_I l LIBRARY Michigan State University _L This is to certify that the thesis entitled A COMPARATIVE STUDY OF LOCATIVE MEANING PRODUCED BY NORMAL AND LANGUAGE IMPAIRED CHILDREN IN A STRUCTURED ENVIRONMENT: A PILOT STUDY presented by Lisa Ann Reath has been accepted towards fulfillment of the requirements for M. A . degree in Speech-Language Pathology Majo professor Date November 1987 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution )VIESI_J RETURNING MATERIALS: Place in book drop to LIBRARJES remove this checkout from J-fl-r‘ullL your record. FINES will ' be charged if book is returned after the date stamped below. PLEASE NOTE: In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a check mark ./ . A 5090:1999?» 10. 11. 12. 13. 14. 15. 16. Glossy photographs or pages _ Colored illustrations. paper or print __ Photographs with dark background _ illustrations are poor copy __ Pages with black marks, not original copy __\/ Print shows through as there is text on both sides of page__ Indistinct. broken or small print on several pages: Print exceeds margin requirements __ Tightly bound cepy with print lost in spine __ Computer printout pages with indistinct print Page(s) lacking when material received, and not available from school or author. Page(s) seem to be missing in numbering only as text follows. Two pages numbered . Text follows. Curling and wrinkled pages Dissertation contains pages with print at a slant, filmed as received Other U-M-I A COMPARATIVE STUDY OF LOCATIVE MEANING PRODUCED BY NORMAL AND LANGUAGE IMPAIRED CHILDREN IN A STRUCTURED ENVIRONMENT: A PILOT STUDY BY Lisa Ann Reath A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Audiology and Speech Sciences 1987 ABSTRACT A COMPARATIVE STUDY OF LOCATIVE MEANING PRODUCED BY NORMAL AND LANGUAGE IMPAIRED CHILDREN IN A STRUCTURED ENVIRONMENT: A PILOT STUDY BY Lisa Ann Reath This study compared locative utterances produced by normal and language impaired children cut a structured elicitation task. Utterances were analyzed to determine whether group differences were present within global and specific contexts. The effectiveness of the structured task in eliciting the responses was measured. Three language impaired subjects and three normal control subjects matched by chronological age participated in the study. Their ages ranged from 7;6 to 7;5. The subjects were screened using the Northwestern Syntax Screening Test, Leiter International Profile _S_c_q_r_g, and Test of Language Development tests. Their screening scores were variable and indicated ‘that the subject groups *were heterogeneous. As was predicted, all subjects evidenced global locative knowledge. However, contrary to what was expected, group differences were not observed within the specific levels of locative knowledge. The structured task was judged to be average in effectively eliciting locative responses. Improvements for the structured task were suggested. DEDICATION This is dedicated to my husband, Mark, whose emanating love and support was always there, especially when it was most needed. 111 ACKNOWLEDGMENTS I would like to thank Ida J. Stockman, Ph.D., Chairperson of the thesis committee, for the time dedicated to this investigation, and the quality of advice so gratuitously offered. My appreciation also is extended to Leo V. Deal, Ph.D., and Philip Davidson, Ph.D., for accepting to be a member of the thesis committee, for their support and advice, and for their willingness to adapt their schedules. I would like to thank Farah Stockman for her creativity and hard work in constructing the landscape scene used in the study, and Dawn Nyhoff, Sharon McWhirter, and Katie Reid for their help needed to gather the data, and for their support. Finally, I would like to thank the subjects participating in the study, and their families for their time and support. This study was conducted as part of AURIG Grant No. 714685, Ida J. Stockman, Ph.D., principal investigator. 1v TABLE OF CONTENTS Page LIST OF TABLES O O O C O C O O O O O O O C . Vii LIST OF FIGURES O O O O O O O O O O O O O O O x LIST OF APPENDICES . . . . . . . . . . . . . . X1 Chapter I. BACKGROUND FOR THE STUDY . . . . . . . . . . l The Semantic Focus in the Description of Children's Language Acquisition . . . . . . 1 Arguments for Linkage Between Language Acquisition and Cognition . . . . . . . . . . . . 3 Implications of the Language and Cognition Link for Language Impaired Children . . . . . . S Examining the Cognitive Performances of Language Impaired Children with Normal Nonverbal' Intelligence . . . . . . . . . . . . 6 Studies Examining the Semantic Categories Produced by Language Impaired Children . . . 8 Statement of the Problem . . . . . . . . . ll Semantic Category Targeted . . . . . . . . l7 Implications for Developing a Structured Task to Elicit Locative Utterances . . . . . . . 18 Purpose of the Study . . . . . . . . . . 19 11. REVIEW OF LITERATURE ON SUBCATEGORIES OF LOCATIVES . 21 Background History of Locative Subcategories . . 21 Dynamic and Static Locative Subcategories in Normal Language Acquisition . . . . . . . 23 III. EXPERIMENTAL PROCEDURES . . . . . . . . . . 26 Subjects . . . . . . . . . . . . . . 26 Subject Description . . . . . . . . . . 26 Subject Selection Procedures . . . . . . . 29 Chapter Page Description of Stimuli and Tasks . . . . . . 31 Operational Definitions of Locative Subcategories Targeted . . . . . . . _. . . . . . 31 Description of Structured Task . . . . . . 32 Responses to the Structured Task . . . . . . 49 Spontaneous Language Sampling . . . . . . . 60 IV. RESULTS 0 O O O O O O O O O O O O O O 65 Structured Task Data . . . . . . . . . . 65 Temporal Reliability . . . . . . . . . . 81 Va DISCUSSION 0 O O O O O O O O O O O O O 83 Explanations for Subject Group Differences Not Revealed Among the Subcategories of Locative Knowledge . . . . . . . . . . . . . 85 Alternate Hypotheses . . . . . . . . . . 85 Subject Characteristics . . . . . . . . . 86 Nature of the Targeted Locative Subcategories . 87 Explanations for the Why the Structured and Spontaneous Elicitation Procedures Showed Particular Similarities and Differences . . . 87 Recommended Changes for the Structured Task Procedure . . . . . . . . . . . . . 90 Recommended Changes for the Scoring System . . 91 Recommended Changes to the Test Presentation Format . . . . . . . . . . . . . 93 APPENDICES . . . . . . . . . . . . . . . . . 98 REFERENCES . . . . . . . . . . . . . . . . 137 vi 10. ll. 12. 13. 14. LIST OF TABLES Subject Characteristics . . . . . . . . . . Operational Definitions and Linguist/Contextual Criteria Used to Assign Utterances to the Global Category of DYNAMIC LOCATIVE UTTERANCES . . . . . Operational Definitions and Linguistic/Contextual Criteria Used to Assign Utterances to Global Category of STATIC LOCATIVE UTTERANCES . . . . . Description of A Events . . . . . . . . . . Locative Words Targeted for Each Locative Subcategory Within the A Events . . . . . . . . . . . . Locative Words Targeted for Each Locative Subcategory Within Event IA . . . . . . . . . . . . . Description of B Events . . . . . . . . . . Subjects Who Exhibited the General Categories of Dynamic and Static Locative Responses to the Structured Task . . . . . . . . . . . . . Subjects Who Exhibited the Specific Categories of Locative Response to the Structured Task . . . . . Positional Locative Words Elicited During the Struc cured TaSk O O O O O O O C O O O O O Directional Locative Words Elicited During the Structured Task . . . . . . . . . . . . . Task Scores for Each Subject by Locative Subcategory . Specific Locative Subcategories Elicited in Structured Task and Language Sample . . . . . . . . . . Mean Percent of Subjects Who Produced Positional Locative Words During the Structured Task and Language Sampling . . . . . . . . . . . . vii Page 27 33 34 38 40 4O 45 66 68 69 71 74 77 79 Table 15. 16. 1.1. 1.2. 1.3 1.4 1.5 K.l L01 L02 L.3 L04 L05 Mean Percent of Subjects Who Produced Directional Locative Words During the Structured Task and Language Sampling . . . . . . . . . . . . Test-Retest Scores Obtained by One Nonclinical SUbj ect O O O O O O O O O O C O O O O 0 Subject Scores for Individual Targeted Directional Locative Words--STATIC PERSPECTIVE LOCATIVE . . . . Subject Scores for Individual Targeted Directional Locative Words--DYNAMIC DIRECTION LOCATIVE . . . . Subject Scores for Individual Targeted Locative Positional Words--STATIC POSITION . . . . . . . Subject Scores for Individual Targeted Locative Positional Words--DYNAMIC ORIGIN . . . . . . . Subject Scores for Individual Targeted Locative Positional Words--DYNAMIC DESTINATION . . . . . . Rank Ordering of Group Mean Percentages Averaged Across All Five Tasks . . . . . . . . . . . Raw Data Used for the Rank Sums Test for Structured Task Mean Scores Between the Subject Groups Averaged Across All Five Tasks . . . . . . . . . . . Distribution of Locative Words Elicited for Static Position Locative Subcategory During the Structured and Spontaneous Language Sampling Tasks . . . . . Distribution of Locative Words Elicited for Dynamic Origin Locative Subcategory During the Structured and Spontaneous Language Sampling Tasks . . . . . . Distribution of Locative Words Elicited for Dynamic Destination Locative Subcategory During the Structured and Spontaneous Language Sampling Tasks . . . . . Distribution of Locative Words Elicited for Dynamic Direction Locative Subcategory During the Structured and Spontaneous Language Sampling Tasks . . . . . Distribution of Locative Words Elicited for Static Perspective Locative Subcategory During the Structured Spontaneous Language Sampling Tasks . . . . . . viii Page 80 82 119 120 121 122 123 124 125 126 127 128 129 and 130 Table Page M.1 Static Positional Locative Subcategory . . . . . 131 M.2 Dynamic Direction Locative Subcategory . . . . . 132 133 M.3 Dynamic Origin Locative Subcategory . M.4 Dynamic Direction Locative Subcategory . . . . . 134 M.5 Static Perspective Locative Subcategory . . . . .' 135 N.1 Raw Data Used in the Rank Sums Test Between Test- Retest Scores 0 O O I O O O I O O O O O O 136 ix LIST OF FIGURES. Figure Page 1. Range of Structured Task Score for Language Impaired and Normal Groups by Locative Subcategory . . . . 75 LIST OF APPENDICES Appendix A. Description of Subjects . . . . . . . . . . B. Human Subjects Affidavit . . . . . . . . . . C. Lansing Public School District Project Approval Form. D. Teacher Questionnaire Forms . . . . . . . . . E. Test Forms Used During the Screening . . . . . . F. Parental Consent Form . . . . . . . . . . . G. Definition of General and Specific Locative Words . H. Language Sample Score Form . . . . . . . . . 1. Subject Scores for Indivdiual Targeted Locative words I O O O O O O O O O O O O . O O O J. Rank Ordering of Group Mean Percentages Averaged Across All Five Tasks . . . . . . . . . . . K. Raw Data Used for the Rank Sums Test for Structured Task Mean Scores Between the Subject Groups Averaged Across A11 Five Tasks . . . . . . . . L. Distribution of Locative Words Elicited for the Five Targeted Locative Subcategories During the Structured and Spontaneous Language Sampling Tasks . M. Rank Order for Mean Number of Subjects Producing The Targeted Locative Words for all Five Subcategories Between the Structured Task and Language Sampling Procedures . . . . . . . . . . . . . . N. Raw Data Used in the Rank Sums Test Between Test- Retest Scores . . . . . . . . . . . . . xi Page 98 104 105 106 110 114 116 118 119 124 125 126 131 136 CHAPTER I BACKGROUND FOR THE STUDY This research focused cut the relational semantic features of normal and language impaired children's language. Relational semantics refers to the various meanings words acquire in relation to other words in syntactic context. For example, "Mark" acts as an agent in the sentence ”Mark is cooking," but as an owner in the sentence "This is Mark's car.” Over the past decade, the study of language development has expanded its focus beyond linguistic form to include the relational semantic aspects. The following section reviews literature concerning this shift in focus, and its significance to the nature of language impairment. The Semantic Focus in the Description of Children's Language Acquisition During the 1960's, much research focused on the development of children's linguistic forms (Freeman. & Carpenter, 1976; Leonard, Bolders, & Miller, 1976). This focus was most significant with the emergence of transformational grammars as vehicles for describing child and adult syntax. (See Bowerman, 1973 for literature review.) Transformational grammars were challenged partially because they inadequately' described. children's productions. For' example, they were unable to explain ambiguous sentences and, consequently, left themselves open to anomalous productions. As a result, their distributional criteria could not predict-early word combinations. For a more complete argument, refer to Bowerman (1973). As a result of this inadequacy, investigative focus shifted from linguistic forms to the semantic concepts underlying the forms. Bloom (1970) was one of the pioneering investigators who systematically examined semantic features of children's early utterances. She used ”rich interpretation,” which considered both the syntactic structure of the utterance and the nonlinguistic context in which it occurred. It was observed that a sentence with the same syntactic arrangement of ‘words could be interpreted. as having several meanings: "Daddy shoe” couLd mean ”this is Daddy's shoe," "Daddy has my shoe,” or ”Daddy get the shoe.” Also, several separate syntactic arrangements could be interpreted as representing one meaning. For example, ”Mama down," ”Mama put me down,” "Me down," and "Mama me down” could all convey the message that a child wants to be put down by the mother. Other investigators, using a semantic research focus, developed varying classification systems to describe the relational meaning of children's language, e.g., Schlesinger (1971). Brown (1973) reviewed 19 reports on 13 children researched by various investigators. Specifically, the goal was to reveal the semantic relations prevalent across languages at early MLU (Mean Length of Utterance) defined stages of language development. The primary data were gathered from spontaneous conversations within natural environments of children exhibiting similar MLU values. The various languages studied were American, English, Finnish, Swedish, Samoan, and Spanish. When Brown (1973) described the children's utterances according to their semantic relations, it was revealed that a small set of the same relations were expressed by children acquiring the specified languages. Brown listed his minimal two-term semantic relations as follow: 1. agent + action 5. entity + location 2. action + object 6. possessor + possession 3. agent + object 7. entity + attributive 4. action + location 8. demonstrative + entity These eight basic semantic relations accounted for about 702 of most of the children's 'utterances, and they' 'were defined semantically, rather than syntactically. That is, there existed no syntactic ordering of these relations. For example, agent + action could also occur as action + agent during a child's early utterance. Bowerman (1973) argued that the universality of these globally defined semantic relations could serve as the basis for children's independent syntactic development of various language stemming from a universal semantic base. Arguments for Linkage Between Language Acquisition and Cognition The evidence that the distribution of children's syntactic forms is determined by the underlying meaning relations prompted questions about the origin of these meanings. That is, if language acquisition is derived from learning, then what is meaning derived from? As meaning referred to that part of cognition which could be coded by symbols, there arose a renewed interest in Piaget's theory of cognitive development for explaining the Origin of meaning. This theory includes a hypothesis about ”abstract organizational patterns" created from interaction with the environment (Rice, 1983, p. 348). Piaget believed these cognitive patterns follow a universal development in children (Rice, 1983). It was hypothesized that if language were derived from meaning and meaning were derived from cognition, then language may be derived from cognition. To support this hypothesis, investigators began to search for evidence demonstrating that language and cognition were linked. This search produced a wide range of results. For example, Slobin (1973, cited in Rice, 1983) observed that children produced new forms to express old meanings and old forms to express new meanings. This type of evidence provided an argument for a strong relation between cognition and language by indicating that children must have meanings in place before they use forms to code the meanings. However, other investigators (e.g., Casby & Ruder, 1983) observed that first word combinations emerged at the same time as symbolic play. The simultaneous emergence of cognitive and linguistic skills suggested that cognition did not necessarily emerge before language nor did language emerge before cognition. These correlational data provided an argument for a more parallel relationship between cognition and language. Other observations suggested only a weak link between cognition and language. For example, synonyms are created and used to express identical meanings--laugh, giggle, roar, chortle, snicker, cackle, break up, split one's sides, and roll in the aisle--all express the same meaning. As another example, a child may have called himself by his name and later refer to himself as "me” or 'I" (Cromer, 1976, cited in Rice, 1983). These illustrations suggest that while cognition appear to be the basis for some primary linguistic forms, cognition may not be the basis for all forms children acquire. Additionally, some observations suggested the relation between cognition and language was more linguistically based. For example, Schlesinger (1982, cited in Rice, 1983) observed that children's general meanings about the world become more specifically shaped and better anchored as they linguistically code the meanings and receive linguistic feedback. This supported the theory that cognitive meanings are anchored through language. It is clear that the literature varies on the issue of linkages between cognition and language. These various proposals range from a strong to a weak link, and from cognition anchoring language to language anchoring cognition. While these discrepancies remain, the fact that a link exists is generally agreed on. For a more critical discussion of supportive evidence for linkages, see Rice (1983) and Rice and Kemper (1984). Implications of the Language and Cognition Link for Language Impaired Children The notion that critical links exist between normal language acquisition and cognition led to the hypothesis that failure to acquire language normally is linked to cognitive impairment. This reasoning was pursued in. two categories of studies. 'The first category sought to» examine closely’ the cognitive performances of language impaired children who had been. judged as having normal nonverbal intelligence, and the second category sought to examine the semantic categories produced by language impaired children. Examininggthe Cognitive Performances of Language Impaired Children with Normal Nonverbal Intelligence The suggestion that specifically language impaired children have cognitive deficits conflicts with the traditional notion that such language impaired children exhibit normal nonverbal cognition. This traditional notion has been supported by the observation that normal range scores are received by language impaired children on conventional cognitive tests, such as the Leiter International Performance Scale (LIPS) (Leiter, 1959). However, studies have begun to show that such language impaired children exhibit cognitive deficits on nonconventional cognitive tasks. These investigations have compared language impaired children varying in age from 32 months to 12 years with their chronological age matched normal peers. All of these children had performed within normal range on various widely used intelligence tests, such as the LIPS (Leiter, 1959). Language impaired groups have performed poorer than the normal groups on the nonverbal cognitive tasks in the areas of symbolic and imaginative play (wain a Yule, 1983; Terrell et al., 1984), anticipatory imagery abilities (Johnston & Savich, 1984), mental rotational abilities (Johnston & Weismer, 1983), and cognitive and semantic processing (Wren, 1982). The results of this research suggested that children with verbal deficits also possessed coexisting nonverbal cognitive deficits. However, if these children did have‘cognitive deficits, the question arose as to why they scored within the normal range on the traditionally used intelligence tests. Johnston (1982) speculated that the widely used intelligence tests were not sensitive enough to reveal the nonverbal cognitive deficits of language impaired children. This claim motivated her review of the LIPS. Johnston observed that the items of the LIPS appeared to consist of two groups: the perceptual items and the conceptual items. She hypothesized that the language impaired subjects would perform better with the perceptual items, and that they would differ in performance from the normal children in both the number and types of items passed. What she observed, however, was that the first hypothesis, and not the second hypothesis, was confirmed by her observation. She proceeded to argue that it would be inappropriate to label children who received scores in the normal range as possessing ”normal‘ nonverbal intellectual functioning" and acknowledged that "a child who succeeds in perceptual tasks may or may not show equivalent levels of development in other areas of cognition” (p. 295). Finally, Johnston suggested it would be appropriate to conclude that the children who receive scores within the normal range on the LIPS demonstrated only that they have age-appropriate visual perceptual processing for static events. In summary, in the first category of studies, the literature demonstrated that language impaired children may have had deficient nonverbal cognition which is not detected by widely used intelligence tests, such as the LIPS. Studies Examining the Semantic Categories Produced by Langggge Impaired Children The second category of studies compared the semantic relations produced by normal and language impaired children. The population observed in this category' of studies also» consisted of language impaired children who received scores within the normal range on conventional intelligence tests. In addition, some of the population in these studies included mentally retarded language impaired children. In all of these studies, the investigators compared various groups of language impaired children with their MLU-matched peers. The children generally ranged in utterance length from 1 to 3.6 morphemes. By comparing normal children to language impaired children, these studies provided implicit evidence for cognitive deficits among the language impaired. Assuming that semantic relations reflect the cognitive organization of experience, the presence of a cognitive/ representational deficit is implied. when the full complement of syntactic relations expected at a gdven age are not represented in impaired language. The following is a description of some of the studies which fall into this category. Freedman and Carpenter (1976) investigated two-word utterances of language impaired children and normal children matched at the same linguistic level (Brown's Stage I level of linguistic development). They were compared on their use of ten basic semantic relations: agent + action possessor + possession action + object entity + attribute agent + object introducer + entity action + location more + entity entity + location negation + entity Type-token ratios for each semantic relation were used to compare the productions of semantic relations between groups. A significant difference was obtained for only one relation, with the language impaired group demonstrating greater diversity in the use of the introducer + entity relation. The investigators concluded from their findings that at the Stage I level of linguistic development, language impaired children demonstrated a linguistic system no different from the system of normal Stage I children. Duchan and Erickson (1976) conducted another study in which semantic relations were presented in different verbal contexts (expanded, telegraphic, and nonsense) to normal developing and mentally retarded language impaired children matched on their MLU's (between 1 and 2.5). The semantics used in this study were agent + action, action + object, possessives, and locatives. They found no significant differences between the performance of the two groups on the verbal comprehension task. Leonard et al. (1976) examined semantic relations reflected in language usage as a function of chronological age (3 and 5 years) and linguistic status (normal and language disordered). They discovered that the language impaired children expressed the same relational meanings as their MLU and age matched normal controls. They 10 interpreted their results as supporting the notion that the language disordered. group reflected semantic relations consistent with an earlier level of development. Coggins (1979) explored the early two—word utterances from Down's syndrome children to determine whether they encode the same relational meanings as children developing normally. Nine semantic categories were used to classify the subjects' two-word constructions: demonstrative + entity negation + entity agent + action action + object agent + object action + locative entity + locative possessor + possession entity + attribute Coggins suggested that Down's children concentrate on the same set of relational meanings as normal children. Fokes and Konefal (1981) researched normal and language impaired children's ability to produce sentences containing specific semantic- case relations under two conditions (manipulation and observation conditions). The three groups of subjects consisted of normal children with a mean age of 3:6 years, normal children with a mean age of 5:6 years, and language impaired children with a mean age of 7:0 years. The language impaired group was operating minimally at an educable level on the Wechler Intellflence Test and demonstrated measured MLU's from 1-3.6 morphemes. No attempt was made to obtain scores for mental age. 11 Manipulation and observation tasks were used to obtain sentences containing the agent 4» action + object 4- locative semantic case relations. The language impaired group demonstrated productive usage of all four of the semantic notions, but they had a tendency to produce more single- and two-word utterances. Their production of four-case relation strings increased with the manipulation task. The relations used most frequently by the language impaired group when producing two—word utterances were: action + locative, action + object, and object + locative. In summary, these comparative studies of language impaired children and their MLU matched peers, collectively, revealed that language impaired children represented the same major categories of semantic relations in their language as normal children. This was true both in the case of the specifically language impaired children and the mentally retarded language impaired children. However, it should also be noted that only a small number of studies compared the semantic relations produced by children. In fact, the five studies just reviewed comprise a large fraction of the existing studies of semantic relations in clinical groups. Statement of the Problem In the background section, it was shown that the focus on the semantic features of the language impaired and the accompanying cognitive implications reflected the same focus observed in. the literature regarding normal language acquisitions It appears that the literature on language impaired children, which has grown out of 12 this shift, leads to conflicting positions about the relation between language and the nonverbal cognition assumed to underlay it. The results