Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The appropriateness of selected subtests of the Stanford-Binet Intelligence Scale, Fourth Edition for… Perley-McField, Jo-Anne 1990

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1990_A8 P47.pdf [ 4.37MB ]
JSON: 831-1.0054616.json
JSON-LD: 831-1.0054616-ld.json
RDF/XML (Pretty): 831-1.0054616-rdf.xml
RDF/JSON: 831-1.0054616-rdf.json
Turtle: 831-1.0054616-turtle.txt
N-Triples: 831-1.0054616-rdf-ntriples.txt
Original Record: 831-1.0054616-source.json
Full Text

Full Text

THE APPROPRIATENESS  OF SELECTED  SUBTESTS OF T H E  STANFORD-BINET INTELLIGENCE S C A L E : FOURTH EDITION F O R HEARING IMPAIRED CHILDREN by JO-ANNE PERLEY-MCFIELD B.P.E., The University of British Columbia,  1973  Dip. Education of the Hearing Impaired, The University of British Columbia, A THESIS SUBMITTED  IN PARTIAL FULFILMENT OF  THE REQUIREMENTS  FOR THE DEGREE OF  MASTER OF ARTS  in THE FACULTY OF GRADUATE STUDIES (Department of Educational Psychology and Special Education) We accept this thesis as conforming to the required standard  THE UNIVERSITY  OF BRITISH COLUMBIA  June c  1990  Jo-Anne Perley-McField, 1990  1984  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at The University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  (Department of Educational Psychology and Special Education) The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date: June 1990  ABSTRACT This study proposed to evaluate the appropriateness of selected subtests of the Stanford-Binet Intelligence Scale: Fourth Edition (SB:FE) for use with severely to profoundly hearing impaired children. The subjects  used in this  study  were  enrolled in a residential/day school for the deaf whose educational methodology was Total Communication. The subjects were tested on both the SB:FE nonverbal selected subtests and the Performance Scale of the Wechsler Intelligence Scale for Children-Revised (WISC-R PIQ). To data  assess  gathered  appropriateness,  from  the  several  procedures  hearing impaired sample  standardized population of the  SB:FE.  were  employed comparing  with  data reported for the  Correlations were  computed between the  WISC-R and the SB:FE and comparisons of the total composite scores for each measure were made to detect any systematic The impaired  results sample  standardized Scores  of  indicated are  sample the  generally  of the  SB:FE  revealed a difference  that  and  correlations  similar  SB:FE. the  the  differences.  to  the  correlations  The analysis  WISC-R  reported  PIQ to  for  the  reported  hearing for  performed between the detect  systematic  the Area  differences  of one standard deviation between these two instruments,  with the. SB:FE results being lower than the WISC-R PIQ results. It was concluded that the selected subtests of the SB:FE and the WISC-R PIQ  could not be used  interchangeably.  Further  research into  this  area  was  advised before using this measure to estimate general cognitive ability for hearing impaired children whose levels of language development may be delayed. Further research  was  validity  of the  also  encouraged  SB:FE  with  to confirm the academic  measures.  ii  suggestion It  was  of greater suggested  predictive that these  findings indicated that the use of language as a cognitive tool may be important in acquiring certain problem solving skills.  iii  T A B L E OF CONTENTS LIST OF TABLES  v  LIST OF FIGURES  vi  Chapter 1 A. B. C. D.  Background to the Problem Purpose of the Study Hypotheses Definition of Terms  Chapter II. SURVEY OF T H E LITERATURE A. Historical Overview B. Use of the Performance Scale of the Wechsler Intelligence Scale for Children-Revised C. Modifications In Test Instructions  1 1 4 5 6 8 8 13 23  Chapter III. METHODOLOGY A. Procedure B. Sampling C. Instrumentation 1. WISC-R Performance Scale 2. Stanford-Binet Intelligence Scale: Fourth Edition—Selected Subtests 3. Stanford Achievement Test (7th Ed.), Special Edition for Hearing Impaired Students (Gardner et al., 1982) D. Analysis  29 29 30 31 31  Chapter IV. RESULTS  39  Chapter V. SUMMARY AND DISCUSSION A. Summary B. Discussion of Results C. Conclusions D. Limitations of the Study E. Directions for Further Research  33 36 37  49 49 50 65 66 67  REFERENCES  70  APPENDIX A: ITEM ANALYSIS CHARACTERISTICS OF SELECTED SUBTESTS OF THE SB:FE  76  APPENDIX B: COMPARISONS OF SB:FE TEST SCORES USING ALTERNATE SCORING  85  iv  LIST OF TABLES Table 1: WISC-R PIQ Concurrent Validity Correlations  19  Table 2: WISC-R PIQ Mean Score Comparisons With Other Measures  21  Table 3: Correlations Between Area Scores of the Stanford-Binet: Fourth Edition and the WISC-R PIQ  40  Table 4: Subtest Pattern Intercorrelations Between the Hearing Impaired Sample and the Non-exceptional Sample  41  Table 5: Item Analysis Characteristics for the SB:FE  42  Table 6: Comparison of Standard Age Area Scores of the SB:FE and Converted WISC-R PIQ Scores for the Hearing Impaired Table 7: Subtest and Area Intercorrelations Between the SB:FE and the  43  WISC-R PIQ  44  Table 8: Correlations of the WISC-R PIQ and the SAT-HI  46  Table 9: Correlations of the SB:FE and the SAT-HI  47  v  LIST OF FIGURES Figure 1: WISC-R PIQ Subtest Profile  54  Figure 2: SB:FE Subtest Profile  5  vi  9  C H A P T E R I.  A. B A C K G R O U N D TO T H E P R O B L E M It is  possible  to trace  the  history of intelligence  testing of the hearing  impaired and see it as a journey of discovery of the many confounding variables involved in any research. Just as the concept of intelligence and its measurement cannot be simply explained, hearing impairment.  neither  Research  is there  based  a simple condition referred to as  on different  interpretations  of these many  variables has produced conflicting results and perpetuated many misconceptions. Pintner, an early investigator  in this field, produced findings from which  he concluded the deaf were intellectual^' inferior to the hearing (Pintner, Eisenson &  Stanton,  conclusions  1941).  