Panagos, 1980; Wilcox & Leonard, 1978). Although the general efficacy of language intervention is well documented, the third question, "Which type of.
206 Journal of Speech and Hearing Research
29
206-217
June 1986
Journal of Speech and Hearing Research, Volume 29, 206-217, June 1986
DIRECT LANGUAGE INSTRUCTION AND INTERACTIVE LANGUAGE INSTRUCTION WITH LANGUAGE DELAYED PRESCHOOL CHILDREN: A COMPARISON STUDY PHILIP S. DALE
KEVIN N. COLE
University of Washington, Seattle Direct language instruction and interactive language instruction are based on highly distinct theoretical positions and incorporate very different behavioral techniques. A purpose of the present study was to evaluate the relative effectiveness of direct and interactive language instruction in promoting a variety of language skills. A second purpose was to determine if an aptitude and treatment interaction influenced the outcome of intervention. Forty-four children of preschool age were randomly assigned to classrooms using either direct instruction or interactive instruction. After 8 months, children in both settings improved significantly and substantially on syntactic and semantic measures. There were no differences between the two groups at posttest, nor were there any significant aptitude by treatment interactions for either cognitive or language pretest measures. The results of this study, and a careful examination of previously reported interactions, suggest that such an interaction has yet to be demonstrated.
effective. Numerous studies have shown that languagedisordered children make gains in language development that can be attributed to language intervention, even though procedures involved in intervention programs differ a great deal (Friedman & Friedman, 1980; Gottsleben, Tyak, & Buschini, 1974; Gray & Ryan, 1973; Hegde & Gierut, 1979; Hegde, Noll, & Pecora, 1979; Leonard, 1975; Mulac & Tomlinson, 1977; Prelock & Panagos, 1980; Wilcox & Leonard, 1978). Although the general efficacy of language intervention is well documented, the third question, "Which type of intervention is most effective?", has been addressed less frequently and with much less consistent results. Existing intervention programs have been based on quite different theories as well as on experience with different populations of children. They differ, therefore, on many dimensions, including the amount of strueture in the pro-
The treatment of language disorders in preschool-aged handicapped children is a relatively recent phenomenon. Three basic questions must be addressed in evaluating this innovation: (a) Is intervention necessary? (b) Do effective interventions exist? and (c) What type of intervention is the most effective? The first two questions can be answered affirmatively. Follow-up studies have documented that language-delayed children are much more likely to have academic, social, and linguistic deficits later in life than are children with adequate language development (Aram & Nation, 1980; Cooper & Griffiths, 1978; Hall & Tomblin, 1978; King, Jones, & Lasky, 1982; Silva, 1980; Weiner, 1974). It is apparent from these studies that children with delayed language do not simply "outgrow" the handicap. Some form of intervention is clearly warranted. After determining that intervention is necessary, the next question is whether language facilitation is, in fact, © 1986, American-Speech-Language-Hearing Association
206
0022-4685/86/2902-0206501.00/0
COLE & DALE: Direct and Interactive Language Instruction gram, the degree to which the child is assigned an active, problem-solving role in language learning, and the similarity of the training to everyday speaking and listening situations. During the past three decades, the theoretical bases for intervention programs have changed substantially in response to changing definitions of language by linguists and psycholinguists, and changes in psychological models of language learning (Bloom & Lahey, 1978; McLean, 1983). In the 1960s the introduction of behavioral models stimulated speech pathology away from its primary focus on "speech correction" activities to a concern with a wider variety of language skills and disorders. Behaviorism provided a clearly defined methodology for fostering the development of first grammatical, and later in the decade, semantic skills. In the following decade, the guiding definition of language was broadened still further to emphasize the central role of social interaction and communication of meaning. The inclusion of these factors as language intervention goals had a major impact on the nature of language instruction. Facilitating meaningful communication in a social context requires that the intervention be conducted in an environment that is less structured and less teacher-directed (MUma, 1978). In such an approach, the specific content is also likely to be guided more by the child than by the adult. These divergent theoretical models have sometimes been implemented in very pure forms and have sometimes been combined in more eclectic approaches. Hart and Rogers-Warren (1978), for example, developed a program that incorporated use of aspects of both direct and more interactive models. The use of a natural context was stressed, but structured reinforcement procedures were also incorporated into the program. Friedman and Friedman (1980) and Leonard (1981) provide a taxonomy of current language intervention programs that reflects the two major eras of theory development identified by McLean (1983). The contrasting treatment approaches are identified as "direct" language instruction and "interactive" language instruction. Direct language intervention programs typically use elicited imitation as a teaching technique. The child is required to repeat a complete or partial model of a specific linguistic construct produced by the teacher. These programs also include operant techniques such as structured reinforcement schedules and procedures, chaining, fading, and specific Considerations for carryover of skills to situations beyond the teaching situation (Gottsleben et al., 1974; Gray & Ryan, 1973). A variety of specific language constructs have been successfully taught using a direct instruction format, including pronouns and present progressive -ing (Gottsleben et al., 1974), auxiliary verbs (Hegde et al., 1979), possessive -s (Hegde, 1980), and is interrogative (Mulac & Tomlinsonl 1977). Unlike direct instruction techniques, in which the child is viewed as a passive learner, the interactive approach to language facilitation considers the child to be an active participant in abstracting the rules involved in forming utterances (Leonard, 1981). In general, interac-
207
tive procedures incorporate repeated exposure to the specific forms to be learned in a context where the child can associate the utterance with the events or stimuli in the environment. This exposure is often viewed as "modeling." After exposure, the child is encouraged to use the syntactic rule, but not the particular utterance, Used by the teacher, Table 1 lists the primary contrasting features of the two instructional models.