of the first category of studies suggest that some nonverbal cognitive deficits may be present in language impaired children. However, to the extent that semantic knowledge indexes cognitive representation in normal and language impaired children, it may be inferred from the results of the second category of studies, that these children are essentially equal in nonverbal cognition. This discrepancy indicates the need for further investigation that could reveal the possible linkages between the language performances of language impaired children and their underlying cognition. There are several possible explanations for why studies of semantic relations have not yielded the same outcomes as the more direct studies of cognitive performances. First, there may be no link between semantic knowledge and cognition: and consequently, cognitive and semantic knowledge may exist as two separate domains of knowledge or skill. Therefore, an existing nonverbal cognitive deficit need not show up as a semantic relational deficit. Another possibility, which provided the focus for this research, was that the cognitive differences between normal and language impaired children may have existed, but were not revealed because of the global nature of the meaning relations studied. The meaning relations were primary and universal, and they referred only to very broad referent categories. To the extent that these categories are basic and universal, any child could be expected to exhibit them. Stockman (1 986) observed , 13 Children all over the world probably use their language to reflect their experiences with moving versus nonmoving objects. The notion of universality leads to the expectation that any child's language will reflect such basic semantic features if they' tap ’something that is basic to human representation (p. 1). As semantic relations coded only very broad cognitive concepts, they may' not have revealed. the more specific concepts occurring within each relation. Consequently, children who are cognitively different on a specifically defined semantic task may not appear different in their use of globally defined semantic relations. Other investigators have also observed that meaning relations have been described in terms which may be too general to capture developmental differences (Johnston, 1982: Stockman, 1986). For example, Johnston (1982) suggested, The taxonomies of meaning relations developed by Schlesinger (1974), Brown (1973), and Fillmore (1968) categorize only the broadest aspects of communicative intention. Applied to normal child language data, such systems have proved rather weak in capturing developmental trends (p. 787). Also, Stockman (1986) observed that, Failure to reveal differences between normal and clinical children may be due partly to the use of a descriptive model that focuses observation on just the general features of meaning. Applying adaptations of case grammar notions (Bowerman, 1973: Bloom, Lightbrown, & Hood, 1975), semantic relational knowledge has been described in terms of broad referent categories (e.g., action, state, locative action, locative state, etc.) represented by two and three term semantic relations that are coded by the major syntactic constituents of early sentences (p. 1). To further illustrate the global nature of semantic relations, it has been demonstrated that the broad semantic concepts, which served as the base for previous semantic relations studies, could be 14 divided into more specifically defined semantic subcategories. Stockman and. Vaughn-Cooke (forthcoming: cited in Stockman, 1986) illustrated this point in the following examples: Consider locative action constructions. They refer to the displacement of objects from one spatial point to another. For example, the constructions, (1) The cat jumps down, (2) the cat jumps from the table, (3) the cat jumps onto the floor or (4) the cat jumps down onto the floor, are all locative action constructions, but they differ in the particular aspect of the locative event coded. Utterance #1 refers to locative direction of movement. Utterance #2 identifies the original place of movement: #3 focuses on locative destination whereas #4 refers to a combination of two of these features (p. 2). The notion of semantic subcategories that more adequately characterize children's semantic representations is analogous to the more global category of "dog” having subcategories such as ”Golden Retrievers,” ”Labrador Retrievers,” and "Doberman Pinchers.” Other researchers also discussed semantic subcategories. For example, Olswang and Carpenter (1982a, 1982b) demonstrated a five- level developmental sequence for the cognitive notion of agent: Huttenlocher et al. (1983) defined eight subcategories of action: and Bloom and Lahey (1978) observed three subcategories of negation. Therefore, while studies have demonstrated few differences between the language impaired and normal children in their productions of globally defined semantic relations, it was speculated that semantic differences may be revealed by using more detailed descriptive models of semantic representation. While this hypothesis provided the direction for this study, a more immediate consideration 15 was to determine an appropriate procedure for eliciting language that could test such an hypothesis. Both standardized and nonstandardized approaches have been used to elicit language in clinical settings. The advantages and disadvantages of these two approaches were also observed by Bloom and Lahey (1978), Lucas (1980), and Mecham and Willbrand (1979). Standardized approaches provide procedures which are repeatable by more than one person. This is achieved by structuring the elicitation conditions, including the stimuli and the responses. However, the result is a rigid format for revealing linguistic knowledge. More recently, nonstandard approaches have emerged in an attempt to provide an alternative to the structure exercised over the responses elicited from children. The goal of these approaches is to create an environment which places less control over the stimuli and the responses than the standard approaches. Therefore, the subjects are allowed to produce utterances of any form in an unrestricted manner. Danwitz (1981) observed that both approaches have their advantages and disadvantages. ‘Within. a structured testing environment language performance does not reflect an accurate picture of a child's linguistic knowledge. Danwitz argued further that the child is limited by the context surrounding the testing. The environment does not allow the child to freely express experiences across various contexts. As a result, the nature and degree of a language disorder are not revealed. 16 However, nonstandard procedures also have disadvantages (Danwitz, 1981). For example, the language sampling procedure does not provide enough structure. As a result, factors influencing the elicited productions are not controlled. Such factors include the environmental setting where the testing occurs, the stimulus materials used, the topics of conversations, and the way in which the topics are discussed (e.g., object description, conversational, free play, etc.). Therefore, while the language sampling is not limited by structural constraints, the use of nonstandard elicitation procedures yield variable and unpredictable responses. The data are unpredictable because no constraints operate on the child to make the utterances predictable. For example, during unstructured language sampling, not even the topics of conversation are controlled. As a consequence, the sampling may not be completely representative of the child's linguistic knowledge. For example, consider the situation in which a child produces the word ”up” during a sample, but not ”down.” It would be difficult to determine whether or not the child had *down” within his system if the context facilitating the production of ”down" were never presented. Advantages of each approach exist. The standardized approaches may be more successful than nonstandardized procedures in revealing the limits of linguistic knowledge. For example, if a subject does not produce "down” within a task, that probes the word "down,” one might be more confident that the word is not known. On the other hand, failure to elicit ”down" in a naturalistic context could be due to the lack of opportunity and not the unavailability of the form. 17 The direction of this investigation incorporated the two approaches by creating a structured elicitation task that systematically probed for existing linguistic knowledge while preserving the spontaneity of the responses elicited. This task was not too restricting in that the child was able to respond in an open- ended manner. However, there was enough structural context to guide the child in the direction of desired responses if he had knowledge to express them. In sum, the general goal of this study was to pilot a standard protocol for eliciting a more detailed representation of children's locative expressions. Its effectiveness, practicality, and accuracy benefits were then weighted against the outcome of a spontaneous sampling procedure. Semantic Category Targeted Locative utterances were chosen as the test case for examining the efficacy of a structured elicitation procedure to reveal group differences for three reasons. First, an expanded descriptive model of locative subcategories was available for use. These subcategories (dynamic and static: origin, direction, and position) will be discussed in further detail in the following section. Second, locative expressions are frequently the targets of language assessment and therapy. However, frameworks for assessing locative knowledge are undeveloped in two major ways. First, they have a global focus. For example, assessment of spontaneous language samples typically classify locative semantic relations using only 18 global definition (e.g., action + location or entity + location). Second, they present only a partial analysis of a child's locative system by focusing on a small subset of common locative words (see Stockman, 1985 for a review of lexical studies). As Stockman (1985) noted, Despite the relatively long research track record on normal children's locatives, the largest number of studies has focused on acquisitional meaning of small subsets of locative prepositions such as ”in,” ”on” and "under'-- typically in isolation of the broader semantic/syntactic context in which they occur. On the other hand, prepositional meaning has been ignored by the small set of studies that has focused on sentential locative meaning (p. 29). Third, not many studies focusing on locative expressions of clinical groups have been conducted. Stockman (1985) observed that with the exclusion of the studies using mentally retarded subjects, just four studies were found which focused on children with a specific language impairment. These studies were Duchan and Siegel (1979), Leonard, Bolders, and Miller (1976), Freedman and Carpenter (1976), and Fokes and Konefal (1981). As noted previously, none of the studies employed a developmental focus, and all of them restricted observations to global locative categories. Implications for Developing a Structured Task to Elicit Locative Utterances Testing locative utterances using a structured elicitation task has clinical and theoretical implications. Theoretically, the data gathered can add to the body of literature concerning language impaired children's nonlinguistic cognition. For example, difference 19 found between the locative subcategory productions of the language impaired and normal children would support the theory that nonlinguistic abilities of language impaired children may not be within the normal range. Clinically, an efficient method for gathering and analyzing such data may be developed. Its use on a larger scale could provide a more specific analysis of children's locative productions in clinical settings. Also, goals for treatment can be guided by the definition of normal developmental stages with the verification that such a stage model is applicable to impaired performance. Purpose of the Study The purpose of the study was to investigate the efficacy of a structured elicitation procedure in revealing differences between language impaired and nonimpaired groups within the locative domain. This investigation was guided by the following questions: Research Question 1: Does a structured elicitation task reveal differences between normal and language impaired groups in: a. the presence or absence of utterances in the two globally defined dynamic and static locative categories of meaning b. the presence or absence of utterances in the more specifically defined subcategories of dynamic and static meaning: namely, static position, static perspective, dynamic origin, dynamic destination, and dynamic direction 20 c. the distribution of words elicited within each of the specifically defined locative subcategories, and d. the appropriateness and specificity of responses elicited for the specifically defined locative subcategories? Research Question 2: Do the results of the structured task parallel the content of the language sampling in: a. the number and type of locative subcategories evidenced, and b. the distribution of words within each locative subcategory? CHAPTER II REVIEW OF LITERATURE ON SUBCATEGORIES OF LOCATIVES The purpose of this literature review is to provide the reader with more detailed information about the prior investigations that have focused on locative subcategories, particularly as considered from a semantic relational perspective. While this is a relatively new area, the literature review provides a framework for the kind of research that is proposed. The locative subcategories are defined within this section. Background History of Locative Subcategories Leech (1970) described locative subcategories demonstrated in the adult English language. He described the various ways provided by the language for coding static locations (i.e., locations not involved with movement). They include reference to position and dimensionality, relative position, extremities and parts of locations, compass