The  fact  about intelligence  inappropriate.  Over  the  that  language  measures  were  used  to  form  for this population did not, at first, appear to be  next  performance tj'pe instruments,  few  decades  however,  the  use  of  non-verbal,  individually administered, came to be regarded as  more appropriate for the assessment of hearing impaired children. A  series  of  studies by  hearing impaired achieved children, there  Myklebust (1960)  global  scores within  appeared to be characteristic  were interpreted  as  indicating the  demonstrated  the  average  that  range  subtest variations.  while  the  for hearing  These variations  hearing impaired population were limited in  their ability for abstract reasoning. Myklebust hypothesized  this was due to the  sensory impairment which caused a different cognitive structuring in the brain. Yet  another line of research, led by Rosenstein  (1961), Furth (1966) and  Vernon (1967) began to indicate that hearing impaired and hearing people were very similar in their cognitive functioning. The number of studies in the last two  1  2 decades investigating the nature of cognitive functioning of the hearing impaired indicates that this area continues to intrigue researchers. The search for the most appropriate instrument to assess intelligence and appropriate  administration procedures  working with the (a)  has  concerned educators  and  psychologists  hearing impaired. The basic principles to emerge  have been  that the instrument must be individually administered and be a non-verbal  performance test; (b) child understands have respected  that the  administration procedures must ensure  the exact nature of the task;  (c)  that the  that  the  instrument must  psychometric properties concerning validity and reliability as well  as a wide research base to validate its use as a diagnostic tool and (d)  that  plurality of instrumentation is necessary for appropriate assessment (Sullivan and Vernon, 1979). The Performance Scale of the Wechsler Intelligence Revised (WISC-R PIQ) (Wechsler,  1974)  instrument  impaired  Abraham,  for  testing  & Stoker,  hearing  1988;  Scale for Children —  has emerged as the leading assessment  Levine, 1974)  children  in  North  America  and in Great Britain  (Kyle,  (Allen, 1980).  Professionals should not, however, become complacent with this situation. As new intelligence measures appear on the market, their use with the hearing impaired population should be explored. The need for plurality of instrumentation has been considered Vernon,  vital to thorough and appropriate assessment practice 1979;  Sattler,  1982),  as  well  as  contributing  to  our  (Sullivan and knowledge  of  cognitive functioning. The  Stanford-Binet Intelligence  Hagen & Sattler,  1986b) has been  Scale: Fourth published. Its  Edition  (SB:FE) (Thorndike,  suitability for use  with the  hearing impaired population needs to be explored. The previous edition of the  3 Stanford-Binet Intelligence Scale (Terman and Merrill, 1973) was considered totally inappropriate for the assessment of intelligence in the hearing impaired population because of its heavy emphasis on language ability (Sullivan and Vernon, 1979). The fourth edition represents a major revision of the scale. The authors decided that  four  quantitative  areas  of  cognitive  ability  should  be  appraised:  verbal  reasoning,  reasoning, abstract/visual reasoning, and short-term memory. It was  also decided that the revised scale would continue to provide an overall score that would represent general reasoning ability. On the surface, the new changes to the Stanford-Binet Intelligence appear  to  make  its  use  with  the  hearing  impaired  population  Scale  potentially  appropriate. The separation of the Verbal Reasoning Area from the Quantitative Reasoning Area, the Abstract/Visual Reasoning Area and the Short-Term Memory Area suggests that these latter areas may not rely heavily on verbal abilities. Man3' subtests of these areas appear to be essentially nonverbal, performance-type tests. As such their use with hearing impaired children may be appropriate. Within  the  Short-Term  the  division of visual memorj' into two  Stanford-Binet,  yield particularly useful  Memory  Area  information. Studies  of  this  fourth  edition  separate  of  the  subtests may  by Watson, Sullivan, Moeller and  Jensen (1982) have reported high correlations between visual memory and English language ability. For the older child especially, the Stanford-Binet Intelligence  Scale: Fourth  Edition provides a wide range of subtests, sampling various areas of nonverbal abilities. A total of eight subtests may be used with the older child, providing more  diagnostic  estimate.  information  and  increasing  the  reliability  of  the  intellectual  4  Yet another attractive feature of this instrument is that it functions as a power  test  responses  rather is  than  regarded  a test as  rewarding fast  especially  responses.  important  impaired children (Sullivan and Vernon,  in  the  The lack  assessment  of timed of hearing  1979). The one timed subtest, Pattern  Analysis, only sets a time limit and does not reward for extra speed. In  addition  to  these  promising  features,  the  new  edition  of  the  Stanford-Binet has other attractions for use with a hearing impaired population. Each measuring instrument provides to some degree a unique perspective of the construct  that  it  purports  to  measure.  The  perspective  offered  by  the  Stanford-Binet Intelligence Scale: Fourth Edition may well provide new information to  the  field  of  cognitive  functioning  and  reveal  exciting  new  areas  for  investigation.  