Program Comparison Studies Both types of programs have been demonstrated to be effective in facilitating syntactic development with language-disordered children. However, there is relatively little comparative research on these approaches that has been conducted independently of the program developers, and that has included random assignment of subjects to treatments. A few studies have evaluated the effectiveness of specific aspects of these programs, and still fewer have compared the programs as they are actually implemented, as complete packages. The results have not been unequivocal. Courtright and Courtright (1976) compared two treatment approaches based on elicited imitation and modeling, respectively, for their ability to facilitate the use of the pronoun they. The 4 subjects in the modeling treatment were presented with 20 stimulus pictures paired with the spoken model "they are [nouns]." Then they were required to respond to each of the pictures without further modeling. The imitation group of 4 subjects was given the added stimulus of "What are they?," and were required to respond. Analyses revealed that the modeling group started with fewer correct responses but ended with more correct responses than did the imitative group on a generalization test. The authors concluded that the use of modeling was a more effective instructional technique, and that the teaching generalized to a greater degree than with an elicited imitation technique. A later study by Courtright and Courtright (1979) also compared modeling with elicited imitation using a different language structure. The subjects were taught a nonsense grammatical form "[noun] + means + to + [verb]ing" to eliminate the possibility of an pretest differences between groups for the target structure. Modeling again was found to be superior to imitation in teaching the construct. The results of the two Courtright and Courtright studies suggest that modeling is a more effective teaching method. However, the implications of these studies of specific teaching processes for direct and interactive programs as a whole are limited, because neither the elicited imitation technique nor the modeling technique used were entirely representative of direct and interactive approaches, respectively. Imitation in direct instruction programs is typically faded to a partial model and cue once the child can perform the complete imitation successfully. In the two Courtright and Courtright studies no fading was done. The subject always received the entire
208 Journal of Speech and Hearing Research
29 206-217
June 1986
TABLE 1. Contrasting features of direct language~instruction and interactive language programs.
Interactive instruction
Direct instruction
Individualized language goals.
Predetermined language goals.
No structured reinforcement procedure.
Specific reinforcement procedures.
Modeling is a primary teaching method.
Elicited imitation is a primary teaching method.
Emphasis on natural context for language learning.
Stimulus materials are included in program, with little emphasis placed on natural context.
The child is encouraged to initiate language production.
Instruction is primarily teacher directed.
Response rate is generally low.
Response rate is high.
Structure and sequencing of instruction is variable.
Structure and sequencing of instruction is predetermined.
The child is viewed as actively involved in the process of abstracting linguistic rules.
The child is viewed primarily as a respondent to operant techniques,
Sequencing is often based on normal language development.
Sequencing is often based on operant principles of ease of instruction.
Note. Adapted from Weller (1979), and Siegel and Spradlin (1982).
target sentence to repeat. The failure to use fading as part of the treatment could account for less generalization of the construct taught. In the modeling conditions, the example structures were presented in a didactic, rather than interactional, sequence, and the children were expected to produce the identical forms in response to the same stimuli; rather than similar forms in new contexts. As a result, these two conditions were somewhat different from the programs they purport to exemplify. This qualification holds for several other studies that have attempted to compare these techniques. Connell, Gardner-Gletty, Dejewski, and Parks-Reinick (1981) attempted to replicate the results of Courtright and Courtright (1979) with 10 language-delayed children. They found that none of the children used the target form after receiving over twice as much exposure to the modeling technique used in the Courtright and Courtright study. Prelock and Panagos (1980) compared the elicited imitation and modeling techniques used by Court_right and Courtright (1976, 1979) with trainably mentally retarded children with an average MLU of 2.11. The modeling technique resulted in the significantly higher use of action-agent-object forms in free speech samples than did the imitation technique. It also facilitated more frequent use of lexical restructuring of the target form, rather than simple duplications of the forms involved in training. However, because this study used the same elicited imitation technique incorporated by Courtright and Courtright, the same caution should be used in generalizing these results to the use of programs that incorporate
a different form of elicited imitation. Thus previous research has not shown a clear superiority of one approach over the other. Weller (1979) compared more complete implementations of direct instruction and interactive techniques. She compared a language development program based on the techniques used by Bereiter and Englemann (1966) versus an interactive approach outlined by Blank (1973). In a single-subject research design, each of 4 subjects received each type of intervention for either functor or descriptor words. The results of the interventions were reported as changes in frequency counts from week to week over the training period. No inferential statistics were used. The direct-instruction subjects used the language constructs more during instruction and during free speech sessions with the teacher. The time allotted to the treatments differed, however, with the direct instruction involving 20-minute group sessions, and the interactive sessions lasting 15 minutes with individual instruction. Because engaged time is a critical factor in teaching effectiveness (Denham & Lieberman, 1980), this difference could account for the results. Friedman and Friedman (1980) compared direct and interactive programs in a study with a larger sample, assigned to the two conditions. Their findings suggest a different answer to the question, "Which type of program is more effective?" Unlike previous investigators, they found that neither program was superior to the other overall. Instead, they found evidence of aptitude by treatment interactions between program types and several child characteristics. A wide variety of systematic
COLE & DALE: D i r e c t a n d I n t e r a c t i v e L a n g u a g e I n s t r u c t i o n
structures were presented to 41 preschool children using either the Interactive Language Development Teaching approach (Lee, Koenigsknecht, & Mulhern, 1975) or the Programmed Language Presentation approach (Koenigsknecht & Lee, 1974). The programs were equally effective in increasing the use of target structures in a free-speech sample. However, a disordinal aptitude by treatment interaction was found for the factors of IQ, degree of syntactic development, and visual-motor ability. Children with high IQ, relatively high syntax production, and good visual-motor performance did significantly better with the interactive instruction; whereas the converse was true for children with lower IQ, very low initial syntax skills, and poor visual-motor integration. The outcomes of this important and widely cited study must be viewed with caution for several reasons. First, the study design did not include random assignment, and the criteria used for selection are not clearly stated. In addition to study design and subject selection, the results themselves are problematic in interpretation. Although significant disordinal aptitude by treatment interactions (ATIs) were found between the three pretest measures of Stanford-Binet IQ score (Terman & Merrill, 1960), initial Developmental Sentence Scoring score (DSS, Lee [1974]), and Developmental Test of Visual-Motor Integration score (Beery & Buktenica, 1967), and tile gain scores on the DSS, there were 10 additional pretest measures that did not yield an ATI. Most of these measured language characteristics that were very similar to those measured by the DSS. It is surprising that the Northwestern Syntax Screening Test-Expressive (NSSTE, Lee [1971])~ for example, did not interact with the DSS gain scores in the same manner as did the DSS pretest, because both tests measure expressive syntactic development. A similar question can be raised for the Peabody Picture Vocabulary Test (PPVT, Dunn [1959]), the Illinois Test of Psycholinguistic Abilities (Kirk, McCarthy & Kirk, 1968), the Northwestern Syntax Screening TestReceptive (NSST-R, Lee [1971]), and others. Friedman and Friedman (1980) do not report whether the regression slopes for the other 10 aptitude measures were in the same direction as the 3 that were found to be significant. If the slopes were in the same direction, then they would offer some supporting evidence for the hypothesis that children with different aptitudes may benefit differently from the two types of instruction; otherwise the three significant ATIs might be the result of chance. Research on the relative effectiveness of the two contrastive types of language programs is far from conclusive. It does suggest that it is important to ask both of the following questions: 1. Is either program more effective in promoting generalized use of language skills including syntax, semantics, and pragmatics? 2. Is there an aptitude by treatment interaction for the factors of intelligence or level of language development? The purpose of the present study was to examine the effectiveness of the two intervention techniques with language-delayed children.