points, and orientation. When describing location with respect to movement, Leech explained that the movement must be discussed in terms of a goal. For example, ”He is coming" refers to movement: but is not associated with a goal, and therefore, is not identified as locative. However, if the words "here” or "to the store" were added (i.e., "He is coming to the store"), then this movement is associated with a goal, and 21 22 would be identified as a locative. These particular movement locatives are referred to as ”dynamic.” Leech made note that dynamic locatives can be viewed "in two lights: as a transition into a state, or as a transition out of an opposite state" (p. 194). This means that movement locatives can refer to either the origin from which the movement is taking place (i.e., He jumped from the house), or the destination to which the movement is arriving (i.e., He jumped onto the table). Leech also noted that a path of a movement can also be described in the English language (i.e., The boy walked in front of me). The path is neither the origin nor destination of the movement. Finally, Leech discussed resultative position which he defines as ”static position resulting from movement.” He used ”The picture is off the wall” or ”He's just out of jail” as examples (p. 194). Leech suggested these examples were static and dynamic locatives occurring in combination. Following Leech, Bloom and Lahey (1978) describe the semantics relating children's language in terms of dynamic and static locative meaning. The definitions separating the dynamic and static locatives were similar to those of Leech. The distinction was based on the presence of a movement or nonmovement verb when a locative preposition or adverb was being used. In the four children used as subjects, they observed that two of the children encoded dynamic locatives before static. The other two children already had both types of locatives at the time of observation. Unlike Leech, Bloom and Lahey did not identify utterances referring to locative paths, destinations, or origins. However, 23 other research. did observe some of these more specific locative subcategories. Macrae (1976) investigated locative words used to complement diectic movement verbs (go and come). The subjects were seven children ranging in age from 1:3 to 2:9. They were visited weekly in their home, and their spontaneous utterances and the nonlinguistic contexts were recorded. Macrae noted "The children paid little attention to destinations . . . they took account of direction (coming up) without committing themselves to ‘the termination of the movement" (cited in. Stockman, 1985, p. 203). Also, Bowerman. noted in. her study' of two IFinnish children. that ”locative nouns never named a location away from which the referent was moving” (1973, pp. 108—109). More recently, Stockman and Vaughn- Cooke (1983, 1984) observed and defined eight locative subcategories. The results of their research, which provides the model for this study, is summarized in more detail in the following section. Dynamic and Static Locative Subcategories in Normal Language Acquisition Stockman and Vaughn-Cooke (1983, 1984: cited in Stockman, 1985) investigated eight subcategories in a preliminary study of 12 working-class children who were observed during natural play activities at home. Their descriptive framework followed the basic locative distinctions in the adult language described by Leech (1970) and Bennett (1975) (cited in Stockman, 1985). A dynamic and static locative expression can exist in terms of the following meaning subcategories: 24 1. Static origin (e.g., he is off the house) 2. Static direction (e.g., the cat is down from here) 3. Static destination (e.g., it is on the table) 4. Static combinative (e.g., it's down on the table) 5. Dynamic origin (e.g., he jumped off the house) 6. Dynamic direction (e.g., the cat jumps down) 7. Dynamic destination (e.g., I set it on the table) 8. Dynamic combinative (e.g., I set it down on the table) Stockman and Vaughn-Cooke's subjects were subdivided evenly into three separate groups according to age (1:6, 3:0, and 4:6 years). 'The children were observed for two hours in their respective home settings. Their language samples were video recorded, and later their utterances were extracted and recorded on paper. The locative utterances were identified, and the contexts in which these utterances occurred were recorded. Their first major finding was that the younger subjects (1:6 years) referred only to locatives within the dynamic context, whereas. the older subject groups (3:0 and 4:6> years) produced locatives within both the dynamic and static locative contexts. These findings were consistent with Bloom, Lightbown, and Hood (1975). The investigators observed further that the locative words used in the two contexts (static vs. dynamic) differed, depending on the child's age. Certain words, such as "up" and ”down,” were used within the dynamic context almost exclusively at the younger ages (1:6 and 3:0 years) whereas other words, such as ”behind” and ”under," were used mainly within the static locative context. The 25 older subjects began to use a particular word within both static and dynamic contexts. The investigators also noted ”that the distributional cluster of locative words seemed to reflect shared semantic fields” (Stockman, 1985, p. 31). For example, at the earliest ages, the words reflecting the dynamic locative notion appeared to code directionality or orientation of movement (e.g., the ball is going up, the ball is coming down). With the older subjects, the dynamic expressions also included a group of words that referred to positionality (e.g., behind and under). Furthermore, these positional words coded the destinative aspects of the movement (e.g., the doll is being put on the table: the ball is landing in the pool). Developmentally, Stockman and Vaughn-Cooke found that the locative expressions advanced from almost exclusive usage of dynamic locative utterances to the additional use of static locative utterances. They found within the dynamic locative context that the subcategory of locative origin and direction emerged earlier than that of locative destinative position. Also, single category expressions appeared before combinative category expressions. Within the static locative context, the emergence of locative concepts occurred in the order of positional locative expressions, combinative, directional, and finally, original locative expressions. CHAPTER III EXPERIMENTAL PROCEDURES Subjects Subject Description The study included six male children in the age range of 6:7 to 7:5. Three of the subjects were language impaired and three were normal controls matched by chronological age. The subject characteristics are presented in Table 1, and are more completely described in Appendix A. The subjects were selected from low income backgrounds, as indexed by parental occupation and estimations of associated incomes. They were chosen from existing school populations of Lansing, Michigan, and had no frank. neurological insult or physical, sensory, or motor disabilities, as indicated by clinical records or parental reports. No severe difficulties were observed. with the subjects' oral periphery, verbal and nonverbal behavior, such as speech intelligibility, vocal attributes, or response to verbal commands as determined by the investigators' subjective judgments during the screening session (see Phase 2 of Screening Selection section in this chapter). The subjects had no history of chronic ear infections as determined by parent interviews. The research was approved for use 26 Table 1 Subject Characteristics 27 LI Subjects N Subjects Sla 82a 83a Slb 52b 83b Age 6:10 6:7 7:4 6:9 6:11 7:5 Race black white white black white white MLU 4.0 7.0 3.7 3.7 8.0 7.4 NSST: expressive- (102 >252 (101 251 >251 (101 (501 (501 receptive- 101 >251 (102 (251 (251 (102 (501 LIPS >6:0 )6:7 5:3 >7:3 >7:3 >7:6 TOLD below below below average average average Note: Five of the subjects tested through appropriate ages on the Leiter International Performance Scale and upper boundaries were not obtained. SCOI’BS . LI - language impaired: N - normal: Test of Language Development results indicate composites MLU - mean length of utterance measured in morphemes: NSST - Northwestern Syntax Screening Test: LIPS - Leiter International Performance Scale: TOLD - Test pf Language Development. 28 of human subjects by the Michigan State University committee on human subjects (see Human Subjects' Affidavit, Appendix B). The normal subjects tested at age appropriate levels on the Leiter Internal Performance Scale (LIPS), a nonverbal intelligence test, and had no reported history of speech-language impairment or therapy. Their Northern Syntax Screening Test (NSST) scores ranged from below the 10th percentile to between the 25th and 50th percentiles expressively and from below the 10th percentile to below the 25th percentile receptively. Their mean length of utterance (MLU) as measured in morphemes ranged from 3.7 to 8.0. Two of the language impaired subjects tested at age appropriate level on the LPS. One subject received a Leiter IQ score of 5:3 (approximately two years below his chronological age). The language impaired subjects' N§§T scores ranged from below the 10th percentile to between the 25th and 50th. percentile on both eXpressive and receptive measures. Their MLU scores ranged from 3.7 to 7.0. Language impairment status was judged by one school clinician whose language program the subjects were enrolled in. The clinician based her judgment partly on scores from the Test of Langggge Development (2922). Composite TOLD scores as reported by the school clinician ranged from 11 to 632 (mean - 181). Two of the subjects had some difficulty with tongue control on the oral peripheral examination. Previous history of ear infections ranged from one to four infections. It can be noted that there was a wide range of variability within each group and that group NSST scores overlapped. Given the 29 practical time span in which to choose subjects, children were chosen as they became available to participate. As a result, few of the identified children actually met criterion for participation in this study. Subject Selection Procedures Subject selection involved two phases of screening activity. The goal of the first phase was to identify clinical and nonclinical children who could serve as potential subjects. The goal of the second screening phase was to aid in selecting the most appropriate subjects. Three graduate students of Michigan State University's Speech-Language Pathology program were instructed about the screening procedures and shared in conducting the screening activity. Phase 1 of Screening Activity The first phase of the screening activity involved the classroom teachers and speech-language pathologists in the Lansing School District. (See Lansing Public School District Project Approval Form in Appendix C.) It was requested that they complete a brief questionnaire (see sample in Appendix D) on every child in the age ranges of interest. The classroom teachers restricted questionnaire information to children. who have never been referred to speech therapy or special education classes. The speech-language pathologists completed questionnaires for the language impaired children. The questionnaire required yes/no responses to questions about demograhic features (e.g., age, sex, educational placement, health, history, and socioeconomic status), professional judgments, and 30 access to the child's school records. From the questionnaire data, children were randomly selected for further screening in the second phase. The children's identity remained anonymous to the investigators responsible for determining the pool of potential subjects. Phase 2 of Screening Activity The second phase of screening activity involved the administration of the fig: and the L_I£_S_ (refer to test screening forms in Appendix E). The LIPS was administered because it (a) has been used frequently by investigators of language impairment, and (b) is assumed not to penalize intelligence scores. Therefore, the use of the LES in this study permitted the selection of subjects who were comparable with those used in other studies. The N§§T was used because it is a quick measure of receptive and expressive linguistic abilities. In addition, a language sample of at least 50 utterances was collected: and an informal observation of body structure integrity was conducted to assist in subject selection. The informal observation included the oral periphery, verbal and nonverbal behavior, such as speech intelligibility, vocal attributes, and response to verbal commands as determined by subjective judgments. A standard checklist was used for the purpose of applying the same core of observations to each child screened (see test screening forms in Appendix E). 