B. PURPOSE OF THE STUDY The  purpose of this study was to evaluate  selected subtests of the new  Stanford-Binet Intelligence Scale: Fourth Edition (SB:FE) on its appropriateness for use  with  hearing  impaired  children  whose  educational  methodology  is  Total  Communication (TC). To assess appropriateness, several procedures were employed. The  patterns of intercorrelations between the Performance Scale of the Wechsler  Intelligence for  the  Scale for Children — Revised and the selected subtests of the SB:FE sample  correlations evidence  of  reported  hearing for  the  impaired  children  non-exceptional  of the construct validity of the new sought  by  comparing total  were  compared  sample  of  SB:FE  for the  population  was  scores  measures.  Evidence of appropriate internal consistencj  the  from these 7  with  SB:FE.  similar Further  hearing impaired two  intelligence  was sought by conducting  5 exploratory item analysis on the SB:FE results. The predictive validity was also explored  03' correlating  this  available from results of the  instrument  with  academic  achievement  Stanford Achievement Tests  measures  — Hearing Impaired  (SAT-HI) (Gardner, Rudman, Karlsen & Merwin, 1982).  C. HYPOTHESES The hypotheses to be tested in this study: 1.  For  the  hearing  correlations  between  Abstract/Visual,  impaired the  population  Performance  Quantitative,  in  this  Scale  of  study, the  the  pattern  WISC-R  and  of the  and Short-Term Memory Area scores of the  Stanford-Binet: Fourth Edition will not be significantly  different  from that  reported for the non-exceptional sample. 2.  For the hearing impaired population in this study, the pattern of subtest intercorrelations  among the  SB:FE  will not be significantly  different from  that reported for the non-exceptional sample. 3.  For the hearing impaired population in this study, the pattern of subtest intercorrelations  among the WISC-R PIQ will not be significantly  different  from that reported for the standardization sample. Exploratory research questions to be addressed: 1.  Exploratory item analysis will be conducted to examine item  characteristics  of the various subtests of the SB:FE to determine significant trends for this sample of hearing impaired children. 2.  For the  hearing impaired population in this study,  systematically 3.  What  is  the  will either  instrument  yield greater composite scores than the other? pattern  of  subtest  intercorrelations  between  these  two  6 instruments? For instance,  are certain subtests characteristically  associated  on both instruments, e.g., Block Design and Pattern Analysis? 4.  What are the correlations produced between the WISC-R PIQ, the and  SB:FE  the SAT-HI academic measures?  D. DEFINITION OF T E R M S Throughout the study the following terms will be used as defined below: Deaf A  term indicating a hearing disability which precludes  linguistic information through audition, with or without  successful  processing of  a hearing aid. A deaf  individual would have a hearing loss of 90dB or greater in the better unaided ear across the speech frequencies (500, Hearing  1000, and 2000 Hz).  Impairment  A term used to describe a person who has any degree of hearing loss. It is a generic term which includes both deaf and hard of hearing persons. A m e r i c a n S i g n Language ( A S L ) American  Sign  Language  (ASL) is  the  language  of  the  signing  adult  deaf  communit3' in North America (Baker and Cokely, 1980). Pidgin Sign A The  English  sign communication that incorporates grammar from both ASL and English. syntax  can vary from being more like ASL to being more like English  (Woodward, 1973).  7  Nonverbal Performance Tests These are tests that do not require verbal responses. by  manipulation  Tasks may be performed  of materials or by indicating choices using a pointing response.  These tests may require verbal directions.  Total Communication A  philosophy  modes  of communication  accepted 1976  requiring the incorporation of appropriate with  and among  by the Conference of Executives  (Gannon, 1981, p. 369).  hearing  aural, manual, and oral  impaired  of American Schools  persons.  Definition  for the Deaf in  CHAPTER II. SURVEY OF THE LITERATURE This chapter has been organized to present three major topics. The first provides  a  brief  historical  overview  of  intelligence  testing  with  the  hearing  impaired. The second topic addresses the issue of which testing instruments are in  current use and reviews the findings of previous research with the WISC-R  PIQ.  The third topic reviews the modifications used in test instructions for the  hearing impaired and presents  the rationale for the administration modifications  used in this study.  A. HISTORICAL  OVERVIEW  There has been considerable research interest in intelligence the  hearing impaired  during  this  century.  In general,  the  testing with  thrust of research  inquiry has been toward establishing the relative intelligence of deaf and hearing children. Researchers  have  evolved from the  position of regarding the deaf as  intellectually inferior, to the present day position which considers that while some differences in test performance may be observed, in general the deaf as a group score within the average range of intelligence. What is apparent in reviewing the research in this area is that much of the inconsistencies intervening  into  in the findings have been a result of the different variables the  testing  situation  and the  different  measuring instruments  used. As experimental control has increased, the observed differences between the deaf and hearing have decreased. The  earl}'  research  conclusions  reached  by  Pintner  and  his  colleagues  illustrate the significance of type of measuring instrument used. Pintner, Eisenson and  Stanton  (1941)  concluded  that  deaf  8  people  were  intellectually  inferior to  9 hearing  people  and  showed  definite  deficits  in  various  aspects  of  cognitive  functioning. What is important to note is that most of Pintner's tests required the use of verbal English language use  of a verbal measure  both in administration and in response. The  to assess the reasoning abilities of a deaf child are  considered inappropriate today (Sattler, 1982; Sullivan and Vernon, 1979). Vernon's  (1968)  summary  of  the  research  on  the  intelligence  of  the  hearing impaired also contributed importantly to the identification of intervening variables in the research process. After reviewing fifty years of research in this area, Vernon concluded that when there are no complicating multiple handicaps, the hearing impaired function at approximately the same IQ level on performance intelligence  tests as do the hearing. Vernon had examined closely the relationship  of etiology of deafness to intelligence.  