209
METHOD
Subjects
Forty-four preschool children (34 boys and 10 girls) with delayed language served as subjects. The children, who ranged in age from 38 mos to 69 mos (M = 54 mos), were randomly assigned to classrooms using either direct instruction or interactive instruction in the Experimental Education Unit located at the University of Washington. Language delay was defined as a score of 1.5 standard deviations below the mean for the child's chronological age on either the Peabody Picture Vocabulary TestRevised (PPVT-R), Form L (Dunn & Dunn, 1981); the Northwestern Syntax Screening Test, Receptive section (NSST-R, Lee [1971]); the Preschool Language Scale (PLS, Zimmennan, Steiner, & Pond [1979]), Verbal Ability or Auditory Comprehension subtests; or a Developmental Sentence Score (DDS, Lee [1974]) one full point below the 10th percentile. Standard deviation scores are not available for the DSS, so the above criterion for delay, recommended by Lee (1974), was used. No significant differences were found between groups on any of these measures (see Table 2). As is shown in Table 3, there was substantial variation in IQ, as measured by the Columbia Mental Maturity Scale (CMMS-ADJ range 52 to 109, Burgemeister, Blum, & Lorge [1972]).
Measures
A battery of language measures, including standardized instruments and language sample analyses, was administered in a pretest-posttest design. Testers were not in-
TABLE 2. Pretest selection measures. Direct instruction (n = 19)
Interactive instruction (n = 25)
17 2
17 8
Boys Girls C.A. (too) PPVT-R (raw) (mo) PLS-A (raw) (mo) PLS-V (raw) (too) NSST-R (raw) DSS scorablea DSS (raw)
M
SD
M
52.16 22.47 35.83 20.47 42.71 17.00 37.50 17.58
8.50 12.21 7.55 8.31 12.46 8.82 13.23 6.43
54.8'0 21.56 35.25 18.42 39.63 15.13 34.68 14.32
.91
4.99
8 5.55
, SD
9.06 13.65 8.98 7.71 11.57 8.30 12.44 8.31 11 1.07
aOnly subjects who produced complete sentences for 50% of the utterances in a language sample and produced at least 25 complete sentences were scored on the DSS.
210 Journal of Speech and Hearing Research
29
TABLE 3. Pretest measures for direct instruction and interactive instruction groups.
Test
MLU DSS PLS-AC PLS-VA PPVT-R BLCT NSST-R NSST-E CMMS CMMS-ADS CAVAT SR CAVAT AB Combined
Direct instruction (n = 19) M SD
Interactive instruction (n = 25) M SD
2.62 1.19 2.69 1.13 5.55 .91 4.99 1.07 20.47 8.31 18.42 7.71 17.00 8.82 15.13 8.30 22.47 12.20 2 1 . 5 6 13.70 21.42 17.60 15.76 15.10 17.58 6.43 14.32 8.31 6.47 7.15 5.40 8.50 19.00 7.94 22.00 7.54 8 7 . 9 0 12.00 8 3 . 8 6 11.10 6.37 7.35 5.21 6.98 2.47 2.65 2.80 '3.67 .11 .84 -.08 .84
t
p
.20 -1.24 -.83 -.71 -.23 -1.12 -1.47 -.45 -1.23 -1.11 -.53 .33 -.76
.84 .23 .41 .48 .82 .27 .15 .65 .23 .27 .60 .74 .81
Note. Raw scores were used for all measures except the CMMSADS, which is similar to an intelligence quotient score (M = 100, SD = 15), and the combined language measure, which was reported as a z score.
formed of the research hypotheses, and at posttest had no access to pretest data analyses. During a 2-week period in October and November, 1982, the following pretests were administered: Columbia Mental Maturity Scale (Burgemeister, Blum, & Lorge, 1972). Carrow Auditory-Visual Abilities Test (CarrowWoolfolk, 1981). Peabody Picture Vocabulary Test-Revised, Form L (Dunn & Dunn, 1981). Basic Language Concepts Test (Englemann, Ross, & Bingham, 1982). Preschool Language Scale (Zimmerman et al., 1979). Northwestern Syntax Screening Test (Lee, 1971). Language samples. The pretest language samples were gathered in the classroom. The examiner interacted with 2 children at a time, using stimulus materials consisting of a play house, dolls, and stimulus pictures. The language sampling sessions lasted 30 min. The sessions were recorded on audio tape and transcribed following the sessions. Two measures were derived from pretest language transcriptions. Mean length of utterance (MLU) was determined using Miller's (1981) guidelines for computation of MLU. Developmental Sentence Scoring analysis (DSS, Lee [1974]) was also computed for the language samples that were large enough to allow the analysis. Posttesting was conducted in June and July, 1983. Only 1 subject, rather than 2, was included in each languagesampling session during posttesting. This change was made as a result of a preliminary examination of the pretest transcripts, some of which contained too few utterances to allow a thorough analysis. Sampling with 1 child increased the possibility of getting an adequate sample of language behavior from each child.
206-217
June 1986
The posttest language samples were analyzed for two pragmatic measures in addition to the DSS and MLU measures. Conversational act categories. Dore's (1978) speech act taxonomy, as modified by Chapman (1981), consists of eight major categories of speech acts including the use of requests, responses to requests, descriptions, statements, acknowledgments, organization devices, performatives, and miscellaneous (see Appendix). Within each of these categories, subcategories provide additional information specific to the category. The category of "descriptions," for instance, contains the subcategories of labeling for identification, events and actions, description of traits or properties, description of location or direction, and descriptions involving time. One hundred consecutive utterances from language samples were analyzed to determine the variety of speech acts used. The two categories of "no answer" and "uninterpretable" were not included in the count of different speech acts used because they, unlike the other categories, did not represent specific uses of language. The Dore taxonomy includes 39 classifications of speech acts. Excluding the two classifications above, the maximum number of categories a child could use in the language sample was 37. Topic continuation assessment. The topic continuation measurement developed by Bloom, Rocissano, and Hood (1976) classifies child utterances that follow another speaker's previous utterance into three categories. The three categories consist of (a) child responses that maintain the same topic and include additional information, (b) responses that maintain the same topic without adding new information (imitative responses), and (c) responses that do not maintain the previous topic. All utterances that followed another speaker's utterance in the language sample were analyzed to determine the proportion of each of the three categories.
