2003 IFA Congress: Montreal, Canada

A Preliminary Analysis of Interaction Characteristics Between Chronic Stuttering Japanese Children and Their Mothers

Keiko Gongdo1, Yoko Walaba2, Sumiko Inoue2,3 Megumi Iizawa3, and Hiroshi Fujino4
1Department of Early Childhood Education, St. Margaret's Junior College, 4-29-23, Kugayama, Suginami-ku, Tokyo, 168-8626, Japan
2The Research Institute for the Education of Exceptional Children , Tokyo Gakugei University, 4-1-1 Nukuilfita-machi, Koganei-city, Tokyo, 184-8501, Japan
3R&D Center, PIGEON Corporation, 6-20-4, Kinunodai, Yawara-mura, Tsukuba-gun, Ibaraki-ken, 300-2495, Japan
4Department of Education for Children with Disabilities, F uculty of Education, Tokyo Gakugei University, 4-1-1 NukuiKita-machi, Koganei-city, Tokyo, 184-8501, Japan


The purpose of this study was to compare interaction characteristics between a group of 9-year-old Japanese stuttering children (N :3) and their mothers during free play to those of a control group of nonstuttering children (N :3) and their mothers. A 10-minute play session for each pair was analyzed in terms of a number of utterances, number of topic initiations, MLU, a number of interruptions and conversational functions. Subjects in the stuttering group showed fewer interruptions during the conversation than the subjects in the non-stuttering group. Conversational functions varied among subjects; however, 2 mothers of the stuttering group showed a tendency to control their children’s behaviors.

  1. Introduction
Parents often report their distress and feelings of helplessness when faced with their child’s dysfluency during a clinical assessment (Rustin et al. 1996). Such negative feelings might affect parent-child interaction style in the course of the development of stuttering. Clinically, it is Very common for clinicians to suggest to parents of a stuttering child that they have their child’s linguistic and/or para-linguistic behaviors corrected through the therapeutic process (Kelly, 1995). Linguistic behaviors include length and complexity of utterances, and para-linguistic behaviors include speech rate, turn-taking and the like. Egolf et al. (1972) reported that decreasing the number of negative behaviors parents have toward stuttering children, such as linguistic aggression, silence and interruptions, resulted in fewer dysfluent utterances by the children. Stephenson-Opsal and Ratner (1988) pointed out a relation between the mothers’ slower speech rate and the reduction of dysfluent utterances by stuttering children. As clinical studies have indicated, parent-child interaction seems to play an important role in the process of onset, development, and recovery of stuttering. A number of studies comparing parents of stutterers and parents of non-stutterers have been conducted in terms of linguistic/paralinguistic characteristics of interaction (Kelly & Conture, 1992; Meyers & Freeman 1985), although, the results vary among the studies and no consensus has been reached.

In Japan, it is recognized by clinicians that in order to improve fluency a non-stigmatizing interaction style, as indicated in the studies of English-speal<ing families, is crucial (Wakaba 1999). Unfortunately, few attempts have been made to investigate characteristics of interaction between Japanese stuttering children and their parents (Wakaba, 1999). Therefore, it is necessary to shed light on the research topic by using Japanese subjects. The final goal of this research project is to 244 Theory, research and therapy in fluency disorders identify the interaction characteristics of stuttering children and their parents that might contribute to the onset, development, and recovery of stuttering. As a first step toward this goal, we carried out a preliminary study on this topic with the participation of a group comprising 3 stuttering children and their mothers, and also a control group comprising 3 non-stuttering children and their parents. The following interaction characteristics were analyzed: (1) quantitative adjustment of utterances, (2) assertiveness in a conversation, (3) characteristics of tum-taking, and (4) communicative functions.