A disproportionate^ higher prevalence of  low IQs for the hearing impaired population had been noticed. Vernon suggested that many of the other  etiologies of profound hearing loss were also responsible for  neurological  (1982)  has  children  impairment which would result in lower intelligence.  suggested  maj' be  that  the  distribution of  intelligence  Sullivan  in hearing impaired  bimodal. The multihandicapped group would  tend  to form a  lower group and non-multihandicapped would form a higher group. However,  despite  advances  in  experimental  assessment procedures, not all the observed differences children disappeared.  There still were enough  intrigue researchers. There is disagreement  control  and  improved  between deaf and hearing  inconsistencies in the  findings  as to how these observed  to  differences  should be interpreted. If these observed differences  are in fact real, it would be  logical  differences  to conclude  Myklebust  (1966)  that  the  represented  deaf this  have view  cognitive and  proposed  a  from the theory  of  hearing. sensory  10 deprivation to explain these differences.  Myklebust found that while the hearing  impaired performed within the average  range of general cognitive  ability, there  were certain subtest variations that indicated a qualitative difference  in mental  functioning. He concluded that the deaf fell below average mainly on tests which required  a  type  of  abstraction  and  reasoning  process.  To account  for  these  observed variations in the mental abilities of deaf children, Myklebust proposed the organismic shift hypothesis. present  from early life,  adequate maintain  the  This stated that when a sensory deprivation is  organism must modify  its  means  for maintaining  environmental contact. A shift in perceptual organization was made to psychological  equilibrium.  This  resulted  in  a  different  cognitive  structuring from normally hearing children. Levine (1981) suggested that the cognitive differences observed in the deaf were due to different experiences.  This experiential defecit theory suggested that  the environment has a powerful influence in the fashioning of human behavior. The absence of sound from the environment of the hearing impaired would create psychological voids. Researchers  such as Rosenstein  (1961), Furth  (1966) and Vernon (1967)  have supported the position that deaf and hearing impaired people are cognitively similar in all important abilities. Observed differences in test results may be due to a remaining lack of experimental control over verbal and other factors  that  influence performance on tasks that are believed to be nonverbal. Research has been done that shows the method of administration of tests has significant effects on test results.  This issue will be  discussed  in more detail  in a  subsequent  section. The validity of using Performance type tests as an estimate of intelligence  11 has been the subject of further research. When Pintner produced his research on intelligence (using language measures) the current theoretical position favoured the belief that language was primary and thinking took place in language. Therefore the use time.  of language  As  the  to assess intelligence  research  in  the  field  of  was  logical and acceptable at that  intelligence  accumulated,  different  interpretations of intelligence began to evolve. A position known as the cognitive dominant hypothesis  began to  receive  research support. This  position proposed  that basic perceptual and cognitive development can proceed independently from language.  The acceptance  language  that  tended to support the  all cognitive use  processes  do not necessarily need  of performance scales to  assess cognitive  ability. Very intelligence reported.  few  research  measures  with  Research into the  studies the  investigating  hearing impaired  the  predictive  school  validity  population have  relationship of nonverbal intelligence  of been  measures and  academic achievement in the hearing population suggests this relationship is not particularly  strong.  Zimmerman  and Woo-Sam  (1972),  in  a  series  of  studies  relating the WISC (Wechsler, 1949) to various achievement measures, reported a median  correlation of  .33  (1957) began to investigate  between  Performance IQ and achievement.  this issue in the  Murphy  hearing impaired population and  found WISC Performance Scale results to correlate 0.64 with teacher's ratings of current  academic  attainment.  supported, however, in studies  This  rather  strong  coefficient  nonverbal  1949). Correlations of .54  intelligence  measures  not  been  using standardized test instruments. Brill (1962)  conducted a study using the Wechsler Adult Intelligence WISC (Wechsler,  has  and  two  and .55  Scale (WAIS) and the  were  standardized  found between  achievement  the  measures.  12 Hirshoren,  Hurley  and  Kavale  (1979)  correlated  the  WISC-R  PIQ with  the  Stanford Achievement Test. The most robust single correlation of .35 was found between (1980)  the  Average  correlated  Grade  the  Score  WISC  and the  and  WISC-R  WISC-R  PIQ. Brooks and Riggs  Performance  IQs  with  reading  achievement using the SAT-HI as well as class observations and the results were found  to  be  explored  nonsignificant.  