed for recall. The phonological loop, one of three components of the model, processes phonological information. It consists of a storage component and a rehearsal mechanism. Phonological information can be held in storage for only a short period of time (seconds) unless the information is rehearsed. By rehearsing, the information can be “refreshed” within the storage component, enabling it to remain within memory for a longer period of time (see Baddeley, 2003, for a more detailed summary of the model). 1.1. Nonword repetition as a measure of phonological working memory Nonword repetition tasks have been widely used to estimate phonological working memory skills in children (e.g., Dollaghan et al., 1993 and Dollaghan et al., 1995; Dollaghan & Campbell, 1998; Gathercole & Baddeley, 1996) and adults (Gupta, 2003). These tasks essentially rely on retrieval and output as the response that provides information about storage and rehearsal capabilities. That is, a participant who is able to retrieve a nonword stimulus and produce it accurately is presumed to have relied upon adequate rehearsal and storage abilities to reach that point. However, there has been considerable discussion about whether nonword repetition tasks are appropriate measures of phonological working memory (e.g., Howard & van der Lely, 1995; Van der Lely & Howard, 1993), especially given the range of other skills required to repeat nonwords accurately. Within the SLI literature, for example, Edwards and Lahey (1998) examined several components of nonword repetition that may impact performance, including the roles of auditory discrimination, motor planning (response time latency) and execution (duration of responses), and prior vocabulary knowledge in children with SLI and their peers with typical language skills. They found that children in the SLI group were less accurate in nonword repetition, and that these limitations could not be attributed to auditory discrimination, motor planning, or lexical skills. Furthermore, overall expressive language scores, but not receptive scores, were found to partially account for the nonword repetition performance of children in the SLI group. The authors interpreted these findings to suggest that difficulties in nonword repetition among children with SLI were due, in part, to their poorer phonological representations of the nonwords. There is substantial documentation establishing a relationship between vocabulary knowledge and nonword repetition skills, although the directionality of this relationship is not clear (e.g., Gathercole & Baddeley, 1990; Gathercole, Willis, Emslie, & Baddeley, 1991; Gathercole, Service, Hitch, Adams, & Martin, 1999). Some investigators have suggested that phonological working memory (as measured by nonword repetition) drives vocabulary development (Gathercole & Baddeley, 1990; Gathercole et al., 1991 and Gathercole et al., 1999), whereas others have argued that increasing vocabulary knowledge enables a child to perform better on nonword repetition tasks (e.g., Dollaghan et al., 1995; Snowling, Chiat, & Hulme, 1991). Findings from a study by Dollaghan et al. provide support for this latter position. They administered a three- and four-syllable nonword repetition task to 30 boys, ages 9;10–12;5 years, with typically developing language. They found that nonwords whose stressed syllable was a real word (e.g., “bathesis”, where “bath” is the stressed syllable) were repeated with greater accuracy than nonwords whose stressed syllable was not a real word (e.g., “fathesis”). In addition, an error analysis revealed that children were significantly more likely to convert part of a nonword into a real word than they were to convert a “real word” segment into a nonword. As the authors point out, these results highlight one way in which lexical knowledge can aid a child in repetition of nonwords. Another explanation for the observed relationship between vocabulary and nonword repetition for which there is some empirical support is that phonological processing serves as an intervening variable (Bowey, 1996, Bowey, 2001 and Metsala, 1999). For example, Metsala proposed that as children develop phonemic segmentation abilities, this skill aids both word learning and the ability to repeat nonwords. In sum, the nature of the association between nonword repetition and language skills is not yet fully understood, but it is clear that nonword repetition is not a pure measure of phonological working memory. In the case of CWS, given that stuttering has an impact on speech output, one might also expect that output processes (e.g., articulatory speaking rate, response time, etc.), in addition to working memory abilities, might impact performance in a nonword repetition task. Moreover, although a child's level of (dis)fluency might give some indication of the child's output processes, it alone is not sufficient to infer the extent to which speech output impacts nonword repetition performance. Another consideration, particularly for CWS, in interpreting performance on a nonword repetition task is the observed association between nonword repetition and expressive phonological skills (e.g., Bowey, 2001 and Sahlen et al., 1999). This link is particularly germane for CWS, because phonological difficulties are a common concomitant disorder within this group of children (e.g., Arndt & Healey, 2001). Therefore, from a methodological standpoint, it would seem important to take consistent phonological errors into consideration when scoring nonword repetition attempts, so that CWS are not penalized for errors they produce consistently in speech production. At present, nonword repetition represents the standard for measuring phonological working memory in individuals with typical and atypical speech–language production. Thus, a nonword repetition task would appear to be an appropriate assessment tool for initial exploration of the phonological working memory abilities of CWS. However, as revealed in studies of children with SLI, it is equally clear that, in addition to phonological working memory, nonword repetition relies upon a range of skills, including those related to speech output. 