Conversation Skills and Hearing Loss [STUDY]
Conversational Skills of Children with Normal Hearing and Children with Impaired Hearing in an Integrated Setting
Jill Duncan, Ph.D., LSLS Cert. AVT
This study examines the conversation skills used by children with impaired hearing and their chronological age matched hearing peers in an integrated setting. Data were collected from a naturalistic (integrated kindergarten context), as the children interacted with their peers, and a quasi-naturalistic setting (tutorial room context), as they interacted in dyads. A modified version of the Social Organization of Discourse Checklist/Conversational Skills (Beattie, 1990) was used to code both sets of data. Children with impaired hearing were matched to their hearing peers based on chronological age. The Wilcoxon Signed Rank Test, a non-parametric test which is used to test the null hypothesis that two populations are the same in terms of their ranks, was used to compare the results of the chronological age matched pairs in each of the dyad and integrated kindergarten contexts. Analysis of the data documented that there were no significant differences in the majority of the conversational skills used by the children with normal hearing and the children with impaired hearing.
Conversational organization theorists stress that each speech community has expectations or rules concerning how conversation will be conducted (Kretschmer & Kretschmer, 1989) and each communication event has a specific set of rules of organization (Brown & Yule, 1983; Stubbs, 1983; Tannen, 1984).
The components associated with Conversational Organization Theory may be studied through the examination of the social organization of discourse. Beattie (1990) defines social discourse as the ability to initiate and maintain a dialogue with communicative partners over several conversational turns. These communication components constitute the rules of being “social" which are essential to successful social interactions. For Beattie, these rules focus on the listener’s and speaker’s role in conversation and include topic initiation, maintenance, shifting and termination. Roth and Spekman (1984) were among the first to introduce the concept of social organization of discourse, which they describe as the child’s ability to function in, and contribute to, an ongoing conversation.
Initiation is one component of the social organization of discourse. It concerns the communicator’s ability to initiate topics successfully as well as the type of initiation strategies used. Mueller (1972) studied the conversation of preschool children in dyads and concluded that there are several variables which determine the success or failure of a message: articulation clarity, use of attention-getting devices, relevance of the content, eye contact, and physical proximity. Mueller found that although securing the communication partner’s attention was the best overall indicator of success, other variables such as mutual desire for communication and intelligibility of the speaker contributed to the dynamics of the conversation.
Conversational maintenance is also an aspect of social organization of discourse and can give an indication of a communication partner’s linguistic as well as communicative competence. Roth and Spekman (1984) stress that topic maintenance is generally dependent on the contingency of the response to a preceding message. Bloom, Rocissano, and Hood (1976) and Keenan (1975) note that some maintenance attempts simply continue the topic without adding new information while others not only maintain the conversation but make a significant contribution with new information. Roth and Spekman (1984) state that studying maintenance in child conversation attempts can provide insight into how the child’s ongoing monitoring of the conversation takes place. It will also identify how the child’s ability to make sense of ambiguous messages as well as use of strategies that signal listener or speaker confusion develops.
Another factor in the social organization of discourse is the way in which a topic is terminated. However, while it is recognized that it is important to exit a topic in a manner that is socially acceptable, there has been to date very little research which addresses the type and appropriateness of termination strategies used by children.
The quality of a social discourse interaction depends on many contextual factors. Some of these factors include environment, communication partner and mood of the participants. The context in which the interaction occurs is also directly related to social discourse expectations and outcomes (Bloom, 1970; Halliday, 1975). There are many possible social contexts. Peer culture is one context in which there are identifiable rules necessary for effective communication. A knowledge of the social rules within the peer culture is essential for successful communication to occur.
Social discourse research involving children with impaired hearing is scarce. Most studies involving children with impaired hearing focus on specific syntactic components of language; few consider social discourse performance (Wood & Wood, 1991). Those studied which do focus on social interaction concentrate on the non-linguistic behaviors of the children at play, and not on language (e.g., Doyle, Connolly & Rivest, 1980; Higginbotham & Baker, 1981; Lederberg, 1991; Lederberg, Ryan & Robbins, 1986; Vandell & George, 1981).