the  relationship  In  1982,  between  Watson,  nonverbal  Sullivan,  intelligence  Moeller and  and  English  Jensen language  abilitj' in deaf children. Using both the WISC-R PIQ and the Hiskey-Nebraska Test  of  Learning  Abilities  (H-NTLA)  (Hiskey,  correlations of .45 between these measures further,  that  1966),  subtests requiring visual memory contributed most  study  correlation  by  of  achievement.  found  and language measures.  the multiple regression analysis which yielded an average further  they  Watson,  .42  between  Goldgar, the  Kroese  H-NTLA  and Lotz and  They found  significantly  to  correlation of .68. A  (1986) found  various  average  measures  of  a median academic  Correlations between the WISC-R PIQ and these academic measures  were lower with a median of .24.  The authors suggested that the strength of  the H-NTLA achievement correlations may be due to the heavier weighting this test gives to visual memory tasks. may  be  particularly important  in  They suggested that visual memory skills the  hearing  impaired child's  acquisition of  language and knowledge. These rather low reported correlations between performance measures and academic measures do not appear to provide much support for the use of these intelligence measures as an indicator of academic aptitude. However, it should be considered  that  these  measures  do  provide  some  information.  If  used  in  conjunction with other assessment instruments, useful information can be provided  13 to  assist  in  educational  planning.  It  is  most  important  however,  for  those  involved with such decision making, to be well aware of the limitations of these test instruments.  B.  USE  OF  THE  INTELLIGENCE The measure  PERFORMANCE  SCALE  OF  THE  WECHSLER  SCALE F O R CHILDREN-REVISED  Performance Scale of the WISC-R has been used as the comparison  in  this  study.  This  was  done  to  parallel studies  conducted on the  construct validity of the SB:FE. The data for these studies are available in the Technical Manual of the SB:FE (Thorndike, Hagen & Sattler, 1986a). The studies were carried out to determine the correlations between the individual  intelligence  non-exceptional intelligence  samples  tests.  These  and  special  studies groups  were of  SB:FE  conducted  children.  Of  and several using  the  both  individual  tests used in these construct validity studies, the WISC-R PIQ was  deemed the most appropriate for use in this study as the comparison measure. Evidence will be presented to show that it is a nonverbal performance type of instrument, considered most suitable for testing the hearing impaired. It has been found  to  be  the  most  frequently  used  intelligence  measure  with  the hearing  impaired. Further, an extensive body of research has accumulated on the use of this instrument with hearing impaired populations. It results  has  been  between deaf  increased.  The  noted  that  most  observed  and hearing have  identification  of  the  differences  decreased  manj'  as  in  intelligence  tests  experimental control has  confounding  variables  involved  in  research work with the deaf has also led to improved assessment procedures for educators, pyschologists, and other professionals working with the deaf.  14 One of the first principles to be formulated defined the type of test used in  assessing intelligence  with the  deaf presented  in deaf  children. Lane  an earty  (1938),  paper on the  a psychologist working  measurement  of mental and  educational ability of the deaf children which indicates the stage of psychological testing practices of that period. Lane acknowledged that the issue of whether a performance test could replace a linguistic test was still controversial, and that the  problem  depended  on  of the  recommendations  comparing  the  definition  of  for  valid  intelligence intelligence  psychological  of  hearing  accepted  and  hearing impaired  the  investigator.  by  measurement  included  the  Her  use  of  nonverbal performance type instruments. Lane (1938) observed that . . . any test requiring linguistic responses or the understanding of verbal instructions becomes a test of the educational achievement and lipreading ability of the deaf as well as a measure of native intelligence. Selection must therefore be made from those performance tests in which the directions are in pantomime and the responses are manipulative (p. 169). She further recommended that individual tests were to be preferred over group tests because of the influence of such external factors as attention to the examiner, understanding of directions, and time factors. The use of individually administered, nonverbal performance tests continues to be recommended  to the  present  day.  Commonly used  intelligence  measures  standardized on hearing impaired school children which meet these requirements include: 1.  the Hiskey-Nebraska Test of Learning Aptitude (H-NTLA) (Hiskey, 1966);  2.  the Non-Verbal Intelligence and Snidjers-Oomen, 1970);  3.  Tests  for Deaf and Hearing Subjects ,  the Ontario School Ability Examination (Amoss, 1949); and  (Snidjers  15 4.  the Wechsler Intelligence Scale for Children-Revised, Performance Scale, Deaf Norms (Anderson and Sisco, 1977).  Commonly used intelligence  measures not standardized on hearing impaired school  children which also meet the requirements are: 1.  the Arthur Adaptation of the Leiter International Performance Scale (Arthur, 1955);  2.  the Columbia Mental Maturity Scale (Burgemeister, Blum & Lorge, 1972);  3.  the Goodenough-Harris Draw-a-Man (Goodenough & Harris, 1963);  4.  the Leiter International Performance Scale (Leiter, 1948);  5.  the Standard Progressive Matrices (Raven, 1960); and  6.  