Intervention Direct language instruction. The direct language instruction program used DISTAR Language I (Englemann & Osborn, 1976) materials in a small group format. Program placement was determined by the placement tests incorporated in the program materials. The DISTAR language program is based on the presentation of predetermined language instruction formats. The program teaches syntactic, semantic, and pragmatic skills believed by the authors to be necessary for success in school. The materials are organized in a sequence designed to provide maximum benefit from instruction rather than in a developmental sequence. The instruction was teacher-initiated, with the child responding to prompts from the teacher to imitate an utterance or produce an appropriate response to a stimulus. Verbal praise was provided when the child responded correctly. A predetermined sequence of teaching activities, based directly on published materials, was followed for all children within an instruction group. The groups con-
COLE & DALE: Direct and Interactive Language Instruction sisted of 2 to 4 children, with one teacher presenting the materials. Interactive language instruction. The interactive language instruction program used procedures and activities from a variety of sources, including McLean and SnyderMcLean (1978) and Rieke, Lynch, and Soltman (1977). Language instruction was provided primarily within an environmerital context that was designed to increase opportunity for language use involving child-initiated communicative intent. Language goals were individually determined by the classroom teacher and a speech-language pathologist (SLP). Every child was assigned individual language goals in syntax, semantics, and pragmatics based on the student's language assessment. These goals were incorporated into all classroom activities, rather than only during a predetermined language instruction time. Primary emphasis, however, was placed on language training during snack time and social-interaction time. A child working on requesting, for example, would be presented with situations during both snack time and free play where desired objects were placed within sight but out of reach. Lack of appropriate or adequate materials might also be used to facilitate childinitiated requesting. For example, a table of 4 children might be presented with three cups during snack time. Each teacher was aware of the individual goals for the children, and the goals were posted on large signs around the room so they could be referred to easily by any of the teachers when interacting with any child. The use of natural consequences for language behavior was also incorporated in the instruction. When the child asked for a cup the teacher responded by giving the cup, and expanding the child's request or making other appropriate responses. Verbal praise (e.g., "I like the way you are talking") was avoided in favor of more natural responses to communication attempts by the children. Within this language instruction program, the child was provided opportunities for communication at various times and in a variety of settings within the classroom. Instruction took place within a conversational setting, rather than a didactic one, and modeling was used to facilitate language development. Attempts to elicit responses directly (e.g., "Tell me the name of this animal."), or elicit direct imitation (e.g., "Say, 'I want a drink of juice.' ") were avoided.
Classrooms
Students attended class 2 hours daily, 5 days a week, for 32 weeks. Two classes used direct instruction and three classes implemented interactive-language instruction. The two direct-instruction classes were taught by the same staff team consisting of a head teacher and two assistant teachers. Two of the interactive classes were taught by one staff team, with the remaining interactive class taught by a third team. The student/teacher ratio was approximately 4 to 1 for all classrooms.
211
Services were provided by speech-language pathologists both within the classroom and in clinical settings outside the classroom. Approximately 80% of the SLP instruction was provided in the classroom. The clinicians followed the instructional models for each classroom.
Procedural Validity Procedural validity data were gathered to document that the two programs were implemented correctly and differed along observable dimensions. Observation system. Data were gathered during 40 30-min observations using a modified version of the Student-Level Observation of Beginning Reading (SOBR) (Leinhardt & Seewald, 1981). Seven factors were selected to measure the contrastive salient features of direct instruction and interactive instruction. All measures were computed as the percentage of time of occurrence of the specific categories of activities during the observation. 1. Involvement in preacademic training: This includes direct-language instruction, concept training, reading, and fine-motor instruction, with the teacher controlling activities. This factor is predicted to be higher for the direct-instruction program. 2. Involvement in social interaction time: This includes free-choice activities such as housekeeping play, sand and water tables, art and craft projects, blocks, and other toys. The teacher may or may not direct the activities. This factor is predicted to be higher for the interactive program. 3. Elicited responding: This refers to responses by the child to the teacher's question, direction, or command. The child's responses may be either verbal or nonverbal. This factor is predicted to be higher for the direct-instruction program. 4. Conversational talking: This refers to talking to the teacher or peers that is either child-initiated or is a response to a previously modeled or expanded utterance (excluding elicited responses). This factor is predicted to be higher for the interactive program. 5. Printed/programmed stimulus materials: If a child's utterance for the categories of elicited responding or conversational talking was facilitated by printed/programmed materials, the information was recorded under this category. This factor is predicted to be higher for the direct-instruction program. 6. Stimulus environmental materials: I f a child's utterance for the categories of elicited responding or conversational talking was facilitated by toys, snacks, art materials, or other classroom stimuli, it was recorded in this category. This factor is predicted to be higher for the interactive program. 7. Elicited imitation: For each elicited response (category 3 above), those that were elicited as a direct imitation of the teacher's stimulus were noted. This factor is predicted to be higher for the direct-instruction program.
212 Journal of Speech and Hearing Research
29 206-217
TABLE 4. Pretest and posttest raw scores for direct language instruction group.
Pretest SD
Test
M
MLU DSS PLS-AC PLS-VA PLS-Overall BLCT NSST-R NSST-E PPVT-R
2.62 5.55 20.47 17.00 37.47 21.42 17.58 6.47 22.47
1.19 .91 8.31 8.82 16.10 17.60 6.43 7.15 12.20
June 1986
TABLE5. Pretest and posttest raw scores for interactive language instruction group.
M
Posttest SD
t
Test
M
3.42 5.48 27.90 25.90 53.79 36.16 20.84 13.16 35.95
1.21 1.22 7.44 9.57 16.10 20.30 7.68 10.20 13.60
3.56** .86 8.37** 9.63** 13.03"* 4.24** 2.32* 4.57** 8.28**
MLU DSS PLS-AC PLS-VA PLS-Overall BLCT NSST-R NSST-E PPVT-R
2.69 4.99 18.42 15.13 33.54 15.76 14.32 5.40 21.56
Pretest SD 1.13 1.07 7.71 8.30 15.50 15.10 8.31 8.50 13.70
M
Posttest SD
t
3.26 5.80 25.17 22.58 47.75 28.29 16.96 8.54 32.71
1.36 1.41 8.45 10.10 17.80 16.20 8.39 9.65 15.20
3.36** 3.01" 6.90** 5.21"* 7.09** 8.97** 2.65** 3.20** 7.11"*
*p < .05. **p < .01.
*p < .05.**p < .01.