The purpose of this study is to examine an aspect of social discourse, conversational skills, used by children with impaired hearing in order to determine if the conversational skills are similar to or distinct from their chronological age matched hearing peers within their integrated educational setting.
The two groups of participants involved in this study were children with normal hearing and children with impaired hearing exposed to oral language enrolled in the integrated kindergartens at the Speech and Hearing Centre, Western Australia (the Centre). Enrolment in each of the three integrated kindergartens ranged from 19-22 children. Two of the integrated kindergartens enrolled children who were four years of age and one integrated kindergarten enrolled children who were five years of age. The Centre is located in suburban Wembley, Perth, Western Australia. Children with impaired hearing from throughout the metropolitan area attend the parent-infant programs, childcare and kindergartens located on the campus, while hearing children who attend the various programs offered at the Centre are from Wembley and neighboring suburbs. According to the "Methods and Philosophies" statement published by the Centre in 1990, “the Speech and Hearing Centre subscribes to a philosophy which promotes auditory-verbal learning for hearing impaired children.” The Centre is committed to providing maximum opportunities for the child with impaired hearing to develop to his/her fullest potential.
The group with impaired hearing consisted of 10 children with profound hearing impairment and one child with a severe hearing impairment (five males and six females), 3.6 - 5.9 years of age (see Table 1). Criteria for inclusion of the children with impaired hearing in this study were: hearing impairment of 80dB or greater (better ear average at 500, 1000, and 2000 Hz; ANSI, 1969); no other known impairments; regular attendance in an integrated kindergarten; and location within the normal intelligence range, as determined by the school psychologist. The mean better ear average for the participants with impaired hearing was 103dB HL, the mode was 100dB HL and the range was 80dB HL to 120dB HL. All of the children with impaired hearing wore bilateral behind the ear hearing aids at the time of the data collection.
All of the children had been hearing impaired since birth except for two who lost their hearing due to meningitis before language had developed (three months and four months respectively). None of them had parents with impaired hearing except one whose mother had a unilateral hearing loss which was profound in nature and it did not have an obvious effect on her expressive or receptive language. None of them had been exposed to a formal signed communication system. All the children had been in a program in which residual hearing had been stimulated. At the time of the data collection, they were considered oral in their communication function.
The hearing group consisted of 11 children (five males and six females); 3.7 - 5.3 years of age (see Table 2). Criteria for inclusion of the hearing children in this study were: normal hearing, regular attendance in an integrated kindergarten, and no known impairments (placement within the normal range of nonverbal development), as determined by the school psychologist.
Bedrosian (1985) indicates that for this type of study it is necessary to select hearing students who are appropriate participants based on specific criteria. The main criteria for the selection of hearing preschoolers in this research were that all children had to have an age appropriate language comprehension level as well as an age appropriate expressive language level. The Peabody Picture Vocabulary Test-R (PPVT-R) was used as the screening procedure. Hearing participants were included if their PPVT-R age norms were equal to or higher than their chronological age.
The number of participants was restricted to a small group of orally educated preschool children with a profound hearing impairment. Eleven was a relatively large proportion of the overall population of the preschool children with profound hearing impairment in an integrated setting. At the time of the data collection there were only an estimated six additional children with profound hearing impairment integrated in schools throughout Western Australia. Most of those children were in remote country schools. Increasing the sample size was not possible.
Children with impaired hearing were matched with hearing children according to chronological age and gender and data were collected in two contexts: integrated kindergarten and dyad. All matches were made within a range of plus or minus six months. There were eleven pairs of children matched by chronological age. See Table 3 for a detailed list of chronological age matched pairs.
The Social Organization of Discourse Checklist/Conversational Skills (Beattie, 1990) was developed using Roth and Spekman’s (1984) organizational framework. The checklist is based on sound empirical and theoretical literature (Bates, 1976; Day, 1986; Miller, 1978; Roth & Spekman, 1984). A modified version of the Social Organization of Discourse Checklist/Conversational Skills (Beattie, 1990) served the purpose of analyzing social discourse information in a socially rich context.