the  Wechsler  Intelligence  Scale  for  Children-Revised,  Performance  Scale,  (Wechsler, 1974). These  lists  are  not  intended  to be  exhaustive  but  mention  those  tests most  commonly in use with the hearing impaired school aged child (Allen et al., 1988; Bragman,  1982;  Kyle,  1980;  Levine, 1971;  Levine, 1974;  Sullivan and Vernon,  1979; Vernon, 1976; Vernon and Brown, 1964). Of these tests, the Performance Scale of the Wechsler Intelligence Scale for  Children  commonly  —  used  Revised in  (Wechsler  psychological  1974)  assessment  has of  been  reported  school  children. The deaf norms are not reported to be widely were produced from a study national sample of 1,228  conducted  by  age  hearing  in use.  Anderson and  as  Sisco  the  most  impaired  These norms (1977) on a  deaf children. These children were from 18 residential  schools and 4 day-schools for deaf children located throughout the United States. A major concern regarding the use of these norms appears to be the lack of standardized administration procedures.  16 In hearing  a  United  impaired,  States  the  national  WISC-R  survey  of  Performance  psychologists  Scale  was  working  reported  as  with the  the most  frequently used intelligence measure by an overwhelming majority of psychologists (Levine,  1974).  Another survey  by  Spragins  and  Gibbins (1982),  though  not  national in scope, indicates that the popularity of this instrument has continued. In their survey  of school psychologists working with hearing impaired children,  they found that the WISC-R was used by 91% of the respondents. In New York and the New England states, a survey indicated that the WISC-R Performance Scale was still ranked the most popular test in use for hearing impaired children (McQuaid position  and was  survey  of  Alovisetti, further  1981).  demonstrated  psychoeducational  That by  the the  assessment  WISC-R results  of  the  PIQ still  of the  hearing  Allen  maintains  this  et al. (1988)  impaired in  both  the  United States and Canada. In the United Kingdom too, the Performance Scale of the WISC-R has been cited as being the most widely used for hearing impaired children (Kyle, 1980). In view of the small number of intelligence impaired  children,  non-deaf  standardized  tests  tests standardized on hearing  have  maintained  an  important  position in psychological test practices. Leading authorities in the field of psj'choeducational assessment have maintained various positions on the issue of whether a  test developed  and standardized on a hearing population can be  considered  suitable for use with a hearing impaired population. Gerweck maintain  a firm  and  Ysseldyke  position  (1975)  against  the  and use  later  Salvia  and Ysseldyke  of tests standardized  on  (1985)  a hearing  population. Thej' have stated that their use with a hearing impaired population violates  the  basic  assumptions  underlying  psychological  assessment.  Anastasi  17 (1982) also criticizes states  that  one  modifications  cannot  assume  used  in standardized testing  that  reliability,  validitj'  procedures. She  and  norms  remain  unchanged in these circumstances. Both Anastasi (1982) and Salvia and Ysseldyke (1985) recommend only the Hiskey-Nebraska Test of Learning Aptitude (Hiskey, 1966) for use with hearing impaired school children. Sattler  (1982)  specifically  recommends  the  Performance  Scale  of  the  WISC-R as a reliable and valid instrument for assessing the intelligence of deaf children and asserts that the 1974 published norms are valid for the assessment of  deaf children. He further claims  measured  by the  WISC-R  there  is  no evidence  Performance Scale,  differs  that  intelligence,  qualitatively  as  in deaf and  hearing groups. A study by Hirshoren, Kavale, Hurley and Hunt (1977) reported that  with  respect  to  both  individual  subjects  and  total  test  reliability,  the  Performance Scale of the WISC-R possessed a satisfactory level of reliability for testing  the  compared  mental  for  each  abilities  of  subtest  and  a  deaf  for  population.  the  composite  Reliability coefficients Performance  Scale  were of  the  WISC-R. For the Object Assemblj', Block Design and Picture Completion subjects, the reliability values did not differ significantly from the standardization sample. One  subtest,  Picture  Arrangement,  was  found  to  have  a  higher reliabilitj'  coefficient. It was suggested that the greater variability found in the deaf sample on this subtest may have contributed to this finding. The composite reliability of the  WISC-R  composite  was  total  based  (r = .90)  on  the  reported  reliabilities  of  four  subtests.  The  is comparable to the total Performance Scale reliability  reported by Wechsler (1974). Braden (1985) reports that the WISC-R Performance Scale  has  the  same  factor  structure  in  deaf  and  hearing  groups.  Braden  concluded that the internal validity, construct validity, and criterion validity of the  18 WISC-R  Performance  Scale  are  essentially  the  same for deaf  children as  for  hearing children. Vernon has written extensively in the field of psychological assessment of the hearing impaired. Vernon includes the Performance Scale of the WISC-R in his published lists of recommended tests and reviewed  as an excellent test for  use with hearing impaired children (Sullivan & Vernon, 1979; Vernon and Brown, 1964;  Vernon,  1976).  The  issues  of  the  lack  of  standardized  administration  procedures is not ignored, however. Although test administration modifications are advocated, their use is noted to affect test reliability by introducing sources of inconsistenc}' and so underscores the importance of administering more than one performance intelligence measure. The WISC-R  research  Performance  comparison measure  information Scale  also  which  has  contributes  in this study.  accumulated to  its  on  the  desirability  Much of this research  for  use  of  use  the  as  a  provides information  that supports its use with hearing impaired children. There have been a limited number of studies exploring the correlations between the WISC-R PIQ and other nonverbal  intelligence  measures.  