RESULTS
Procedural Validity To demonstrate observable differences between the two programs, observational data for seven primary factors (teacher-directed instructional activities, social-interaction time, elicited responding, conversational talking, printed/programmed materials, environmental materials, and elicited imitation) were examined using a t test for the combined measures. The measures were reflected if necessary (by subtracting the score from 100) so that a high score always indicated a characteristic that should be more frequent for the direct instruction class and less frequent for the interactive program. The mean-combined measure for the direct-instruction group was 47.62 (SD = 3.29), and for the interactive group, 36.59 (SD = 2.96). A t test confirmed a highly significant difference [t(57) = 13.36, df = 57, p < .001] between the two programs in the direction predicted by the SOBR. The seven factors were also examined individually. In every case, the difference was statistically significant and was in the predicted direction.
Pretest Analyses Means and standard deviations by group for each of the pretest measures are shown in Table 3. Two-tailed t tests revealed no significant differences between the treatment groups on any variable. Two pretest measures were used to determine the subjects' aptitudes that were hypothesized to influence the effect of the two treatment programs. The two measures were intelligence, as assessed by the Columbia Mental Maturity Scale (CMMS), and the pretest level of language development. A combined measure of pretest language development was computed by converting the raw scores for the following seven language measures into z scores and averaging them:
1. The Peabody Picture Vocabulary Test-Revised, form L; 2. The Receptive subtest of the Northwestern Syntax Screening Test; 3. The Expressive subtest of the Northwestern Syntax Screening Test; 4. The Auditory Comprehension subtest of the Preschool Language Scale; 5. The Verbal Abilities subtest of the Preschool Language Scale; 6. The Basic Language Concepts Test; and 7. Mean Length of Utterance. Descriptive and comparative statistics for this measure are included in Table 3.
Effect of Intervention Significant differences were found between pretest and posttest scores for both groups on every measure except DSS, which had a substantially smaller number of cases. See Table 4 for the direct-instruction subjects, and Table 5 for the interactive-instruction subjects.
Aptitude by Treatment Regression Analyses Following Pedhazur (1982), a regression analysis was performed for each of the two aptitude measures. The treatment groups were dummy coded to allow comparisons between the two groups for both main effects and interactions. The two aptitude measures were the Columbia Mental Maturity Scale-Adjusted DeViation Score and the combined language measure defined above. The Developmental Sentence Scoring measure was not included in the regression analyses because only 27 of the subjects (12 from the direct-instruction group, and 15 from the interactive group) produced enough complete sentences to allow computation of the posttest DSS. Tables 6 and 7 present the results of these analyses. There were no main effects for treatment, a few main effects for the aptitude measures (particularly for the combined-pretest-language measure), and no aptitude by treatment interactions. It is also possible to examine the
COLE & DALE: Direct and Interactive Language Instruction
213
TABLE 6. Regression analyses for posttest scores regressed on Columbia Mental Maturity Scale-Adjusted Deviation Score.
Measure
MLU PLS-AC PLS-VA BLCT NSST-R NSST-E PPVT-R Conversational Act Categories Topic Maintenance Measure A Topic Maintenance Measure B Topic Maintenance Measure C
Regression coefficients DI INT
Treatment F p
CMMS-ADS F p
Interaction F p
0.1 0.1 0.2 0.8 0.7 1.2 0.1
.99 .81 .67 .39 .41 .29 .97
5.3 7.6 2.9 10.5 3.0 4.6 3.5
.03 .01 .1 .002 .08 .04 .07
2.1 1.3 1.2 0.8 1.1 0.4 0.7
.16 .25 .28 .37 .30 .55 .41
.01 .14 .07 .52 .07 .20 .19
.06 .33 .34 .92 .28 .37 .50
+.05 +.19 +.27 + .40 +.21 +.17 +.31
0.1
.83
0.1
.81
0.4
.53
.05
-.01
-.06
2.5
.12
4.1
.05
0.4
.54
.81
.43
-.38
1.9
.18
0.3
.57
0.1
.93
-.04
-.06
-.02
0.2
.68
0.2
.68
0.5
.49
-.04
.17
+.21
Difference
TABLE 7. Regression analyses for posttest scores regressed on combined language measure.
Measure
MLU PLS-AC PLS-VA BLCT NSST-R NSST-E PPVT-R Conversational Act Categories Topic Maintenance Measure A Topic Maintenance Measure B Topic Maintenance Measure C
Regression coefficients DI INT
Treatment F p
CMMS-ADS F p
Interaction F 19
0.4 0.9 0.9 2.1 2.1 2.3 0.6
.53 .36 .34 .16 .16 .14 .46
26.6 131.8 121.3 79.1 82.2 65.9 91.4
.001 .001 .001 .001 .001 .001 .001
0.1 0.6 0.4 0.7 0.17 0.1 0.6
.85 .43 .54 .41 .68 .73 .45
1.06 8.02 10.74 19.83 7.42 9.87 13.34
.99 9.24 9.95 16.43 8.15 9.06 15.70
-.07 + 1.22 -1.15 -3.40 +.73 -.81 +2.36
6.6
.43
6.6
.02
0.1
.71
1.14
1.54
+.40
4.2
.05
13.5
.001
0.4
.53
17.41
12.22
-5.19
3.1
.09
17.7
.001
3.2
.09
-2.65
-6.27
-3.72
0.1
.82
0.4
.51
0.6
.44
-.29
3.16
+3.45
d i r e c t i o n of results for each analysis to d e t e r m i n e if there is a c o n s i s t e n t trend. F o r this p u r p o s e , the critical feature o f the r e g r e s s i o n e q u a t i o n s is the difference b e t w e e n the slopes (i.e., r e g r e s s i o n coefficients) for the a p t i t u d e measures b e t w e e n t h e two t r e a t m e n t groups. In particular, it m a y b e h y p o t h e s i z e d that the regression coefficients s h o u l d b e c o n s i s t e n t l y m o r e positive for the interactive groups than for the direct-instruction groups, if direct instruction is m o r e facilitative for l o w e r ability subjects. T h e r i g h t m o s t c o l u m n s of T a b l e s 6 and 7 p r e s e n t the a p p r o p r i a t e r e g r e s s i o n coefficients and the difference b e t w e e n them. A p o s i t i v e difference reflects an interaction t r e n d c o n s i s t e n t with the findings of F r i e d m a n and F r i e d m a n (1980). F o r the C o l u m b i a M e n t a l Maturity S c a l e - A d j u s t e d Deviation Score, 8 o f 11 r e g r e s s i o n slope differences w e r e in the p r e d i c t e d direction, a n d for the c o m b i n e d language m e a s u r e , 5 o f 11 w e r e in the p r e d i c t e d direction. Overall, 13 of 22 r e g r e s s i o n slope differences w e r e in the pred i c t e d direction.