Detailed definitions of each category, subcategory and sub-subcategory can be found in Table 4. The category of conversational skill, has three subcategories, initiation, maintenance and shift. The subcategory initiation as seven sub-subcategories, percent initiation, initiations auditory, initiations visual, initiations physical, initiations combinations, initiations appropriate and initiations successful. The maintenance subcategory has six sub-subcategories, percent maintenance, significant contribution, minimally contingent response, nonverbal device, appropriate maintenance and successful maintenance. The subcategory shift has three sub-subcategories, percent shift, appropriate shift, successful shift.
The study involved collecting communication samples from two settings. One was in a naturalistic setting (integrated kindergarten) as the children interacted with their peers, and the other was in a quasi-naturalistic setting (tutorial room), which involved the pairing of children. Two different communication settings were chosen since communicative abilities and interactional styles can vary as a function of the communication partner (Ervin-Tripp, 1973). Another reason for two communication settings was the lack of control over the distribution of utterances in the naturalistic kindergarten setting. The quasi-naturalistic setting of the tutorial room helped to balance this.
All communication samples were recorded on video to assist the transcribing of the complex communicative interactions, to allow a multilevel analysis of the same corpus, and to allow the findings to be verified by two coders for inter-reliability purposes. Video recordings occurred over a two-week period beginning the second term of a four-term school year. At the time of videotaping the children had been enrolled in the integrated kindergarten for approximately 12 weeks.
The procedures for collecting the data in the naturalistic setting involved videotaping students on one occasion as they interacted in the integrated kindergarten during ‘free-choice’ activities for 20 minutes. Children were left largely to their own devices and could freely choose their communication partner during this time. They were not encouraged to involve themselves in teacher-directed projects such as art and crafts. Children were not given specific instructions. In fact, none of the children demonstrated an awareness of being filmed. The purpose of this study was to examine the interactions of preschool children with normal hearing and impaired hearing in a naturalistic context.
Two remote-controlled video cameras were mounted near the ceilings of each of the three integrated kindergartens located on the campus of the Centre. A frequency modulated lapel microphone was worn by individual participants. The camera operator sat in a separate video control room near the integrated kindergartens and had visual and auditory access to the information as it was being recorded. This approach allowed social discourse and linguistic performance to take place in an unobtrusive yet meaningful context.
The procedure for collecting the data in the quasi-naturalistic setting with specific pairs of children required each pair to be videotaped alone on one occasion for 10 minutes. Pairing was made randomly with three constraints: each pair included one child with normal hearing and one child with impaired hearing, partners were from the same integrated kindergarten and partners were matched for gender.
Each session was conducted in a small tutorial room adjacent to the integrated kindergarten rooms. Toys known to facilitate social interaction were placed in the tutorial room, as McLoyd, Thomas & Warren (1984) had reported higher levels of interactive behaviors with low-specificity than with high-specificity toys. Lego, a low specificity toy described by McLoyd et al.was chosen and placed on a table in the tutorial room. The two children sat opposite each other at the table with the Lego. A one-way glass mirror on the upper half of one door provided visual access to the room. A video camera mounted on a tripod was located behind the one-way mirror. Sound was recorded on the videotape via a microphone suspended from the tutorial room ceiling. Children were familiar with the tutorial room setting and were instructed that they could play with the Lego on the table while they waited for the teacher.
Editing and Transcribing the Communication Sample
Careful transcription process was utilized to ensure accuracy. The author viewed each videotape three different times. First, the videotapes were viewed and time-coded. Secondly the videotapes were viewed and the transcription of all verbal communication was coded. Finally, the videotapes were viewed for nonverbal behaviors such as gestures and facial expressions that enhanced the understanding of the discourse. The author then coded all transcripts using a modified version of the Social Organization of Discourse Checklist/Conversational Skills (Beattie, 1990).
Two teachers of the deaf, trained in the transcription procedure, served as inter-reliability coders and completed transcription verifications on 15% of all videotapes. All of the transcript verification fell within a range of 85 and 89 percent accuracy.
In the initial stages of the coding procedure, the first coder coded five transcripts on five different occasions. After each coding, a discussion with the first coder led to the refinement of the descriptive criteria for each of the categories and subcategories of the Social Organization of Discourse Checklist/Conversational Skills (Beattie, 1990). A final inter-reliability check was conducted using an additional 10% of the transcripts by the second coder at the conclusion of the coding process.