Table  1 displays  results  from  some  of these  studies. Other studies report mean score same group comparisons of the WISC-R PIQ  with  other  nonverbal  intelligence  measures.  The results  of these studies,  presented in Table 2, do not yield consistent findings as to whether the WISC-R PIQ generally scores higher or lower than other measures. In summary, the accumulated evidence regarding the use of the WISC-R Performance Scale in the assessment of heanng impaired children is regarded as substantial enough to support its value as a comparison measure in this study.  19 Table 1 WISC-R PIQ Concurrent Validity Correlations Study Ritter (1976)  Subjects  Criterion  31 children x age = 6.9 yr. HL mild to moderate  AA-LIPS  .78  CPM  .50  H-NTLA  .89  CPM CPM CPM CPM  .56 .56 .65 .83  Hirshoren, Hurley & Kavale (1979) 59 children x age =10.6 yr. prelingually deaf Evans (1980)  125 5-12 yr. prelingually deaf ages 5-6 ages 7-8 ages 9-10 ages 11-12  Ullissi & Gibbons (1984)  40 children x age 8 yr. prelingually deaf  LIPS  .82  James (1984)  34 age HL 50 age HL  children 6-11 x = 85dB children 11-16.5 yr. x = 80dB  CPM  .87  SPM  .78  Watson & Goldgar (1985)  71 children x age =13.2 yr. HL profound 44 HL severe 22 HL mod. severe 5  H-NTLA  .85  Phelps & Ensor (1986)  49 children x age= 11.51 yr. HL-prelingually deaf (moderate to profound)  H-NTLA  .91 WISC-R deaf norms  Porter & Kirby (1986)  49 children ages 7-12 yr. x age=10.3 yr. HL x = 102dB • pantomine gesture group • ASL group  KABC non-verbal scale .68 .64  20 Table 1 continued AALIPS: Arthur Adaptation of the Leiter International Performance Scale (Arthur, 1955) CPM: Raven's Colored Progressive Matrices (Raven, 1965) H-NTLA: Hiskey-Nebraska Test of Learning Aptitude (Hiskey, 1966) LIPS: Leiter International Performance Scale (Leiter, 1948) SPM: Standard Progressive Matrices (Raven, 1960) KABC: Kaufman Assessment Battery for Children (Kaufman & Kaufman,  1983)  21 Table 2 WISC-R PIQ Mean Score Comparisons With Other Measures Study Moores, Weiss & Goodwin (1976)  Quigley (1969)  Ritter (1976)  Hirshoren, Hurley & Hunt (1977)  Anderson & Sisco (1977)  Subjects 60 preschool children 3yr interval LIPS first 32 children x age = 3.8 yr. 4yr interval LIPS first 31 children x age = 6.9 yr. HL = mild to moderate 59 children x age=10.6 yr. HL = prelingually deaf  1228 national sample age 6-0 to 16-11 HL=70dB +  WISC-R Criterion x (SD) x (SD)  110.10  LIPS 116.60 range 82-157  102.00  LIPS 114.00  82.20 (16.00)  AALIPS 88.80 (18.10)  88.07 (17.84) range 52-129  H-NTLA 89.86 (16.53) range 57-131  95.70 (17.55) n=1202  LIPS 97.44 (16.90) n = 320 H-NTLA 96.78 (16.64) n = 219  James (1984)  Watson & Goldgar (1985)  34 age HL 50 age HL  children 6-11 x = 85dB children 11-16.5 yr. x = 80dB  71 children x age= 13.20 yr. HL 44 prfound HL 22 severe HL 5 mod/severe  92.36  CPM 97.54  101.41  SPM 99.06  99.13 (17.42)  H-NTLA 100.21 (22.36)  22 Table 2 continued Study Ullissi & Gibbons (1984)  Phelps & Ensor (1986)  Porter & Kirby (1986)  Watson et al. (1986)  Brooks & Riggs (1980)  Subjects 40 children x age = 8.0 yr. HL=prelingually deaf 70dB + 49 children x age =11.51 yr. HL=prelingually deaf (moderate to profound)  49 children ages 7-12 yr. HL = prelingually deaf HL x = 102dB '  53 children x age= 13.90 yr. HL = mod-severe to profound 40 children x age =10.90 yr. HL=40-80 + dB  WISC-R Criterion x (SD) x '(SD) 96.30 (13.00)  91.90 (16.93) 58-130 (deaf norms)  LIPS 97.40 (23.30) H-NTLA 90.41 (20.46) 57-139  KABC 107.92 98.8 n = 25 (11.80) pantomine/ gesture 108.38 96.8 n = 24 (10.90) (ASL)  105.00 (11.82)  H-NTLA 105.00 (15.80)  90.35 (19.81)  WISC 96.08 (17.57)  AALIPS: Arthur Adaptation of the Leiter International Performance Scale (Arthur, 1955) CPM: Raven's Colored Progressive Matrices (Raven, 1965) H-NTLA: Hiskey-Nebraska Test of Learning Aptitude (Hiskey, 1966) LIPS: Leiter International Performance Scale (Leiter, 1948) SPM: Standard Progressive Matrices (Raven, 1960) KABC: Kaufman Assessment Battery for Children (Kaufman & Kaufman, 1983)  23  C. MODIFICATIONS IN TEST INSTRUCTIONS The method of test administration has been determined to be a crucial variable  in  Considerable  the  psycho-educational  work  has  been  done  assessment  in  this  area  of for  the it  hearing  has  been  impaired. noted  that  misinterpretation can occur both with regard to the task demand given by the examiner  and  the  task  response  given  by  the  child.  There has  been  some  research providing empirical evidence that different procedures have a significant effect  on  test  results.  Graham  and  Shapiro  (1953)  compared  the  WISC  Performance Scale results using two different administration methods. Two groups of  children with  normal hearing were  individually matched  with  members  of  group I, the hearing impaired children, using the Goodenough Draw-A-Man Test. The WISC Performance Scale was  administered individual^' to the members of  Groups I and II through pantomime instructions. The members of Group III were given  the  WISC  in  accordance  with  the  standardized  instructions.  Reported  Performance IQs were: Group I  Group II  Group III  96.1  95.5  101.5  The difference between groups II and III was found to be significant at the 5% level.  The  authors  concluded  equaled in intelligence  that,  because  the  three  groups  were originally  as measured by the Goodenough, it would seem that the  difference in WISC performance resulted from the method of administration. Sullivan  (1982)  has  also  demonstrated  that  different  administration  modifications result in significantly different test scores for different categories of deaf  children.  methodology,  From  a  school  using  Total  Communication  (TC) instructional  45 children with severe to profound prelingual hearing losses were  24 sorted  into  three  handicapped  etiological  group,  and  a  groups. group  These of  were:  a  unknown  genetic  etiology.  