  1. Method

Subjects for the study were 3 children (KK, SD, AS ) who chronically stutter and their mothers. The ages of the children were 9;0, 8;3, and 8;8. All mothers were the children’s biological mothers and their ages were 38, 33, and 35, respectively. Based on ITPA scores (PLQ obtained by IT PA Japanese version were 8;6, 8;ll, and 7;2 respectively) and also the fact that parents did not express concerns about their children’s language competencies beyond the stuttering, the children were assumed as having normally-developing language skills. The ages of onset of stuttering were 3;O, 5;7, and around 3 years old, respectively. The levels of stuttering at the onset of the 3 children were very severe according to their mothers. However, the current severity levels of stuttering were 2, 3, and 2 on the Iowa Scale of Severity of Stuttering (Johnson, et al. 1963). The control group consisted of 3 non-stuttering children (CK, HS, SK ) and their mothers. The ages of the children were 8;4, 8,10, and 8;6. All mothers were the children’s biological mothers and their ages were 35, 43, and 32. Based on ITPA scores ( PLQ were 8;4, 8;l0, and 8;6 respectively), the control group children were also evaluated as having normally-developing language skills. All of the children were first-born males and lived with both parents. The children also came from a socio-economical middle class background.

Data collection

Each child and mother was asked to sit at the table face-to-face in a university playroom. They were also asked to play with clay for about 20 minutes. They were given no special instructions except “Do what you normally would do.” The sessions were recorded by DAT and VHS.

Data analysis

Prior to the current study, a 10-minute videotaped segment was transcribed and analyzed according to certain categories, which will be mentioned below, by 2 trained researchers (the first and the third author of the current paper) to verify the analysis’ reliability. Agreement of coding by the two was above acceptable levels of reliability (above 90%). For the current study, the researchers then further transcribed and coded the verbal behaviors of the subjects. The first 10-minute. segment of each videotape was transcribed for verbal behaviors produced by the subjects. Unintelligible utterances were eliminated from the data.

To examine the characteristics of quantitative adjustment of utterances, the number of utterances and MLU (by words and by moras) for each subject were measured. Assertiveness in a conversation was examined by counting the number of topic initiation that each participant (either a child or a mother) made during the 10-minute interaction. Mean number of turns per topic was also obtained for each participant. The number of interruptions was also counted to see turn-taking characteristics. Finally, all Verbal behaviors were coded according to categories such as direction, question, request for clarification, praise, and negative statement. The variables are defined below.

Turns: A string of one or more utterances belonging to a single speaker separated from other utterances by the other speaker or separated from other utterances by him/hers_elf by more than 500 msec.

Topic initiation: How many times each participant (either mother or child) starts a new topic.

Conversational functions

Direction: An utterance that directly controls the conversational partner’s behavior. For example, “Look at me” “Stop it.”

Question: An utterance that requests new information. Both yes/no questions and wh-questions are included in this category. This category also includes requests for clarification: i.e., a speaker asks the conversational partner to clarify his/her utterance.

Praise: An utterance that gives a positive feedback to the conversational partner in an exaggerated manner such as “Good job!” “You are neat!” and “How could you do it!”

Negative statement: An utterance that gives a negative feedback to the conversational partner such as “Can’t you make it better?” or “You are clumsy with clay.”

  1. Results and Discussion
The number of utterances during the 10-minute interaction for each subject is shown in Figure 1. All the mothers of stuttering children tended to produce relatively more utterances than their children did, while the mothers of non-stuttering children varied.


Figure 1. Number of utterances for each participant

Figure 2(a) shows the results of MLU by content words and Figure 2(b) show the results of MLU by moras. The results demonstrated that 2 mothers (KK, SD) of stuttering children tended to produce longer utterances than their children did, while all of the mothers of non-stuttering children produced MLUs that were similar to their children’s.

In the current study, assertiveness in a conversation was examined by the number of_topic initiations that a child and mother made (Figure 3), and also by mean number of turns per topic (Table 1). All the mothers, except AS (stuttering group), showed a tendency to introduce new topics more often into the conversation than their children. Mean number of turns per topic varied from 5.8 to 11.9 and no notable differences were found between groups.

The most interesting result of the current study was the number of interruptions, shown in Figure 4. Although statistic analysis cannot be done because of the limited number of subjects, the non-stuttering children and their mothers produced more interruptions than the stuttering children and their mothers did. One possible explanation is that the stuttering children have been in speech therapy in the past and the mothers were advised not to interrupt while their children were speaking.