Both reliability coders participated in an intensive training program designed to teach this coding system. This labor-intensive procedure using two inter-reliability coders was undertaken because of the relatively subjective nature of coding social discourse and the large number of categories and subcategories.
Using Cohen’s Kappa as an index of inter-rater reliability, exceptionally high agreement was found for coding (among the researcher and the first and second inter-rater coders) the Social Organization of Discourse Checklist/Conversational Skills (Beattie, 1990) (k = .83).
Statistical Analyses of Data
A nonparametric procedure was chosen for the analysis of data because of the small sample size and because it was necessary to choose a procedure which did not make strong assumptions about the shape of the distribution of the data. In addition, nonparametric tests work best with frequencies and rank-ordered scales.
The Wilcoxon Signed Rank Test, a nonparametric procedure, was used to determine information about the existence and direction of changes in paired data. It is analogous to the paired t test in that it is used to test the null hypothesis that two samples come from populations which are the same in terms of means. It is used instead of the t test because it requires neither the assumption of normal distribution in each of the two samples nor that of equal variance and can be used with ordinal scale dependent variables. An alpha level of .05 was used for all statistical tests.
Results and Discussion
As mentioned earlier, conversational skills were coded as initiating, maintaining, terminating, and shifting topics. It was expected that the social discourse strategies used by children with impaired hearing would be delayed when compared with chronological age matched children with normal hearing. The overall findings of the study, however, indicate that the social discourse strategies of preschool children with impaired hearing were largely similar to those of their chronological age matched hearing peers. Of the 32 categories and sub-categories analyzed, only eight were significantly different –
- Percent Initiation – integrated kindergarten context
- Initiations Physical – integrated kindergarten context
- Percent Maintenance – integrated kindergarten context
- Significant Contribution (Maintenance) – dyad context
- Minimally Contingent Response (Maintenance) – integrated kindergarten context
- Appropriate Maintenance – dyad context
- Appropriate Shift – dyad context
- Successful Shift – integrated kindergarten context
The conversational skill of initiation is the skill of introducing conversational topics with a communication partner. Initiation was coded as percent initiation, percent auditory initiation, percent visual initiation, percent physical initiation, percent combination of auditory-visual-physical initiation, initiations inappropriate, and successful initiations.
As mentioned, the only two significant differences in the initiation skills were within the integrated kindergarten context. There was a significant difference in the percentage of initiations in the integrated kindergarten context (Appendix A). Of these, a significant proportion were physical initiations in which the children with impaired hearing used nonverbal communicative acts to initiate communication. Interestingly, results indicated that there were no significant differences in auditory, visual or auditory-visual initiation strategies in either the integrated kindergarten context or the dyad context.
There were no significant differences in the percentage of initiations in the dyad context. Initiations in this context were physical (nonverbal) more often than auditory (verbal). Children, both hearing and hearing impaired, tended to wave a Lego, touch or tap the communication partner just before the verbal communication.
The conversational skill of maintenance is the participant’s ability to keep a conversation going. Maintenance is generally dependent upon the contingency of a response to the preceding message (Beattie, 1990). Maintenance was coded as percent maintenance, significant contribution, minimally contingent response, nonverbal devices, appropriate maintenance attempt, and successful maintenance attempt.
There was a significant difference of the percentage of maintenance attempts in the integrated kindergarten setting but not the dyad setting (Appendix B). Hearing children made more overall communication maintenance attempts in the integrated kindergarten. This outcome is most likely because the integrated kindergarten context had more background noise and more potential communication partners. The participants with impaired hearing often appeared unaware of attempted initiations by their peers so did not attempt to maintain the conversation.
A significant difference was found in the number of maintenance attempts coded as making a significant contribution to the communication in the dyad context. The hearing partners in this context made a larger percentage of significant contributions. Maintenance attempts coded as making a significant contribution or minimally contingent response identify the sophistication of the communication partner. A significant contribution is one which maintains the topic and adds new information (Beattie, 1990). The new information may include elaboration of any preceding contribution. A minimally contingent response is a contribution which maintains the topic but does not add new information (Beattie, 1990). As a consequence of a minimally contingent response the conversational responsibility is not necessarily transferred to the partner and may return to the speaker quickly. Minimally contingent responses may be nonverbal devices such as head nods, facilitating expressions or even body posture.