31 The testing during this phase was conducted in the home environment. Written parental consent was obtained prior to investigator contact (see parental consent form. in. Appendix: F). Language impaired subjects were selected first. Then an attempt was made to select normal children who *were comparable, except for language status. Description of Stimuli and Tasks Responses were elicited under two conditions. First, a structured task was used to elicit locative responses in a standard way. Second, a language sample was obtained to observe spontaneous locative responses. The language sample was taken for the purpose of validating the results gathered from the structured task. In the following section, the structured elicitation task is described first, and the spontaneous language sampling procedure is described second. Operational Definitions of Locative Subcategories Targeted As reviewed in the literature, Stockman and Vaughn-Cooke (1983, 1984: cited in Stockman, 1985) described eight locative subcategories. Four of these subcategories were dynamic, and four were static in nature. Of these eight subcategories, five were targeted (static locative position, static locative perspective, dynamic locative origin, dynamic locative direction, and dynamic locative destination) during this investigation. Operational 32 definitions for the dynamic and static subcategories targeted appear in Tables 2 and 3, respectively. Description of Structured Task Twelve stimulus events were constructed to elicit spontaneous responses that reflected the five categories of dynamic and static locative meanings. The twelve events required the child to attend to a simulated landscape scene with movable objects that included a dog, a bone, and two to three doghouses. The landscape scene, made of plaster, contained a hill, a tunnel, and a river. Props, such as trees and fences, were included on the scene. This scene was mounted on a 32 x 48 plywood base, and was positioned on plastic boxes about 12 inches from the floor. This height allowed the child easily to touch and manipulate the props. The task required the child to attend to the stimuli placed on the landscape scene. He was asked to describe the locative position of the dog with respect to the doghouses and the bone in response to various questions that related to each stimulus event. The events were engineered to elicit particular types of locative responses as described in the succeeding sections. Criteria for Constructing Events Two factors were considered in structuring the events. First, the events were structured to evoke particular subcategories of locative utterances in response to standard questions. The question— response format is discussed in the next section. 33 .anqano>us: anonm ssnusuwz .sounsouuee< nssnu .so«oausnoo and: connsv< .ee .a .conennau nooncnnu ecu nqanoz no .gnmoo: uo eonsuunasu Hessian: o:n bu venuwsasm oceano=«uoun o>wuoooa .Anoo—v .n .H .colxooum "ounsom .uuo nu seen H .oao; lonu mauou one hush .onnon man so nu use .onono on» “on“ an on so» .onoen ecu noonuo nu use .ssov uswou one 50:9 .030; sonu added one hush .onOno can Aenv :«lmu.=o> .ssov usuou one mosh .on connouon onus no ocean ozn lenu has: on can no on spun-don o.n00qao noonouon uo suede vs- eoqnuonqo oak .ouqa no ocean cannouon can on vooenoouv on one an no soc-nonn: can no soda can no sedan-ea onossnoaao so on poo-unauv aw noonoo neononon och .ou vonnOuon sounoonuv newness onn nu o>ol sou noonno nnsonouon ecu no nounno can no use-uses on» Iona usovuso on on vonnouon ensue Assownoonav och .usoao nauseous ecu asunsv coo-moouv on and no poo-Hoouv on nuoaoo nsonouon ssh .oosao + sonu snag no counuonuo mo anon-on can as: assuage-bu o>unsuo~ urn unsolouoloo o>unouo~ + ono> saunas s sovsnuc« sons: olunu cannons»- . no one) o>wuooo~ caucus < .ooaao + on :uoo no souuoonuo no anon-cu unneeded as» as: assuage-bu o>nn000n osn unsoloaeloo sown-ooa + one: sounds a aovsnosu sous) ossnu onnoensao - no one: o>unaoo~ caucus < .sowuuonwv up anon-ow UdUBI‘UI USU Ila— up—gfldn—‘OU asuneoou 0;» muse-oualod o>nueoon + nno> sounds a novsuocq sons: sauna cannons»- o no one: o>unooou unwsgm .snOu accused onmsao a no scones s so as: .nsoIOHQIou noun-nod + ono> defines a advanced sous: osnnu bananas». no one: o>unoooH sumsum .voosnoouv on nuanao usonouon on» songs Iona ouaao on sound-on no condo can suuwnc aouossxu swnonano monumensou annonnno assessmena soundness oosonouon osuuoooa oases»: .uoonoo coo-neouv o no sown-annoov on» ..o.« "woo-Anode aw nounoo so songs on nodes no sound can sounssuuaon sequuooae ousonOuon o>unsuoq oases»: .nsoso>oa mo anon «own-no o moons noonao so no noun-usowno on» sounoonun nowuuoooo addenduon oswnsooa oases»: .nsoloaol elo- uo unseen s as noonoo so we newneuou osn ea sneeze ae>uuaooa sequuoooo ousonouon o>unoooa onsesaa anemoneo sounqsuuon «oceansnooo osnnsooa muoz~e0no can on one so» .oflpnn urn amends on on .ssov one hush .050: Bonu one hush .o>0no any on one so» .ssov one mosh .ousuo no suns vonnouun can no nouns sou an no soc-nanns ecu no ulna on» no onus o~oo>noooo so no venous uvsonna an noonoo neonOuon ssh .noxnsl some can an vouuuoono onus a no nouns use no venous on noonoo neonouon ssh .usoso monsoon. can asunse tonouodon no undead-«v no: on nuoqao unencuon ssh .oooun no onsnsOu onus-so- onn as: usosouasoo 0:» mucosoHnIOo sounded" + onus sonuossos a aovsuosu nous: sauna dune-usa- no one: dawn-cod o—usum .sonnensouno assounuonav was sownuooe up onsuo0u ouncosoa as» as: unassunIOU can uncosouAIoo osuuouoH + one: abundance s novsdosu sous: sauna aqueous»- no one: sewn-eon sausum .snou accused caucus no sconce a on no! nsoioqeloo osunsoo~ ecu “unassunloo o>unsuo~ + one: sequence: a newswosu sous: sauna canons»- no one: «squooOu ouwsum .eousuo~ nu noonoo so sons) no usuon nozoosga usunouxo ssh .uoonoo nuances no ends madness can no olnon sq vouuuoooa sown-now noun-«nu .souusuOu can an vonasaon sass has nosu use-obo- osn on oosonouon noosnws .uoonao no sewn-eon usunouxo soundness oosonOuon osunsuoa sauna-om ounonm o>unooaanom own-um ocean-ooa ouuium noun-awn onnonuno noonnOncou sunonwnu ounouswsum nonuncnnon n-uonuunoao uneven-u saunaooa muuz<¢nnnb u>uu_ .05) (refer to Appendix M for the raw data.) Temporal Reliability Table 16 shows the scores received during the first and second te sting sessions. The score ranges were 15-23 and 12-26, re spectively. Test-retest score differences were not statistically Significant (Senders, 1958, T - 28, T - 21, p Z .05) (refer to l 2 AP'P’eradix K for raw data). Table 16 82 Test-Retest Scores Obtained by One Nonclinical Subject Subcategory First Test Results Second Test Results SPOS 22 26 DO 20 22 DDEST 22 22 DDIR 15 12 SPERSP 23 18 N253: SPOS - Static Position DO - Dynamic Origin DDEST - Dynamic Destination DDIR - Dynamic Direction SPERSP - Static Perspective CHAPTER V DISCUSSION It was hypothesized that semantic knowledge and nonverbal knowledge were linked. The literature review showed that cognitive knowledge had been shown to differ among the subject groups (wain & Yule, 1983; Terrell et al., 1984; Johnston & Savich, 1984; Johnston & Weismer, 1983; Wren, 1982). To the extent that semantic knowledge indexes cognitive representation, it follows that semantic differences may be expected between normal and language impaired groups. Group differences, however, were not observed in studies that have compared the semantic relations expressed by normal and language impaired children (Freedman 8: Carpenter, 1976; Duchman 8: Erickson, 1976; Leonard et al., 1976; Coggins, 1979; Fokes 5x Konefal, 1981). This study was motivated by the argument that the global nature of the semantic categories tested was not sensitive to the differences. It was argued further that if semantic categories were more specifically defined, possible differences would be observed. The purpose of this study was to pilot a structured elicitation task for comparing language impaired and normal groups locative utterances. It was predicted that group differences would not be observed when locative categories were globally defined as dynamic 83 84 and static locatives, but that differences would be observed in the distribution of locative subcategories SPOS, SPERSP, DO, DDIR, and DDEST within the global categories of dynamic and static locative reference. As predicted, no significant differences between the groups were demonstrated at a global level. However, unlike what was predicted, group differences still were not observed at a specific level of locative knowledge. A secondary goal of this study was to determine if the structured task elicited locative responses comparable to a spontaneous language sample. To accomplish this, the data gathered from the structured task were compared with those gathered from the language sampling. The data comparison yielded similarities and differences. Both procedures were similar in that they revealed subjects' use of dynamic and static utterances. They differed in the kinds of locative subcategories revealed and in some of the targeted words elicited within a subcategory. The accuracy of the structured task for eliciting locative knowledge was gauged by a temporal reliability check. This reliability check was accomplished by administering the structured task twice to the same subject on separate occasions. Results from both sampling forms were comparable. This chapter addresses why subject group differences were not revealed and why the structured and spontaneous elicitation procedures showed particular differences and similarities within 85 their data” In addition, suggestions for improving the structured task are offered. Explanations for Subject Group Differences Not Revealed Among the Subcategories of Locative Knowledge Three reasons are offered for the failure to reveal group differences among the subcategories of locative knowledge. These involve alternate theoretical hypotheses, subject group characteristics, and the nature of the targeted locative subcategories. The following sections discuss these proposed reasons. Alternate Hypotheses In the absence of observed group differences, we are led to conclude that language impaired children are not semantically deficient for locatives. Consequently, we must entertain the notion that: (1) language impaired children's cognitive deficiency is not at the base of their language impairment, or, (2) language impaired children are cognitively deficient in ways that do not pertain to semantics. More research needs to be conducted focusing on either the semantics of language impaired children or on language impaired children who exhibit specific cognitive deficits. Subject Characteristics Subject characteristics may also explain why the data did not reveal group differences. While an attempt was made to create homogeneous language impaired and nonimpaired groups, subjects in 86 each group exhibited wide variability in their screening test scores. These variations were reluctantly accepted after realizing that the time available for the study was a limiting factor. Despite efforts to find children from all available channels, few were available to select for testing purposes. Heterogeneity"was observed in both. cognitive and linguistic skills. Some of the children in the language impaired group demonstrated higher than expected linguistic skills and lower than expected Leiter IQ scores. On the other hand, variables could also have entered into the test scores obtained through the school clinicians where no control over the types of tests used to determine pathology nor over the conditions for testing the children.‘was present. The children in the normal group were also heterogeneous. Some scores from this group were lower than expected: for both the linguistic and cognitive skills. The two group characteristics merged as the reduced scores from the normal group combined with increased scores from the language impaired group. As group characteristics overlapped, distinguishing characteristics *were minimized. Similar subject group characteristics would be expected to yield similar group performances on the locative tasks of this study. Nature of the Targeted Locative Subcategories It was hypothesized that group differences would be revealed if semantic concepts were tested at a.specific level. Since the five 87 targeted locative subcategories were assumed to be specific in nature, group differences were expected to be observed. However, group differences may not have been revealed because the targeted subcategories were still too broadly defined to capture more finely tuned differences. That is, smaller divisions may exist within the locative subcategories. For example, just as ”Put it here” codes a destinative position in a dynamic sense, it may also code a more specifically defined subcategory within a dynamic destinative sense. Such finely tuned subcategories of locative utterances, however, need to be identified. Explanations for Why the Structured and Spontaneous Elicitation Procedures Showed Particular Similarities and Differences The results gathered from the structured task and the language sample were compared to determine the accuracy' with which. each elicited the targeted lexicon and the general and specific locative categories. The comparison revealed similarities and differences among the responses to the two tasks. Differences occurred in the kinds of locative subcategories evidenced and their respective targeted words elicited. It is difficult with the small number of subjects participating in this study to determine the accuracy of these results or to draw strong conclusions from them. However, plausible explanations supporting the conclusions are offered. The first explanation attempts to reason why differences were observed within the kinds of subcategories of locative knowledge evidenced” This explanation 88 involves the number of opportunities available for the subjects to respond. The more opportunities available to exhibit knowledge, the higher the probability that knowledge will be exhibited. Consider the following hypothetical example of a child who has SPOS locative knowledge. Two investigators attempt separately to elicit this knowledge from a child. The first investigator, who positions 100 identical objects on top of identical toy houses and questions the child 100 different times about the object locations, is more likely to reveal the child's knowledge than the second investigator who has only six of the same stimuli for questioning the child. This same idea can be related to the DDIR, DO, and SPERSP locative subcategories. For each subcategory, more opportunities were available for the children to demonstrate their knowledge during a particular task. The language sampling provided more opportunities to reveal DDIR locative knowledge, whereas the structured task offered more for revealing DO and SPERSP locative knowledge. On the structured task, prompt questions created opportunities to respond within a particular locative subcategory. Approximately 12 to 18 opportunities to respond to questions 'within the DDIR context existed for the subjects during the structured testing. But it appeared that the language sample appeared to offer even more opportunities. For example, some of the common questions heard from the investigators during the language samples were ”Where are they going," ”Where are they driving,” and ”Where are they walking?” It 89 seemed that the language sample was naturally facilitive to the DDIR locative responses. However, the opposite seemed apparent ~for the SPERSP and DO locative subcategories. While only 12 to 18 opportunities were available for the subjects to respond during the structured task, there were fewer opportunities for them to respond during the language sampling. Because fewer spontaneous prompt questions were used, the number of response opportunities were reduced. It appeared that increased structure of the environment led to a greater number of prompt questions for the SPERSP and DO subcategories. Therefore, the probability that responses would occur in these locative subcategories is increased. One difference was observed among the targeted words elicited between the structured and spontaneous elicitation procedures. Within the three positional subcategories (DO, SPOS, and DDEST), the structured task elicited the word between 671 of the time and the language sample elicited it 111 of the time. The mean difference between the two procedures in eliciting this word was 56. This large difference suggests that the structured task is more accurate than the language sampling in eliciting between within the three positional categories. This may be due to the spontaneity of the language sampling which seldom offered opportunity for such a word to occur. It appeared that the environment needed to be purposefully planned to foster such a response. 