Difference
DISCUSSION
I n the p r e s e n t study, t h e r e was little difference bet w e e n the effectiveness of a direct-instruction program and an interactive p r o g r a m in facilitating l a n g u a g e d e v e l o p m e n t in l a n g u a g e - d e l a y e d children. D e s p i t e the substantial differences in program content, confirmed by observational analysis, t h e r e w e r e no significant differences in m e a n posttest scores, nor w e r e t h e r e any significant a p t i t u d e b y t r e a t m e n t interactions. W i t h the exception o f DSS, w h i c h was restricted to a small s u b s a m p l e , both programs r e s u l t e d in significant c h a n g e from p r e t e s t to posttest on all measures. T h e p r i m a r y focus of this study was to d e t e r m i n e the relative effectiveness of the two programs. No u n t r e a t e d control group was i n c l u d e d to e v a l u a t e the a b s o l u t e effect of the program relative to maturational change. H o w e v e r , an i n d i r e c t e v a l u a t i o n is p o s s i b l e u s i n g the d e v e l o p m e n tal quotients that are available for two of the tests:
214 Journal of Speech and Hearing Research
29
206-217
June 1986
TABLE8. Comparison of subject characteristics between Friedman and Friedman (1980) and the present study. Present study Direct Interactive N CA NSST-E NSST-R PPVT IQ Subjects scorable on DSS DSS
19 52.2 mo 6.47 (7.51) 17.58 (6.43) 22.47 (12.20) 87.89 (12.00) 8 5.55 (0.91)
25 54.8 mo 5.40 (8.50) 14.32 (8.31) 21.56(13.70) 83.86(11.10) 11 4.99 (1.07)
Friedman and Friedman (1980) Direct Interactive 16 51.5 mo 9.33 (6.68) 22.50 (5.89) 43.75 (6.34) 102.00(15.49)
25 52.3 mo 5.24 (4.68) 21.76 (5.71) 43.12 (12.62) 99.88 (13.97)
16 4.75 (2.07)
25 3.84 (1.53)
Note. Raw scores were used for all measures except IQ. Standard deviations (SD) are given in parentheses. Preschool Language Scale (PLS) Language Quotients and Peabody Picture Vocabulary Test-Revised (PPVT-R) Standard Scores. These quotients are independent of chronological age, and therefore of maturation. For the PLS, Language Quotients increased significantly from 74.5 to 82.0 It(41) = 4.9, p < .001]; and for the PPVT-R, Standard Scores increased nonsignificantly from 71.7 to 74.1 [t(35) = .93, p < .4]. Thus there is some evidence that performance on language tests improved more than would be predicted by maturation alone. The lack of significant differences between directlanguage instruction and interactive-language instruction contrasts with the findings of Courtright and Courtright (1976, 1979) and Prelock and Panagos (1980). These authors found one component of interactive instruction (modeling) to be more effective than one component of direct instruction (elicited imitation) in promoting syntactic skills in language-delayed children. In the present study, a comparison was made of the effect of complete language-development programs, rather than of the components of programs implemented in a manner that is substantially different from the way in which actual programs use the components. Weller (1979) did compare complete programs and concluded that the direct-instruction program was superior to the interactive-instruction program in facilitating language production in preschool children. Weller, however, did not subject her data to probability analysis, therefore it is not clear if the differences found in her study were statistically significant. In addition, the time in treatment differed by 25% between the two types of programs, with direct instruction being given more time in treatment. This difference may have influenced the outcome of the study. Friedman and Friedman (1980) also found no significant differences between treatment groups; as in the present study, both groups made significant gains from pretest to posttest. Friedman and Friedman did find several significant aptitude by treatment interactions. However, when two of the aptitude variables identified by Friedman and Friedman (IQ and level of language development) were examined in the present study, no significant interactions were found. It should be noted
that the direction of regression slope differences for the interactions between IQ and the syntactic and semantic measures (though not the pragmatic measures) were consistently in the hypothesized direction, though none attained statistical significance. However, such a trend was not found for the aptitude measure of pretest language performance. The differences in results between the two studies may be due to differences in subject characteristics, use of random assignment, and/or data analysis. Subject differences. The number and ages of the subjects in this study and in the Friedman and Friedman (1980) study were very similar (see Table 8). The subjects differed in two important ways, however. The subjects in the present study averaged 15 points lower in IQ than the Friedman and Friedman subjects. In addition, the subjects in the present study were delayed in a variety of language areas, both expressively and receptively. The Friedman and Friedman subjects, in contrast, performed within normal limits on the Peabody Picture Vocabulary Test (Dunn, 1959) and the Illinois Test of Psycholinguistic Abilities (Kirk, McCarthy, & Kirk, 1968). Their subjects scored an average of 1 year below the mean on the Expressive subtest of the Northwestern Syntax Screening Test (Lee, 1971). This is within the range of low normal (25th percentile). Perhaps aptitude by treatment interactions are found only with subjects who are delayed solely in the area of expressive syntactic development but not for children who are more globally delayed in language functioning. It is also possible that the aptitude by treatment interactions found by Friedman and Friedman only occur for children with mild to moderate delays, rather than for children with more severe language impairments. Intellectual development may also be relevant. The mean cognitive level of the Friedman and Friedman (1980) subjects was in the normal range. The mean cognitive level of the subjects in the present study was approximately one standard deviation below normal. Perhaps the interaction between IQ and treatment in the Friedman and Friedman study is found only for subjects within one standard deviation of normal cognitive development. Subjects within that range might be more responsive to differences in programs than are children with less-well-developed cognitive skills.
COLE & DALE: Direct and Interactive Language Instruction
Assignmen~ differences. In the present study, subjects were randomly assigned to the two treatment groups. In contrast, Friedman and Friedman (1980) assigned subjects to treatment groups based on their availability during scheduled treatment times and on their position on a treatment waiting list. Although there were no significant differences in pretest scores between the two treatment groups in the Friedman and Friedman study, the mean pretest DSS score of the direct-instruction group was 19% higher than the mean pretest DSS score for the interactive group. Analysis differences. Both studies used regression analyses to examine the interaction between treatment and aptitude variables, but they differ in how the regression analyses were applied. Friedman and Friedman (1980) examined all pretest measures available for the subjects in a post hoc search for interactions. The present study was prospective, examining only two aptitude variables, IQ and language development, which were selected in part because of the earlier significant findings. Friedman and Friedman examined 13 aptitude variables and found three that produced significant disordinal aptitude by treatment interactions with DSS gain scores. That 10 aptitude measures did not produce significant ATIs raises an important question. Why did measures that were very similar to the DSS, such as the expressive subtest of the NSST, fail to produce significant ATIs with the aptitude measures used? A consistent direction of regression slope differences for these 10 measures would support the hypothesis of aptitude by treatment interactions. Unfortunately, Friedman and Friedman do not report the direction of regression slopes for the 10 nonsignificant outcome measures. In the present study, the direction of all slopes resulting from aptitude by treatment interactions was examined, and only 14 out of 22 differences were in the hypothesized direction. Thus the finding of significant ATIs may be a result of chance variation.