Although the children with impaired hearing attempted to maintain conversation at the same rate as their hearing peers in the dyad context, they used more minimally contingent responses and fewer significant contributions. The children with impaired hearing in this context generally used one or two word phrases to maintain the communication and did not add new information, whereas the hearing children contributed to the communication by adding new information. The hearing children generally took responsibility for maintaining the communication in the integrated kindergarten context, however, those maintenance devices were minimally contingent responses.
There were no significant differences in the amount of non-verbal maintenance attempts in either the integrated kindergarten or the dyad context. The children tended to use nonverbal communication equally in both contexts.
Although the children with impaired hearing attempted to maintain conversation at the same rate as their hearing peers in the dyad context, there was a significant difference in the number of appropriate maintenance attempts. An appropriate maintenance attempt can be seen as one that fits the communication partners, the context and the time (Beattie, 1990). In the dyad context, all of the maintenance attempts of the hearing children were appropriate, whereas only 75% of the maintenance attempts of the children with impaired hearing were appropriate. Children with impaired hearing frequently attempted to maintain the communication with comments that were unrelated to the topic at hand.
There were no significant differences in the category successful maintenance in either context.
The conversational skill of shift is the participant’s ability to change the topic of conversation to a new topic. Shift was coded as percent shift, appropriate shift, successful shift.
Overall results indicate that there were no significant differences found in the number of shift attempts in either the integrated kindergarten or dyad context (Appendix C). Children with hearing and impaired hearing in both contexts attempted to shift the conversation topic at the same rate. However, there was a significant difference in the successful shift attempts in the integrated kindergarten setting only. Hearing children made more overall successful communication shifts. After establishing a topic, the children with impaired hearing were more likely to stay with the initial topic than to shift the topic.
Similarly, there was a significant difference in the number of appropriate shift attempts in the dyad context. Again, hearing children were more successful at shifting conversational topics than their chronological age matched peers with impaired hearing. Finally, there were no differences in the dyad context in successful shift attempts. There were also generally fewer shift attempts in the dyad context. In most cases, the conversation topic focused on the topic of play, Lego.
Interestingly, the ability to shift a conversation topic may be attributed to the typical educational style used with most children with impaired hearing. Adults interacting with children with impaired hearing frequently exhaust one topic before moving on to the next topic. There are many reasons for this, one being the need to expose the child with impaired hearing to vocabulary, sentence structure and pragmatics associated with specific topics of discourse.
The conversational skill of termination is the participant’s ability to end a conversation appropriately.
None of the children used any type of conversational termination skill. As indicated earlier, the context did not require the use of this type of social discourse skill. This was especially true in the dyad context. There was no need to terminate the conversation because the context set the conversation agenda (Lego play). In contrast, the children in the integrated kindergarten context drifted from activity to activity without formal termination of conversation. Interestingly, Keller-Cohen (1977) reports that termination strategies other than merely walking away from the conversation are not learned until much later. When young children have no further comments about the topic they usually stop and do not fill in or tie the topics together (Kretschmer & Kretschmer, 1980).
In-depth analysis of the data suggests that there were some differences in communication between participants that were related to the context. The context related social discourse difference emerged in several predictable ways. In the dyad context there were fewer initiation attempts and a larger number of maintenance attempts. Setting out the bucket of Lego predetermined the topic in the dyad context. One would therefore expect fewer initiations and perhaps more maintenance attempts, which was indeed the case. There were also more shift attempts in the dyad context. All participants had more communication partners in the integrated kindergarten context and were required to initiate or respond to an initiation more and as a result were less likely to shift.
Interestingly, hearing participants in the integrated kindergarten context who did attempt to shift made shifts that were either inappropriate or unsuccessful.
The aim of this research was to examine the social discourse/conversational skills of children with normal hearing and children with impaired hearing in an integrated setting. Kindergarten data were collected from the natural social context of the participants and no structured elicitation was used. As was intended, these optimal conditions existed for the participants to exercise their communication skills. For example, all of the children in both the integrated kindergarten and dyad context engaged in social speech most of the time, and less than one percent of the interactions recorded were nonsocial in nature. This result is a product of the context in which the children were videotaped.