9O Similarities also existed among the targeted words elicited between the structured and spontaneous elicitation procedures. For the two locative subcategories targeting directional words (DDEST and SPERSP); Mg was not elicited by either procedure. Also, for the SPERSP subcategory specifically, through was not elicited by either procedure. There are several possibilities why such similarities would be observed. One possibility involves the normal developmental aspect for acquiring these words. These words may not be fully developed by the age of 7 or they may exist only within restricted contexts. Such restricted contexts were evidenced when several of the subjects were questioned in a nonstandard way following the testing session. Although they did not produce the word 21228 during the structured task, they all were able to move the dog along the river upon request. The possibility exists that these children had knowledge about m only under very specific conditions and could not generalize 2.13.8 to more abstract situations. The more abstract conditions probably were present during both procedures, causing these words (along and through) not to be reflected within either set of data. Recommended Changes for the Structured Task Procedure Following the administration of the structured task and upon reviewing the results, procedural changes are recommended in addition to those previously discussed. These recomendations change the 91 scoring system and the test presentation format used. The changes are thought to make the testing procedures more succinct and more reflective of locative knowledge. Recommended Changes for the Scoring System The existing scoring system was developed for recording all targeted subcategories in the same format. However, upon reviewing the data, it appeared that all the subcategories could not conform to the one format. In particular, the breakdown in this format occurred with the allowance of points for responses using general locative words such as M and M in contrast to specific words, such as g and BM- Such responses would be categorized within the General Locative to the Targeted Subcategory response category. It is argued that such general lexicon does not exist for the directional locative subcategory. The basic definition of a general locative could not logically apply to the direction of the movement, since any direction by definition is specific. The Webster definition (Webster's New Collegiate Dictionary, 1979) of direction is ”the line or course on which something is moving or is aimed to move or along which something is pointing or facing” (p. 320). To 8emantically convey such a definition, a specific locative must be uSued. To support this argument, data were reviewed for general locatives elicited in a directional sense. It was found that no such 1°<=atives were elicited. An attempt was made by the investigator to 92 arrive at an existing directional general locative using all available information. No general locative could be observed. Considering this notion, it seems moot to appropriate points to a response category which does not exist. It would seem more appropriate to modify the scoring format for the directional subcategory such that no points are allocated to a General Locative Response Category. The observation did not bias the comparisons made with the test scores because they occurred only within a particular subcategory (e.g., "intra-subcategory analysis") not between the subcategories (e.g., "inter-subcategory analysis”). This latter type of comparison is discouraged since the two subcategories are qualitatively different. A further modification of the present scoring system would be to appropriate points which are sensitive to variations in responses across targeted locative words within the SPERSP locative subcategory. For example, a subject may use the response ”over there" to convey each targeted word for the SPERSP locative subcategory. While this response is correct, it lacks variety and does not convey the changing perspective occurring with each event. In contrast, another subject responds with ”Over the grass," ”Over the river," and "Over the fence” for three separate targeted locative wOrds. These responses vary in that they identify the object that defines the perspective in each event. The present scoring system records these responses as equal in numeric value. However, since the essence of the SPERSP locative 93 subcategory is change in perspective, it is suggested that responses such as these be differentiated in terms of quality and numeric value. This differentiation should be conveyed through the scoring system. In sum, it seems appropriate to use the same scoring format for the positional locative subcategories only (SPOS, DO, and DDEST). Because of the nature of directionality, the remaining two subcategories (SPERSP and DDIR) should be scored apart from the positional locative subcategories. In addition, each of the two directional subcategories should have a separate scoring format. This is because the SPERSP locative subcategory has changing perspectives as part of its nature, whereas the DDIR subcategory does not. Recommended Changes to the Test Presentation Format To make the presentation possibly more succinct, it is proposed that the underlying format for presentation be modified. At present, this basic format consists of targeting four locative subcategories within one event. This format is present for all of the six A events. Such a design was in the interest of minimizing time for administering the task. However, in retrospect, it is questionable whether such a format served this purpose. It is necessary within Such a format to present multiple stimuli in one event. That is, one movement transports the dog from an originating location, in a Particular direction, to a destinative position. After such a tho"-?ement, the subject is questioned about all three aspects of this 94 movement. At this point, the subject is expected to remember all aspects of the movement and relate the prompt questions to the aspect. Breakdown is possible within the recall process or the ability to relate such questions to the appropriate aspects of the movement. Three solutions to this problem are proposed. One is to introduce a training event prior to testing, a second is to interject questions at different times, and the third is to reduce the number of stimuli presented within each event. The training event would serve as an example for the child to learn the relative nature of the responses. To reduce training specific locative words, words occurring later in the task may be used within this event. It would be improbable that the subject would remember the target locative words by the last event if he did not already have them within his linguistic system. I For this structured task, for example, 93, M, and through were the last words to be targeted. For the training event, the dog and bone should be set in a manner to elicit these words and the prompt questions should be asked. What would be significant about this training event would be the treatment of inaccurate answers. As the subject responds with an answer not relative to the targeted response, inaccurately, or insufficiently, the investigator would be allowed to inform the subject what answer is expected. For example, consider the following situation where the prompt question, "Where did the dog come from?” is asked and the response 95 is, ”From his house.” The investigator would then be allowed to say, ”That's correct, but I'm looking for a more specific answer like, 'The dog is o_n his house." After this training example, the events would be followed in the original manner. Another way to aid the subject in understanding what aspects of the event are important is to change the time at which the questions are interjected. The structured task was designed such that all dynamic questions were asked after the movement. However, it may have been helpful within the DDIR subcategory to interject a prompt question while the movement was occurring. At this moment, the question, "Where is the dog walking?” (asked to elicit the DDIR subcategory) would offer a more concrete referent for responding to the questions. Further investigation would need to be conducted in this area to determine the effectiveness. A final suggestion to aid the subjects in understanding what aspects of the event are important is to reduce the amount of stimuli presented at one time. To accomplish this, it is suggested that only one locative subcategory be presented within each event. For example, one series of events could be designed for the purposes of targeting the DDEST locative subcategory only. Stimuli could consist Of a dog and a house separate from the outdoor setting. The dog for each event is moved to the house in a straight path. The dog's Originating position would be constant for each event. However, the destinative position would be variable. The sole varying aspect with each event is related to the locative word targeted. 96 For example, for one event the dog is moved to the top of the house and the subject is asked, ”Where did the dog move to?” This event would target the word 93. The next event consists of the dog's moving under the house. The same question is again asked for purposes of eliciting the word 92513;. With only one aspect of the motion changing with each event, the targeted stimuli could become more salient for the subject. Therefore, the subject can more easily relate the prompt question to the appropriate aspect of the movement and respond appropriately. It cannot be calculated whether such a format will be more time consuming. The existing format typically takes 45 minutes to an hour to administer. Only further research with this format can estimate the time needed for testing. A final modification recommended is to change the stimuli presented. It has been noted that part of the stimuli used for the structured task consists of a set simulating on outdoor setting. Designed into this setting were obstacles (e.g., a hill, a tunnel, and a river) meant to facilitate directional locative words. However, upon reviewing the data, particular directional locatives Words were not elicited often relative to the remaining directional locative words. These words were mg and around and may not have been elicited because of the obstacles designed to facilitate their Production. To facilitate the production of glppg, a river was placed in the natural setting. This river was fairly long and winding. The 97 assumption was that the subject would walk the dog along the river bank and comment on this motion upon questioning. However, no subject was observed to move the dog in this manner. Instead, the subject often took a direct path from one end of the river to the other. Therefore, instead of following the winding bank, the dog was often observed to go over the grass. Indeed, this was the directional response frequently observed among the subjects (e.g., ”The dog walked over the grass”). It is proposed that the river be redesigned such that moving the dog along the river is a natural inclination. Such a river may consist of just one large curve which would force the subject to walk the dog around its outer circumference, and therefore, along the river bank. To facilitate the production of around, a hill was placed on the set. The hill was to provide a curvature which the dog was to walk around. In fact, the design of the hill accomplished its purpose and all the children walked the dog around a bend of the hill to reach the bone. However, this movement did not seem to facilitate the production of around. It is hypothesized that possibly the motion of walking around the hill was not salient enough to elicit such a response. This may, in part, be due to the circumference of the hill at its base. This circumference is small relative to the motion needed to move the dog around it. It may be beneficial to increase the diameter to the Point where moving the dog in a curving motion is clearly salient. APPENDICES APPENDIX A DESCRIPTION OF SUBJECTS APPENDIX A DESCRIPTION OF SUBJECTS Following is a detailed description of the language impaired and normal subject groups. Their family, medical, and educational histories are foregrounded, and their screening results and any significant observations are presented. Language Impaired Subjects Subject 1A: CC Age at Testing: 6:10 Background History: CC is the oldest of two children. He experienced two ear infections during his childhood to date. Medication was administered for one to two weeks during the infections. It is unknown whether pressure equalization tubes were used. CC has Hay Fever and uses medication to relieve the symptoms. Family history of