CONCLUSIONS Direct-language instruction and interactive-instruction appear to have comparable effects in facilitating language skills in language-handicapped preschool children. Neither program appears to be superior overall, for either specific aspects of language development or for subgroups of children identified by degree of language delay or cognitive ability. The apparent success of these two programs is likely to be due at least in part to the fact that both were well-staffed, organized around coherent theoretical positions, and carefully monitored. It is possible that these factors are more influential than is the model underlying an early intervention program. However, it is also possible that a merging of the two models would be more effective than the exclusive use of either. A concept or language structure might be presented initially using direct instruction for rapid acquisition and later extended using interactive methods for generalization to a variety of contexts. Alternatively, introduction by interactive methods might lead to comprehension with transfer to productive competence being accomplished with the aid
215
of direct-instruction methods. Despite the failure to observe aptitude by treatment interactions in the present study, the hypothesis that different language-impaired children might be helped by different programs remains highly plausible. The results obtained here demonstrate only that two developmental measures, IQ and a global measure of language development, do not reliably predict relative success in a program. Another possibility is that stylistic, rather than developmental, measures may interact with treatment effectiveness. For example, children who are highly imitative, who are relatively holistic in their early syntax, or whose production is approximately equivalent to their comprehension may be particularly well-suited for the elicited-imitation technique used in direct instruction; whereas children who are less imitative, more analytic in early syntax, or whose comprehension substantially exceeds their production may be better suited for the interactive approach. Further research assessing stylistic as well as developmental measures will evaluate this possibility.
ACKNOWLEDGMENTS The research reported here was conducted as part of the first author's doctoral dissertation, submitted to the University of Washington. We wish to thank Dr. Joseph Jenkins, Director of the Experimental Education Unit, University of Washington, for making this.study possible. The invaluable contributions of Drs. Rebecca Fewell, Percy Peckham, and Nancy Hatfield are also gratefully acknowledged. Funding for this study was provided in part by a grant from the Washington Association for Retarded Citizens to Joseph Jenkins.
REFERENCES ARAM,D., & NATION,J. (1980). Preschool language disturbances and subsequent language and academic difficulties. Journal of Communication Disorders, 13, 159-178. BEERY, K., & BUKTENICA, N. (1967). Developmental Test of Visual Motor Integration. Chicago: Follett. BEREITER, C., & ENGLEMANN, S. (1966). Teaching disadvantaged children in the preschool. Englewood Cliffs, NJ: Prentice Hal/. BLANK, M. (1973). Teaching learning in the preschool: A dialogue approach. Columbus, OH: Charles E. Merrill. BLOOM, L., & LAHEY, M. (1978). Language development and language disorders. New York: John Wiley & Sons. BLOOM, L., ROCISSANO, L., & HOOD, L. (1976). Adult-child discourse: Developmental interaction between information processing and linguistic knowledge. Cognitive Psychology, 8, 521-552. BURGEMEISTER, B., BLUM, L., & LORGE, I. (1972). Columbia Mental Maturity Scale. New York: Harcourt Brace Jovanovich. CARROw-WOOLFOLK,E. (1981). Carrow Auditory-Visual Abilities Test. Hingham, MA: Teaching Resources. CHAPMAN, R. (1981). Exploring children's communicative intents. In J. Miller (Ed.), Assessing language production in children: Experimental procedures (pp. 111-138). Baltimore: University Park Press. CONNELL, P., GARDNER-GLETTY,D., DEJEWSKI,J., & PARKSREINICK, L. (1981). Response to Courtright and Court_right. [Letter]. Journal of Speech and Hearing Research, 24, 146-148.
216
Journal of Speech and Hearing Research
COOPER, J,, & GRIFFITHS, P. (1978). Treatment and prognosis. In M. Wyke (Ed.), Developmental dysphasia (pp. 159-176). New York: Academic Press. COURTRIGHT, J., & COURTRIGHT, I. (1976). Imitative modeling as a theoretical base for instructing language-disordered children. Journal of Speech and Hearing Research, 19, 655-663. COURTRIGHT, J., • COURTRIGHT, 1. (1979). Imitative modeling as a language intervention strategy: The effects of two mediating variables. Journal of Speech and Hearing Research, 22, 389--402. DENHAM, C., & LIEBERMAN, A. (Eds.). (1980). Time to learn. Washington, DC: The National Institute of Education. DORE, J. (1978). Requestive systems in nursery school conversations: Analysis of talk in its social context. In R. Campbell & P. Smith (Eds.), Recent advances in the psychology of lan-
guage: Language development and mother child interaction (pp. 271-292). New York: Plenum Press. DUNN, L. M. (1959). Peabody Picture Vocabulary Test. Circle Pines, MN: American Guidance Service. DUNN, L., & DUNN, L. (1981). Peabody Picture Vocabulary Test-Revised. Circle Pines, MN: American Guidance Service. ENGLEMANN, S., & OSBORN, J. (1976). DISTAR Language I: An Instructional System. Chicago: Science Research Associates. ENGELMANN, S., ROSS, D., & BINGHAM, V. (1982). Basic Language Concepts Test. Tigard, OR: C.C. Publications. FRIEDMAN, P., & FRIEDMAN,K. (1980). Accounting for individual differences when comparing the effectiveness of remedial language teaching methods. Applied Psycholinguistics, 1, 151-170. GOTTSLEBEN, R., TYACK, D., & BUSCHINI, G. (1974). Three case studies in language training: Applied linguistics. Journal of Speech and Hearing Disorders, 39, 213-224. GRAY, B., & RYAN, B. (1973). A language program for the nonlanguage child. Champaign, IL: Research Press. HALL, P., & TOMBLIN, J. (1978). A follow-up study of children with articulation and language disorders. Journal of Speech and Hearing Disorders, 43, 227-241. HART, B., & ROGERS-WARREN,A. (1978). A milieu approach to language teaching. In R. Schiefelbusch (Ed.), Language intervention strategies (pp. 193-235). Baltimore: University Park Press. HEGDE, M. (1980). An experimental-clinical analysis of grammatical and behavioral distinctions between verbal auxiliary and Copula. Journal of Speech and Hearing Research, 23, 864--877. HEGDE, M., & GIERUT, J. (1979). The operant training and generalization of pronouns and a verb form in a language delayed child. Journal of Communication Disorders, 12, 23-34. HEGDE, M., NOLL, M., & PECORA, R. (1979). A study of some factors affecting generalization of language training. Journal of Speech and Hearing Disorders, 44, 301--320. KING, R., JONES, C., & LASKY, E. (1982). In retrospect: A fifteen-year follow-up report of speech-language-disordered children. Language, Speech and Hearing Services in Schools, 13, 24-32. KIRK, S., MCCARTHY, J., & KIRK, W. (1968). The lllinois Test of Psycholinguistic Abilities. Urbana, IL: University of Illinois Press. KOENIGSKNECHT,R., & LEE, L. (1974). The assessment of gram-
matical development in children and the clinical presentation of grammatical structure to children with language problems. United States Office of Education final report. Evanston, IL: Northwestern University Press. LEE, L. (1971). Northwestern Syntax Screening Test. Evanston, IL: Northwestern University Press.