Figure 2a. Mean length of utterances (content words) for each participant


Figure 2b. Mean length of utterances (moras) for each participant


Figure 3. Number of topic initiations for each participant


Table 1. Mean and standard deviation of the number of turn per topic for each participant


Figure 4. Number of Interruptions for each participant

Conversational functions of the subjects’ utterances are shown in Figure 5. Conversational functions selected for the current study, such as directions, questions, praises, and negative statements have been recognized as the functions that possibly put the stuttering children’s fluency under pressure. Total frequency of occurrence of the 4 functions were 54.6% (KK), 33.1% (SD), 30.4% (AS) for the mothers of stuttering children and 24.7% (CK), 18.1% (HS), and 31.0% (SK) for the mothers of non-stuttering children. Conversational functions of the subjects’ utterances are shown in Figure 5. The mother of KK showed very directive and negative interaction style toward KK. The mother of SD, on the other hand, used directions only 5 times (3.3% of total utterances) and more praise was used compared to other subjects. Her interaction style seems to be non-directive and positive. However, an excessive use of praise might be a sign of manipulation of children’s behaviors. Therefore, it can be said that both the mother of KK and the mother of SD had a tendency to control their children's behaviors, although the manners in which they did this were quite different.


Figure 5.

The overall results indicated that two mothers of stuttering children did not quantitatively adjust their utterances to their children’s. These mothers also showed a tendency to be directive to the children, while the mother of AS in the stuttering group demonstrated a we1l-adjusted communication style with her child.

Because of the limited number of the subjects reported here, we have to be careful not to generalize the results. However, this study gives us a future direction for research in terms of characteristics of interaction between chronic stuttering Japanese children and their parents.

This study was conducted as a part of a 2-year interdisciplinary research project supported by Japanese government: Grant-in-Aid for Scientific Research, Category B (1) No 14390015 (Title: Study on onset, development, spontaneous recovery, and fundamental therapeutic information of stuttering by interdisciplinary research method). The final goal of the project is to find clues to identify the types of stuttering soon after onset so that clinicians can give a better diagnosis and prospect of treatment. Data collection and analyses are in progress on 40 three-year-old children who recently began to stutter and 20 nine-year-old children who chronically stutter, with the same number of non-stuttering children for tie control groups. Thus, the research question obtained by the study here will continue to be investigated with a larger number of subjects in conjunction with other environmental and biological factors.

Egolf, D.B., Shames, G.H., Johnson, P.R., & Kasparin-Burrelli, A. (1972). The use of parent-child interaction patterns in therapy for young stutterers. Journal of Speech and Hearing Disorders, 37, 222-232.
Johnson, W., Darley, F. L., & Spriestersbach, D. C. (1963). Diagnostic methods in speech pathology. New York: Harper & Row Publishers.
Kelly, E.M. (1995) Parents as partners: Including mothers and fathers in the treatment of children who stutter. Journal of Communication Disorders, 28, 2,93-l05.
Kelly, E. M. & Conture, E. G. (1992) Speaking rates, response time latencies, and interrupting behaviors of young stutterers, nonstutterers, and their mothers. Journal of Speech and Hearing Research 35, 1256- 1267.
Meyers, S. C. & Freeman, F. J. (1985) Are mothers of stutters different? An investigation of social- communicative interactions. Journal of Fluency Disorders, 10, 193-209.
Rustin, L., Botterill, W., & Kelman (1996). Assessment and therapy for young dysfluent children- Family interaction. Whurr Publishers.
Wakaba, Y. (1999) Therapy process of stuttering children with early onset. Unpublished doctoral dissertation, University of Nagoya, Nagoya, Japan.


In preparation for the 2018 World Congress the IFA is implementing Japanese translations of some pages on the site. Choosing Japanese below to see these translations.

Not all pages are translated, but you can use Google translate to see a machine translation using the switch below

Google Translate

Follow the Joint World Congress