The data collected in the dyad context was only quasi-naturalistic in that pairs of children were randomly chosen, a predetermined context was arranged (Lego toys) and the length of interaction was timed. Eliciting social discourse in this contrived manner may have produced different results and therefore brought a different perspective to the social discourse skills of the children with impaired hearing.
There are obvious benefits from having communication samples that were taken from a naturalistic as well as a quasi-naturalistic context. A combination of the two contexts provided a more comprehensive description of the peer communication within social contexts. It also demonstrated whether the participant’s style of communication was flexible and met the needs of the particular context.
This research has demonstrated that children with impaired hearing in an integrated setting have similar social discourse skills as their chronological age matched hearing peers. Difficulties which were experienced by the children with impaired hearing were due, in most part, to extraneous factors inherent in the hearing impairment and not in the communicative competence of the child.
The results reported in this research must be interpreted in the light of the sample studied and the educational setting in which the participants interacted. The children with impaired hearing, although profoundly deaf, had been immersed in oral language from an early age. Social integration began immediately after diagnosis in the form of playgroup and/or childcare located at the Centre. In addition, the parents and caregivers at the Centre were involved in an intensive language and communication education program. Teachers of the Deaf visited the participants with impaired hearing at home and provided them with regular therapy sessions during school hours. Participants with impaired hearing had daily access to an educational audiologist who insured that hearing aids and earmolds were appropriately fitted and working to the maximal benefit of the child.
This study has looked at the way children with impaired hearing structure conversation through turn-taking, topic initiation and topic maintenance and the ability to control these devices which facilitates cohesion and coherence to conversation. The results of this study identify several potential areas of research. Do children with impaired hearing in segregated educational environments have similar discourse strategies as their integrated hearing peers? Longitudinal study of the social discourse skills of children with impaired hearing may identify the stages of social discourse skill development.
It is also important to determine how adult communication partners can assist in facilitating the development of conversational competence. This involves the examination of child language development and adult interaction. It is well known that before children begin to use formal language they have already learned much about language. From birth children are learning aspects of social discourse, including turn-taking, topic maintenance and conversation termination. As children interact with their peers and adults, they learn to communicate. Often in deaf education adults interact with children with impaired hearing in a way that is different from their interaction with hearing children. A clear understanding of the implications of specific interactional styles will have a bearing on the facilitation of the communication development of oral children with impaired hearing.
It is also important to determine how Teachers of the Deaf can best build on the natural language learning abilities of children with impaired hearing. A clearer understanding of the adult role in communication development will lead to facilitating student learning. If documentation existed which demonstrated the impact of specific styles of communication with children who have impaired hearing it would lead to a better understanding of how learning can be maximized.
Conversational Organization Theory argues that there is a set of rules that governs how social discourse is conducted (Kretschmer & Kretschmer, 1989). These rules include turn-taking, topic initiation, topic maintenance, topic shifting and termination of conversation. Mastery of these rules or organizational features facilitates conversational fluency. Implicit in Conversational Organization Theory is the idea that communicative interactions are social. Clearly, the child’s social discourse abilities must be analyzed in a social context and in relation to those of his communication partners. In this way the communicative strength of the child is emphasized. It must be recognized that each individual has a unique style of communication performance, which is related to intelligence, personality and social experience. This individual and unique characteristic is especially distinct in children with impaired hearing. Educators of the deaf are sometimes overly concerned with language input and often forget that children with impaired hearing are competent individuals who have individual strengths and experiences.
Kretschmer & Kretschmer (1990) state that it is a crucial task for any language learner to acquire a full range of communication functions within one particular language. The profoundly deaf children in this study have demonstrated that with early integration and amplification, communicative competence is a possibility. In sum, the educational environment in which the children with impaired hearing in this study were enrolled provided support necessary to facilitate development of all areas of social discourse. It is likely that the modeling by the normal hearing children provided a positive social influence for the children with impaired hearing.
Source: Volta Review, 2001