29
206-217
June 1986
LEE, L. (1974). Developmental sentence analysis. Evanston, IL: Northwestern University Press. LEE, L., KOENIGSKNECHT, R., & MULHERN, S. (1975). Interactive language development teaching. Evanston, IL: Northwestern University Press. LE1NHAm)T, G., & SEEWALD, A. (1981). Student level observation of beginning reading. Journal of Educational Measurement, 18, 171-177. LEONARD,L. (1975). Developmental considerations in the management of language disabled children. Journal of Learning Disabilities, 8, 232-237. LEONARD, L. (1981). Facilitating linguistic skills in children with specific language impairment. Applied Psycholinguistics, 2, 89-118. McLEAN, J. (1983). Historical perspectives on the content of child language programs. In J. Miller, D. Yoder, & R. Schiefelbusch (Eds.), Contemporary Issues in Language Intervention.'ASHA Reports, 12, 115-126. McLEaN, J., & SNYDER-MCLEAN, L. (1978). A transactional approach to early language learning. New York: Charles E. Merrill. MILLER, J. (Ed.), (1981). Assessing language production in children: Experimental procedures. Baltimore: University Park Press. MULAC, A., & TOMLINSON, C. (1977), Generalization of an operant remediation program for syntax with language-delayed children. Journal of Communication Disorders, 10, 231-244. MUMA, J. (1978). Language handbook, Englewood Cliffs, NJ: Prentice-Hall. PEDHAZUR, E. J. (1982). Multiple regression in behavioral research, (2nd ed.) New York: Holt, Rinehart & Winston. PRELOCK, P., & PANAGOS, J. (1980). Mimicry versus imitative production in the speech of the retarded. Journal of Psycholinguistic Research, 9, 565-578. RIEKE,'J., LYNCH, L., & SOLTMAN,S. (1977). Teaching strategies for language development, New York: Grune and Stratton. SIEGEL, G., & SPRADLIN,J. (1982). Language training in natural and clinical environments. Journal of Speech and Hearing
Disorders, 47, 2-6. SILVA, P. (1980). The prevalence, stability and significance of developmental language delay in preschool children. Developmental Medicine and Child Neurology, 22, 768-777. TERMAN, L., & MERRILL, M. (1960). Stanford-Binet Intelligence Test. Boston: Houghton and Mifflin. WEINER, P, (1974). A language delayed child at adolescence. Journal of Speech and Hearing Disorders, 39, 202-212. WELLER, C. (1979). Training approaches on syntactic skills of language-deviant children. Journal of Learning Disabilities, 12, 46-55. WILCOX, M., & LEONARD, L. (1978). Experimental Acquisition of wh-questions in language-disordered children. Journal of
Speech and Hearing Research, 21,220-239. ZIMMERMAN, I., STEINER, V., & POND, R. (1979). Preschool Language Scale-Revised Edition. Columbus, OH: Charles E. Merrill. Received October 29, 1984 Accepted October 23, 1985 Requests for reprints should be sent to Kevin N. Cole, Experimental Education Unit, WJ-10, Child Development & Mental Retardation Center, University of Washington, Seattle, WA 98195.
COLE & DALE:
APPENDIX CONVERSATIONALACT CATEGORIES (SouRCE: MILLER, 1981 )
Requests: Yes/no questions seeking true-false judgments Wh-questions seeking factual information Clarification questions about the content of a prior utterance Action requests seeking that the listener do something Permission requests Rhetorical questions seeking acknowledgment from listener to allow speaker to continue
Responses to Requests: Yes/no answers supplying true-false judgment Wh-answers supplying solicited factual information Clarifications supplying the relevant repetition Compliances verbally expressing acceptance, denial, or acknowledgment of a prior action or permission request Qualifications supplying unexpected information in response to the soliciting question Repetitions repeating part of prior utterances
Direct and Interactive Language Instruction
217
Internal reports expressing emotions, sensations, and mental events including intents to perform future actions Attributions reporting beliefs about other's internal states Explanations expressing reasons, causes, and predictions
Acknowledgments: Acceptances, neutrally recognizing answers or nonrequests Approval/agreements positively recognizing answers or nonrequests Disapproval/disagreements negatively evaluating answers or nonrequests Returning the floor, acknowledging rhetorical questions and some nonrequests
Organization Devices: Boundary markers indicating openings, closings, and other significant points in the conversation (e.g., topic switches) Calls soliciting attention Speaker selections explicitly indicating speaker of next turn Politeness markers indicating ostensible politeness Accompaniments maintaining verbal contact, typically conveying information redundant with respect to context
Perf ormat ives : Descriptions: Identifications labeling objects, events, etc. Events, actions, propositions, etc. are described Properties, traits, or conditions are described Locations or direction are expressed Times are reported
Statements: Expression of rules, procedures, definitions, etc. Evaluations expressing attitudes, judgments, etc.
Protests registering complaints about the listener's behavior Jokes displaying nonbelief toward a proposition for humorous effect Claims establishing rights by being said Warnings alerting the listener of impending harm Teasing comments, taunting or playfully provoking the listener
Miscellaneous: Exclamations expressing nonpropositional emotional reactions
Direct Language Instruction and Interactive Language Instruction with Language Delayed Preschool Children: A Comparison Study Kevin N. Cole, and Philip S. Dale J Speech Hear Res 1986;29;206-217
This information is current as of July 25, 2011 This article, along with updated information and services, is located on the World Wide Web at: http://jslhr.asha.org/cgi/content/abstract/29/2/206