speech disorders was reported. Other medical and developmental history is unremarkable. Speech and Language History: It was reported that CC began using first words meaningfully at the age of three. The use of two- word sentences began at four years old. CC was enrolled in therapy through his school district for approximately two months. It was reported that this therapy focused on articulation. CC's mother reports that he has difficulty understanding instructions and believes this difficulty is persistent during school. Screening Results: MLU (in morphemes): 4.0 Observation Checklist: Not remarkable Oral Peripheral: Diadochokinetic rates slightly below average 98 99 20 Repetitions 10 Repetitions [p ] - 8 seconds [p t k ] - 14 seconds It ] - 8 seconds (this sequence difficult [k ] - 9 seconds for CC to perform) NSST: Expressive Score: 5/22--below 10th percentile Receptive Score: 14/22--at 10th percentile LIPS: tested through at least age 6:10 Responsiveness to Testing (subjective): average School Test Results: Test of Language Develgpment (TOLD) sentence imitation--2nd percentile word discrimination--9th percentile grammatic closure--9th percentile grammatic completion-—2nd percentile oral vocabulary--37th percentile picture vocabulary--lst percentile composite scorers: syntax--66 speaking-~72 spoken language—-67 listening--67 semantics-- 76 Subject 2A: MF Age at Testing: 6:7 Background History: MP is the oldest of three children. He experienced one ear infection as an infant. Medication was administered for one week during the infection. MF was described as a late starter in relation to his general motor development. Other medical and developmental history is unremarkable. Speech and Language History: It was reported that HP began using first words meaningfully at the age of 1:5 years. However, his talking stopped for a period, and he began talking again at the age of 2:0. Two-word phrases began at 2:6 years. It was reported that MF stutters ‘when he is upset. MF has been enrolled in therapy since kindergarten. Educational History: MF was enrolled in Head Start for one year. He attended kindergarten for two years. 100 Screening Results: MLU (in morphemes): 7.0 Observation Checklist: not remarkable Oral Peripheral: difficulty following model for tongue protrusion, retraction, elevation, depression, and lateralization. NSST: Expressive Score: 18/22--between 25th and 50th percentile Receptive Score: 17/22--between 25th and 50th percentile LIPS: tested through at least age 6:7 Responsiveness to Testing (subjective): good School Test Results: Test of Langppge Development (TOLD) sentence imitation--9th percentile word discrimination--l6th percentile grammatic closure--25th percentile grammatic completion-~25th percentile oral vocabulary--50th percentile picture vocabulary--63rd percentile composite scores: syntax--83 speaking--87 listening--97 semantics--103 Subject 3A: 88 Age at Testing: 7:4 Background History: SS is the second oldest of four children. He experienced numerous ear infections between the ages of one and three years. Pressure equalization tubes were placed in his ears at 6 months of age. Medication was used to clear his infections. SS was stressed during the birth process. However, no drugs or instruments were used. SS was slow to meet general motor development milestones as described by his mother. It was reported that SS appears to be uncoordinated with his fine motor skills and it is observed that his mouth opens while concentrating on fine motor tasks. At the ages of 4 months and 7 months, SS experienced 105 degree fevers. Tylonol was administered to relieve the fevers. His fever at 7 months was associated with seminella poisoning. SS also has suffered form chicken pox. 101 Educational Historya SS has attended school since 2 years of age. During his kindergarten year, he attended regular education for half a day, and special education for the other half. Speech and Language History: It was reported that $8 began using first words meaningfully at 8 months, and talked in two-word phrases shortly after reaching 2 years of age. Speech learning appeared to stop between the ages of 2 and 4, it was commented that he did not have clear pronunciation during this period. SS has been in speech therapy since the age of 2. Screening Results: MLU (in morphemes): 3.68 Observation Checklist: not remarkable Oral Peripheral: not remarkable NSST: Expressive Score: 3/22--below 10th percentile Receptive Score: 12/22--below 10th percentile LIPS: Basal: III Ceiling: VII IQ: 5:3 Interpretation--It is suspected that his lowered IQ score reflects visual perceptual impairments rather than mental disorder. Responsiveness to Testing (subjective): good School Test Results: Test of Langppge Development (TOLD) sentence imitation--lst percentile word discrimination--5th percentile grammatic closure--16th percentile grammatic completion--l6th percentile oral vocabulary--37th percentile picture vocabulary-~5th percentile composites scores: syntax--72 speaking--76 spoken language--74 listening--76 semantics--82 102 Normal Control Subjects Subject 1B: BD Age at Testing: 6:9 Background History: BB is the oldest of two children. BD suffered from Chicken Pox at the age of 5 years. Other medical and developmental history is unremarkable. Speech and Language History: BD is reported to have developed age appropriate speech and language skills. Screening Results: MLU (in morphemes): 3.7 Observation Checklist: not remarkable Oral Peripheral: not remarkable NSST: Expressive Scores: l6/22--25th percentile Receptive score: l6/22--< 25th percentile LIPS: Tested through at least age 7:3 Responsiveness to Testing (subjective): extremely shy and quiet Subject 2B: SR Age at Testing: 6:11 Background History: SR is the second of three children in his family. SR suffered from Chicken Pox at the age of 3 months. Other medical and developmental history is unremarkable. Speech and Language History: SR is reported to have developed age appropriate speech and language skills. Screening Results: MLU (in morphemes): 8.0 Observation checklist: not remarkable Oral Peripheral: enlarged tonsils NSST: Expressive Score: 18/22--> 25th percentile Receptive Score: 15/22--< 25th percentile LIPS: tested through at least age 7:3 Responsiveness to Testing (subjective): average 103 Subject BB: JS Age at Testing: 7:5 Background History: JS is the second oldest of four children in his family. .78 suffered from Chicken Pox at the age of 6 years. Other medical and developmental history is unremarkable. Speech and Language History: JS is reported to have developed age appropriate speech and language skills. Screening Results: MLU (in morphemes): 7.4 Observation checklist: not remarkable Oral Peripheral: Not remarkable NSST: Expressive Score: l7/22--below 10th percentile Receptive Score: 12/22--below 10th percentile LIPS: Tested through at least age 7:6 Responsiveness to Testing (subjective): good APPENDIX B HUMAN SUBJECTS AFFIDAVIT 10‘: MICHIGAN STATE UNIVERSITY L'Mvusm LUMMImt on mac" mvowmc. usr MKSWG . mcmm” . “““M mam Sl'BjEG'S mums: m anumlsmnos IUIIDING (“"5”""6 December 1, 1986 Dr. Ida J. Stockman Audiology & Speech Sciences 378 Communication Arts Building Dear Dr. Stockman: Subject: Proposal Entitled, "Locative Distinctions of Clinical and Normal Children" UCRIHS' review of the above referenced project has now been completed. I am pleased to advise that since the reviewers' comments have been satisfactorily addressed, the conditional approval given by the Committee at its November 3, 1986 meeting has now been changed to full approval. You are reminded that UCRIHS approval is valid for one calendar year. If you plan to continue this project beyond one year, please make provisions for obtaining appropriate UCRIHS approval prior to November 3, 1987. Any changes in procedures involving human subjects must be reviewed by the UCRIHS prior to initiation of the change. UCRIHS must also be notified promptly of any problems (unexpected side effects, complaints, etc.) involving human subjects during the course of the work. Thank you for bringing this project to our attention. If we can be of any future help, please do not hesitate to let us know. Sincerely, C., Henry E. Bredeck, Ph.D. Chairman, UCRIHS HEB/jms APPENDIX C LANSING PUBLIC SCHOOL DISTRICT PROJECT APPROVAL FORM 105 11, LANSING scuiqggp DISTRICT Committed to Quality November 25 , 1986 Ida J. Stockmn, Ph.D. Associate Professor Department of Audiolog & Speech Sciences 378 Coumunication Arts & Sciences Building Michigan State University East Lansing, MI ”88le Dear Dr. Stockmn: In regard to your research study titled, "Locative Distinctions of Abnormal and Clinical Children", the request to conduct the study in the Lansing School District has X been approved, has not been approved. The following cements apply to your study: Any teacher participation must be clearly voluntary. Written parent permission for student involvement must be on file in the (each) school. If you have any questions or need additional information, please contact me (374-4347). “999...... Pat P ersen FNaluation Specialist PP/mlc cc: Research Review Cosmittee Manbers Research & Evaluation Services Office 500 W. Lenawcc St. [amino Michigan 48933 APPENDIX D TEACHER QUESTIONNAIRE FORMS 106 .GOHumuwmoou use» How 50% xcwzH coaumauowcfi Ham umzu umusmmm on has now .Hmauawwawcou uawx on Haas so» hp umvfi>oun .afico mamauaca an vancu none sensuous .xuwsu m mo wmmumcfi muwnm mcu ca va xumE :owummac m mumHn .muowmumo m .auowmuwu umzu a“ ucmamvsfi “50% mafiumu«vafi .Eooummmao use» 6H vafino some wmvan ca 30: usonm cfimuuwuaa mum so» MH woman wumfiunonamm ecu ca Axv Moose m mumHa mwmwam no mwfiuowmumu mawaoHHOM ecu co coaumauowcfi wvH>oum cu vwxmw one now I. I: .3: x . :2» e: 5.5: :3. 107 N Caz; LaSG Ilull ~33£IH I .. 3.91:... ‘E 108 , I figs: III IIIHIIII I _=¢.-...m III Lap—usih .Hmauawnfiwaou was: on Haaa so» we vmufi>oum .cho Hmfiuacfi mam: umHH was wane umuwu an vafiso some amauamua mmmoam .wumnw wan :H ANV a woman .uamawwsn :fimuumu m ufisuwn uo: ow mvuoouu Hoonum «H .aofiumuwnooo Mao» How so» Janna coaumauomaw HH< .uumawvsfi uaoh ou wafivcoammuuou momma wnu :H nxv m momam .maoanoun cowumasuwuum How umsh hamnmcu w>amomp 0:3 cmuvaaco owDHUCH uoc waoosw umfia mask .manmufiuu coauuwaum uuwnASw H30 woos HHH3 wam 50% 0:3 vmonmmo “30% no :wuvaazo umuamaafi mwmawama mnu new mmwuomwumu waa30HHow unu :H coaumfiuomafi mwfi>oua ou vmxmm mum so» uwmmmmwmmmwmw 109 APPENDIX E TEST FORMS USED DURING THE SCREENING 110 APPENDIX E TEST FORMS USED DURING THE SCREENING Observation Checklist 1. Ability to formulate and produce the CV syllable /pA/, /tA/, and /kA/. 2. No extraneous vocalizations present during speech. 3. Points to nose, leg, and chair when asked a ”where” question. 4. Absence of drooling. 5. Symmetry of the facial features. 6. Symmetric mouth retraction. 7. Symmetric tongue protrusion. 8. Ability to lateralize tongue and move tongue up and down. 9. No obvious paralysis of muscles. 10. Normal gait. 111 £65 :3 dune—£0 .60 .6300; .3 .0 .< d .D a. $85.5 .583 .3- .3 .......I 2x .......I. III=x III Ix ..II.. 5 III IIIA—S II => II II; .II II...> I... I2 I IIIE . II. = 3:23 ”am. I Bears 59m as? 2...: as» . zoEmuueamcea comu.moa a roasts» oco.m stop a: museum canoes cups: nausea «not» so cm "a commmmm Euom muoom mHaEmm owmswsmq m NHszmm¢ ”ocmumm» ca name some ”memz u f b D>Zdt~ Ult-dO-fi-‘U APPENDIX I SUBJECT SCORES FOR INDIVIDUAL TARGETED LOCATIVE WORDS 119 Table I.1 Subject Scores for Individual Targeted Directional Locative Words Numeric Value for Each Targeted Word for the Subject sub5°°t STATIC PERSPECTIVE LOCATIVE Subcategory* Total around down up through across along LI Group: Sla 4A 2 2 2 5 2 17 82a 4A 5 5 2 5 2 23 53a 2 2 2 2 2 2 12 Group Total 52 Mean 17 N Group: 51b 2 2 2 2 2 2 12 82b 2 AB 5 2 5 4A 22 83b 5 2 5 4A 2 AB 3; Group Total 56 Mean *Numeric Values Assigned to the Response Categories 0 - No Verbal Response/ Nonlocative Response - Incorrect Locative Response - Nontarget Locative Subcategory Response - General Locative Response to the Targeted Subcategory - Nontarget Locative Response to the Targeted Subcategory Semantic Sense Response to the Targeted Subcategory - Targeted Locative Word Response m‘FwaH >w I 120 Table 1.2 Subject Scores for Individual Targeted Directional Locative Words Numeric Value for Each Targeted Word for the Subject sub5°°t DYNAMIC DIRECTION LOCATIVE Subcategory* Total around down up through across along LI Group: Sla 4A 4A 4A 4A 5 4A 25 52a AB 5 5 5 2 2 23 83a 4A 5 5 2 5 2 23 Group Total 71 Mean 24 N Group 81b 2 2 5 2 2 2 15 82b 4A 5 5 5 5 2 26 83b 2 5 5 5 5 4A 29 Group Total 67 Mean 22 *Numeric Values Assigned to the Response Categories 0 - No Verbal Response/ Nonlocative Response - Incorrect Locative Response - Nontarget Locative Subcategory Response - General Locative Response to the Targeted Subcategory — Nontarget Locative Response to the Targeted Subcategory Semantic Sense Response to the Targeted Subcategory - Targeted Locative Word Response U'IJ-‘4FwNi-s >13! I 121 Table I.3 Subject Scores for Individual Targeted Locative Positional Words Numeric Value for Each Targeted Word for the Subject Subject STATIC POSITION Locative Subcategory* Total front in behind under between on LI Group: Sla 3 5 5 1 4A 5 23 82a 5 5 5 5 5 5 30 83a 4A 5 4A 5 5 5 gg Group Total 81 Mean 27 N Group: 81b 3 5 3 3 5 3 22 82b 3 5 3 5 5 5 26 83b 3 5 S 5 5 5 gg Group Total 76 Mean 25 *Numeric Values Assigned to the Response Categories 0 No Verbal Response/ Nonlocative Response Incorrect Locative Response Nontarget Locative Subcategory Response General Locative Response to the Targeted Subcategory Nontarget Locative Response to the Targeted Subcategory Semantic Sense Response to the Targeted Subcategory Targeted Locative Word Response UII-‘J-‘WNI— >u: I 122 Table 1.4 Subject Scores for Individual Targeted Locative Positional Words Numeric Value for Each Targeted Word for the Subject Subject DYNAMIC ORIGIN Locative Subcategory* Total front in behind under between on LI Group: Sla 5 5 3 3 3 5 24 82a 3 5 5 3 5 5 26 83a 3 5 3 3 5 S 23 Group Total 74 Mean 25 N Group Slb 3 3 3 3 S 3 20 82b 3 5 3 S 5 1 22 83b 4A 5 5 3 5 5 21 Group