1.2. Memory processes in individuals who stutter Within the area of fluency disorders, several models, such as the covert repair hypothesis (Postma & Kolk, 1993) and the EXPLAN theory (e.g., Howell, Au-Yeung, & Sackin, 1999), include phonology and phonological processing as components in their explanation of stuttering. Moreover, as previously indicated, phonological disorders frequently co-occur with stuttering in young children (Arndt & Healey, 2001). Therefore, given that current models of stuttering and clinical observations of CWS emphasize the role of aspects of phonology in understanding the population and the disorder, it seems appropriate and timely to devote attention to the phonological working memory capabilities of those who stutter. To date, however, only a few studies have examined broad aspects of memory in individuals who stutter. In one such study, adults who stutter (AWS) and their fluent controls were asked to reproduce, in writing, sets of four CVC syllables following their presentation and an intervening task (Bosshardt, 1993). Findings revealed that AWS reproduced significantly fewer syllables in the correct position than their fluent counterparts. Taken together, findings were interpreted to suggest that phonological encoding is weaker and phonological rehearsal time is slower in AWS. In an earlier study, Bosshardt (1990) examined both children (mean age = 13.8 years) and AWS, along with their normally fluent counterparts in their rates of subvocalization and reading, as these measures have been shown to correspond with phonological working memory skills (Baddeley, Thomson, & Buchanan, 1975). Participants read one set of phrases (definite article + noun) aloud and another set of phrases silently, using a key press to indicate when they had finished reading each phrase. The time interval from each stimulus presentation to the participant's response was then recorded for both conditions. Results revealed that both children and AWS exhibited longer response times than their normally fluent counterparts in the two reading conditions. Moreover, for the silent reading condition, this difference was greater between the groups of children than between the groups of adults. Although this study did not attempt to examine phonological memory processes per se, the authors pointed out that their findings suggest that individuals who stutter (including children) subvocalize more slowly, a quality that may impact their ability to rehearse information (within the context of the phonological loop; see Baddeley, 1986) as efficiently as their peers. More recently, Hakim and Ratner (2004) examined phonological working memory skills in CWS ages 4–8 years. They evaluated participants’ performance on the Children's Test of Nonword Repetition (CNRep; Gathercole, Willis, Baddeley, & Emslie, 1994), a test widely used in studies of phonological working memory in children with typically developing language (e.g., Gathercole et al., 1999 and Gathercole et al., 1997) and language impairments (e.g., Bishop et al., 1999 and Bishop et al., 1996). On the CNRep, children are presented with nonwords of 2–5 syllables in length via tape recording and then asked to repeat each nonword. Items are scored as correct or incorrect. They found that, at the three-syllable length, CWS repeated significantly fewer items correctly and exhibited more phoneme errors (both findings represented large effect sizes) than normally fluent children. However, the fluency of nonword productions did not change as a function of increased word length—that is, CWS were just as fluent on longer nonwords as on shorter ones. As Hakim and Ratner (2004) acknowledged, their nonword repetition findings were likely impacted by two factors. First, as their tabled data suggest, ceiling and floor effects were likely operating for the shorter and longer syllable lists, respectively. Second, with small sample sizes of eight children per group, it is unlikely that anything but a large statistical effect could be detected, due to insufficient power (see Jones, Gebski, Onslow, & Packman, 2001, for discussion of this issue). It should be noted that some reports of differences in nonword repetition in children with and without language impairments have employed larger samples, resulting in greater statistical power (e.g., Edwards & Lahey, 1998). However, when clinical populations are involved, one relevant question is whether small effect sizes (i.e., those that can be detected with larger samples) are clinically significant (see Jones et al., 2001). Thus, it seems reasonable to argue, particularly in preliminary work on the nature of any differences between groups in aspects of phonological memory, that the between-groups differences of interest are those that constitute relatively large effects. Of studies that have sought to examine memory skills in individuals who stutter, Hakim and Ratner's (2004) work is particularly interesting because it links conceptually the fairly extensive literature on phonological working memory in typically developing and language disordered populations to questions of the phonological memory capabilities of CWS. The present study is intended as a follow-up study to and partial replication of their work. We chose to address the issue of the ceiling effects observed by Hakim and Ratner by recruiting a sample of younger CWS within a more restricted age range (3–5 years of age). Younger CWS provide the opportunity for examination of phonological working memory during a time of substantial language and motoric development, relative to school-age CWS. As such, the overall purpose of the present study was to characterize the phonological working memory skills of young CWS relative to their normally fluent peers. In particular, our research questions were: 1. Do young CWS differ from their normally fluent peers in the number of correct responses produced on a nonword repetition task? Similarly, do young CWS differ from their normally fluent peers in the number of phoneme errors produced on this task? 2. Is there a relationship between speech–language performance and nonword repetition performance for CWS and their normally fluent peers? 3. For CWS, is there a difference in the fluency of nonword repetition responses as nonword length (in syllables) increases?