Total 69 Mean *Numeric Values Assigned to the Response Categories 0 No Verbal Response/ Nonlocative Response Incorrect Locative Response Nontarget Locative Subcategory Response General Locative Response to the Targeted Subcategory Nontarget Locative Response to the Targeted Subcategory Semantic Sense Response to the Targeted Subcategory Targeted Locative Word Response UII-‘l-‘th-o >5 I Table 1.5 123 Subject Scores for Individual Targeted Locative Positional Words Numeric Value for Each Targeted Word for the Subject Subject DYNAMIC DESTINATION Locative Subcategory* Total front in behind under between on LI Group: 81a 3 5 3 3 3 5 22 82a 3 5 3 5 5 2 23 83a 3 5 3 3 3 3 29 Group Total 65 Mean 22 N Group Sla 3 5 3 3 5 3 22 52b 3 5 3 3 5 3 22 83b 3 3 5 5 5 5 26 Group Total 70 Mean 23 *Numeric Values Assigned to the Response Categories 0 mbbwn» >U No Verbal Response/ Nonlocative Response Incorrect Locative Response Nontarget Locative Subcategory Response General Locative Response to the Targeted Subcategory Nontarget Locative Response to the Targeted Subcategory Semantic Sense Response to the Targeted Subcategory Targeted Locative Word Response APPENDIX J RANK ORDERING OF GROUP MEAN PERCENTAGES AVERAGED ACROSS ALL FIVE TASKS 124 Table J.l Rank Ordering of Group Mean Percentages Averaged Across All Five Tess Group Mean Rank Group LI Group N Group Percentages 22 1 LI 1 28 2 N 2 33 3 LI 3 39 4.5 LI 4.5 39 4.5 N 4.5 44 6.5 LI 6.5 44 6.5 N 6.5 50 8 N 8 61 9 N ’ 9 67 10 LI 10 Tl-ZS T2-30 n1- 5 n2- 5 Note. The percentages reflect the ratio of correct responses to the total number of response opportunities. N - Normal Control LI - Languagee Impaired Source: Senders, 1958, pp. 437-440. APPENDIX K RAW DATA USED FOR THE RANK SUMS TEST (SENDERS, 1958) FOR STRUCTURED TASK MEAN SCORES BETWEEN THE SUBJECT GROUPS AVERAGED ACROSS ALL FIVE TASKS 125 Table K.1 Raw Data Used for the Rank Sums Test (Senders, 1958) for Structured Task Mean Scores Between the Subject Groups Averaged Across All Five Talia Mean Scores Rank Group LI Group N Group 17.3 1 LI 1 19 2 N 2 21.7 3 LI 3 22.3 4 N 4 22.7 5 N 5 23.3 6 N 6 24 7.5 LI 7.5 24 7.5 LI 7.5 25.3 9 N 9 27 10 LI lO T1-29 T2-26 n1- 5 n2- 5 Note. LI - Language Impaired N - Normal Control APPENDIX L DISTRIBUTION OF LOCATIVE WORDS ELICITED FOR THE FIVE TARGETED LOCATIVE SUBCATEGORIES DURING THE STRUCTURED AND SPONTANEOUS LANGUAGE SAMPLING TASKS Table L.1 126 Distribution of Locative Words Elicited for Static Position Locative Subcategory During the Structured and Spontaneous Language Sampling Tasks Front In Behind Under Between On Structured Task Data Language Impaired Group S a - + + - - - 1 Sza + + + + + + S a + - - + + + 3 Normal Group Slb - + - - + - S b - + - + + + 2 S b - + + + + + 3 Total Mean 33: 832 so: 671 ' 831 67: Language Sample Data Language Impaired Group S a - + - - - + 1 82a + + + — - - 83a - + - + - + Normal Group S b + + + + - - l S b - + + + - + 52b - + - - - + 3 Total Mean 33! 1001 501 502 01 671 Note. + - Target Locative Elicited - - Targeted Locative Not Elicited 127 Table L.2 Distribution of Locative Words Elicited for Dynamic Origin Locative Subcategory During the Structured and Spontaneous Language Sampling Tasks Front In Behind Under Between On Structured Test Language Impaired Group S a + + - ~ - - l S a - + + - + + 82a - + - - - + 3 Normal Group 51b - - - - + - 82b - + - + + - 83b - + - - + + Total Mean 171 672 171 171 ' 671 501 Language Sample Language Impaired Group 8 a - + - - - - l S a - + - - - + 82a - + - - + + 3 Normal Group 82b - + - - - + 82b - + - - - - 83b - - - - - + Total Mean 01 1001 01 02 171 671 Note. + - Target Locative Elicited - - Targeted Locative Not Elicited 128 Table L.3 Distribution of Locative Words Elicited for Dynamic Destination Locative Subcategory During the Structured and spontaneous Language Sampling Tasks Front In Behind Under Between On Structured Task Data Language Impaired Group S a - + - - - + 1 S a - + - + + + 82a - + - - - - 3 Normal Group Slb - + - - + - S b - + - - - + 2 S b - - + + + + 3 . Total Mean 02 831 171 331 501 501 Language Sample Language Impaired Group Slb — + + - - + S b - + - + - + 82b - + - + - + 3 Normal Group S b + + + + + 1 82b - + - + - + 83b - + - - - + Total Mean 17! 1002 331 672 171 1002 Note. + - Target Locative Elicited - - Targeted Locative Not Elicited Table L.4 129 Distribution of Locative Words Elicited for Dynamic Direction Locative Subcategory During the Structured and Spontaneous Language Sampling Tasks Along Up Around Down Across Through Structured Task Data Language Impaired Group Sla - - - - + - 823 - + - + - 4- 53a - + - + + - Normal Group Slb - + - - - - S b - + — + + + 2 S b - + — + + + 3 c Total Mean 01 832 01 671 672 50% Language Sample Data Language Impaired Group S a — + - + + - 1 82a — - - + + - S a - + + + + - 3 Normal Group 8 b - + + + + + 1 S b — + - + + + 2 S b - + - + + - 3 Total Mean 02 831 331 1001 1001 331 Note. + - Target Locative Elicited - - Targeted Locative Not Elicited 130 Table L.5 Distribution of Locative Words Elicited for Static Perspective Locative Subcategory During the Structured and Spontaneous Language Sampling Tasks Along Up Around Down Across Through Structured Task Data Language Impaired Group Sla - - - - + - Sa - + - + + - Sza - - - — - - 3 Normal Group 81b - — - - - - 82b — + - - + - 83b 7 — + + - _ + - Total Mean 0% 501 171 171 672 02 Language Impaired Group Sla - - - - + - S a - + — + - - 82a - - - - + - 2 Normal Group Slb - + - - + - 82b - + - + + - 83b - - - - + - Total Mean 01 501 02 331 671 0% Note. + - Target Locative Elicited - - Targeted Locative Not Elicited APPENDIX M RANK ORDER FOR MEAN NUMBER OF SUBJESCTS PRODUCING THE TARGETED LOCATIVE WORDS FOR ALL FIVE SUBCATEGORIES BETWEEN THE STRUCTURED TASK AND LANGUAGE SAMPLING PROCEDURES Table M.1 Static Positional Locative Subcategory Percentage Niggtr Group 32::cgzgig :2::I::: Ranks 0 1 LS 1 33 2.5 LS 2.5 33 2.5 S 2.5 50 5 LS 5 50 5 LS 5 50 5 S 5 67 8 S 8 67 8 S 8 67 8 LS 8 83 10.5 S 10.5 83 10.5 S 10.5 100 12 LS 12 T1-44.5 T2-33.5 nl-6 n1-6 Note. The percentages represent the number of subjects who produced a targeted locative word divided by the total number of subjects (6). LS - Language Sampling S - Structured Task 132 Table M.2 Dynamic Direction Locative Subcategory Percentage Rank Group Structured :2:§I:fi; Number Task Ranks Ranks 0 2 LS 2 0 2 SS 2 O 2 SS 2 33 4.5 LS 4.5 33 4.5 LS 4.5 50 6 SS 6 67 7.5 SS 7.5 67 7.5 SS 7.5 83 9.5 SS 9.5 83 9.5 LS ’ 9.5 100 11.5 LS 11.5 100 11.5 LS 11.5 T1-34.5 T2-43.5 n1- 6 n2- 6 Note. The percentages represent the number of subjects who produced a targeted locative word divided by the total number' of subjects (6). LS - Language Sampling S - Structured Task 133 Table M.3 Dynamic Origin Locative Subcategory Percentage Rank Group Structured gzggligg Number Task Ranks Ranks 0 2 LS 2 0 2 LS 2 O 2 LS 2 17 5.5 LS 5.5 17 5.5 S 5.5 17 5.5 S 5.5 17 5.5 S 5.5 50 8 S 8 67 10 S 10 67 10 S 10 67 10 LS 10 100 12 LS 12 T1-44.5 T2-33.5 n1- 6 n2- 6 Note. The percentages represent the number of subjects who produced a targeted locative word divided by the total number of subjects (6). LS - Language Sampling S - Structured Task 134 Table M.4 Dynamic Direction Locative Subcategggy Percentage Nfizbtr Group ::;:°;:;:: giggIzgg Ranks 0 1 SS 1 17 3 Ass 3 l7 3 LS 3 17 3 LS 3 33 5.5 LS 5.5 33 5.5 SS 5.5 50 7.5 SS 7.5 50 7.5 SS 7.5 67 9 LS 9 83 10 SS 10 100 11.5 LS 11.5 100 11.5 LS 11.5 T1-34.50 T2-43.5 n1- 6 n2- 6 Note. The percentages represent the number of subjects who produced a targeted locative word divided by’ the total number’ of subjects (6). LS - Language Sampling S - Structured Task 135 Table M.5 Static Perspective Locative Subcategory Percentage Nfigbtr Group :zztc;:;:: Szggligg Ranks 0 2 5 LS 2 5 0 2 5 L8 2 5 O 2 5 SS 2 5 O 2 5 SS 2 5 17 5.5 SS 5.5 17 5.5 SS 5.5 33 7 L8 7 50 8.5 LS 8.5 50 8.5 SS 8.5 67 10.5 SS 10.5 67 10.5 LS 10.5 100 12 LS 12 w: :22“: 1 2 Note. The percentages represent the number of subjects who produced a targeted locative word divided by the total number of subjects (6). LS - Language Sampling 8 - Structured Task . - '3' . Ah- 411“" _A... a. APPENDIX N RAW DATA USED IN THE RANK SUMS TEST (SENDERS, 1958) BETWEEN TEST-RETEST SCORES 136 Table N.1 Raw Data Used in the Rank Sums Test (Senders, 1958) Between Test- Retest Scores Scores Rank Group Test Retest 12 1 Retest 1 15 2 Test 2 18 3 Retest 3 20 4 Test 4 22 6.5 Retest 6.5 22 6.5 Retest 6.5 22 6.5 Test 6.5 22 6.5 Test 6.5 23 9 Test 9 26 10 Retest 10 T1-28 T2-27 n - 5 n - 5 REFERENCES 137 REFERENCES Bloom, L. (1970). Language develqpment: Form and function in M.I.T. Press. emerging grammars. Cambridge: Bloom, L., & Lahey, M. (1978). Language develppment and language disorders. New York: John Wiley. Bloom, L., Lightbown, P., & Hood, L. (1975). Structure and variation in child language. Monographs of the Society for Research in Child Development, 40 (Serial No. 160). Chicago: University of Chicago Press. Structural relationships in children's In T. Moore (Ed.), Cognitive Bowerman, M. (1973). New utterances: Syntactic or semantic? development and the acquisition of language (pp. 197-213). York: Academic Press. Brown, R. (1973). A first language: The early stages. Cambridge: Harvard University Press. Casby, M. W., & Ruder, F. K. (1983). Symbolic play and early language development in normal and mentally retarded children. Journal of Speech and Hearing Research, 29, 404-411. Relational meaning encoded in the two-word Coggins, T. E. (1979). Journal of utterances of Stage 1 Down's Syndrome children. Speech and Hearinngesearch, 22, 166-178. Formal versus informal assessment: In D. Geffner (Ed.), Topics in Danwitz, M. W. (1981). Salem, MA: Aspen Systems Fragmentation versus holism. language disorders (pp. 95-106). Corporation. Duchan, F. J., & Erickson, J. G. (1976). Normal and retarded children's understanding of semantic relations in different verbal contexts. Journal of Speech and Hearing Research, l2: 767-776. (1979). Incorrect responses to locative Duchan, J., & Siegel, L. Language, Speech, and Hearing Services commands. A case study. in School. 19, 99-103. (1981). Children's use of four semantic Fokes, J., & Konefal, J. Journal of Communication Disorders, lg, cases in two conditions. 497-506. 138 Freedman, P. P., & Carpenter, R. L. (1976). Semantic relations used by normal and language-impaired children at Stage I. Journal of Speech and Hearipg Research, 12, 784-795. Huttenlocher, J., Smiley, P., & Charney, R. .(1983). Emergence of action categories in the child: Evidence from verb meanings. Psycholegical Review, 29, 72-93. Johnston, J. R. (1982). Interpreting the Leiter IQ: Performance profiles of young normal and language-disordered children. Journal of Speech and Hearing Research, 25, 291-296. Johnston, J. R., & Ramstad, V. (1983). Cognitive development in preadolescent language impaired children. British Journal of Disorders of Communication, 18, 49-55. Johnston, J. R., & Savich, P. A. (1984). Anticipatory imagery ability in normal and language-disabled children. Journal of Speech and Hearing Research, 21, 494-501. Johnston, J. R., 5 Slobin, D. (1979). The development of locative expressions in English, Italian, Serbo-Croatian, and Turkish. Journal of Child Language 9, 529-545. Johnston, J. R., & Weismer, S. E. (1983). Mental rotation abilities in language-disordered children. Journal of Speech and Hearing Research, 29, 397-403. Lee, L. (1969). The Northwestern Syntax Screening Test. Evanston, 11: Northwestern University Press. Leech, G. (1970). Towards a semantic description of English. Bloomington: Indiana University Press. Leiter, R. G. (1959). Part I of the Manual for the 1948 revision of the Leiter International Performance Scale. Psychological Service Center Journal, 11, 1-72. Leonard, L. B., Bolders, J. G., & Miller, J. A. (1976). An examination of the semantic relations reflected in the language usage of normal and language-disordered children. Journal of Speech and Hearing Research, 12, 371-392. Lucas, E. V. (1980). Semantic and pragmatic language disorders. Rockville, MD: Aspen Systems Corporation. Mecham, M. J., & Willbrand, M. L. (1979). Language disorders in children. Springfield, IL: Charles C. Thomas. Macrae, A. J. (1976). Movement and location in the acquisition of deictic verbs. Journal of Child Language, 3, 191-204. 139 Newcommer, P. L., & Hammill, D. D. (1977). Test of Language Development. Austin, TX: Empire Press. Olswang, L. B., a Carpenter, R. L. (1982a). The ontogenesis of agent: Cognitive notion. Journal of Speech and Hearing Research, 3;, 297-306. Olswang, L. B., & Carpenter, R. L. (1982b). The ontogenesis of agent: Linguistic expression. Journal of Speech and Hearing Research, 2§,. 306-314. Rice, M. L. (1983). Contemporary accounts of the cognition/language relationships: Implications for speech-language clinicians. Journal of Speech and Hearing Disorders, £8, 347—359. Rice, M. L., & Kemper, S. (1984). Child language and cognition. Baltimore: University Park Press. Schlesinger, I. (1971). Production of utterances and language acquisition. In D. Slobin (Ed.). The ontegenesis of grammar. New York: Academic. Senders, V. L. (1958). Measurement and statistics. New York: Oxford University Press. Stockman, I. J. (1985). Locative distinctions of normal and clinical children. Grant Application, Michigan State University. Submitted to National Institutes of Health. Stockman, I. J. (1986). Locative distinctions of normal and clinical children. Grant Application submitted to Michigan State University. Terrell, B. Y., Schwartz, R. G., Prelock, P. A., Messick, C. K. (1984). Symbolic play in normal and language-impaired children. Journal of Speech and Hearing Research, 21, 424-429. wain, 0., a Yule, W. (1983). Imaginative play in language disordered children. British Journal of Disorders of Communication, 18, 197-205. Woolf, H. B. (Ed.). (1979). Webster's New Collegiate Dictionary. Springfield, MA: G. & C. Merriam Co. Wren, C. T. (1982). Identifying patterns of processing disorder in six-year-old children with syntax problems. British Journal of Disorders of Communication, 11, 83-92. IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII mrl“mlmywuurfllumugnlflwmmm