group,  Three  a multiply  administration  conditions were employed with the Performance Scale of the WISC-R: pantomime, visual  aids,  and  total  etiological groups was General  results  communication. randomly assigned  indicated that  significantly higher scores. children  in  residential  communication  Each  the  of  the  15 subjects in  the  three  to one of three administration groups.  total  communication administrations produced  The implication drawn was that for hearing impaired  and  methodology,  public the  schools  use  of  who  TC  are  in  instructed  the  in  a  administration  of  total the  standardized subtests of the WISC-R Performance Scale would yield higher scores. Bragman (1982) has presented a review of modified test instructions used with hearing impaired populations. The three methods customarily used have been pantomime; demonstration; and some form of language speech,  lipreading,  researchers Murphy  have  written  used  words,  or  manual  these procedures singly  (1957) used all three methods  communication involving  communication.  the  past  or in combination. For instance  to some degree.  should be used throughout the test so that  In  He stated that speech  students who could lipread would  take advantage of it. He also employed some pointing and gesturing as well as additional demonstrations. Neuhaus (1967) believed only pantomime should be used regardless  of  how  well  the  child was  able  to  speechread.  He felt  that  any  advantage a good speechreader had over a poor one should be eliminated. Many researchers in the past have been rather vague in referring to procedures of test administration.  Sattler  (1988;  1982)  has  published  two  modified  instruction  procedures: the Pantomime instructions from Sullivan (1982) and a combination of Murphy  (1957), Neuhaus  (1967), and Reed (1970).  Sattler  (1988) has further  25 recommended the use of TC to administer tests to children who use TC as their educational  modality.  Performance  Scale  supplemental  items  (1979)  Ray  published  administration for  an  procedures.  demonstrations  as  adaptation These  well  as  of  involve  Alternate  the the  WISC-R use  of  Instructions  for  signing. Each advantages  of  the  three  categories  and disadvantages.  of  The use  modified  test  of pantomime  instructions  has  certain  does not always  clearly  convey what is needed, nor is it necessarily free of language biases (Goetzinger and Houchins, 1969; Reed, 1970; Tweney and Hoemann, 1976). It has also been found to depress scores. (Graham and Shapiro, 1953; Sullivan, 1982). Pantomime administration  would  have  considerable  advantages  if  the  test  standardized in that manner. Otherwise, the effects of pantomime  had  been  may be as  variable as other test modifications. The  use  administration  of  may  demonstration have  the  on  serious  extra effect  practice of  items  rehearsal.  to  modify  The effects  test  on  the  validity of the test are not known. However, since the value of each test item depends to a great extent on the novelty of the task, it can be assumed that extra demonstration is very likely to alter the nature of the task. This type of test modification is seen as a more drastic alteration than pantomined instructions and less likely to produce valid results. The use following  of language  communication has  standardized procedures.  the  The disadvantages  advantage are  the  of more variable  closely  language  levels of the hearing impaired children. Unlike hearing children, hearing impaired children cannot be assumed to have a common language base. There appears to be little consistency  in the modifications used in actual  26 assessment situations.  Levine's (1974) survey of psychologists  working with the  hearing impaired revealed that 52% used mainly speech and 31% used mainly signs  in  reported  test great  was  used by  with  speech;  administration. inconsistency 77.2% 1.3%  The  Sisco  (1977)  in mode of test instructions.  of the used  Anderson and  sample;  only  2.2%  gestures;  used  7.3%  Oral;  used  norming study  Total Communication  4.4%  used fingerspelling  pantomime;  6.6%  employed  other, unidentified procedures. In another survey McQuaid and Alovisetti (1981) reported that 71% of the respondents indicated they used Total Communication for test administration. It is important  to  administrations communication  note here were  that  the  standardized.  encompassing  many  use  of TC does not Total  necessarily  Communication  different  communication  is  a  modes.  mean  the  style  of  Audition,  speechreading, signing and speaking, gestures, and pantomime are all incorporated. The exact proportions of each modality used in any communication event are not stipulated.  Gesturing  and  pantomime  would  logically  reflect  the  personal  characteristics of the communicator. Some of the signs used may well be unique to the geographical region or even the specific school attended by the child. The results  of a recent survey taken in the United States and Canada by Allen,  Abraham and Stoker (1988) indicated that 48.8% of the respondents  used both  speech and sign simultaneously in assessment situations. These research results appear to indicate a growing trend in the field of assessment of the  hearing impaired: to  use  a Total  communication. This may reflect the growing acceptance education  of  the  hearing  impaired. As  mere  children  Communication mode of and use of TC in the are  taught  using TC  methodology, the more it becomes the most appropriate mode of communication in  27 which to assess those children. In the United States, P.L.94-142, has stimulated additional incentives for the use of sign language. Section 121a 532 states: tests and other evaluation materials are provided and administered in the child's native language or other mode of communication, unless it is clearly not feasible to do so. (Federal Register, Aug 23, 1977 p. 42494).  It  is  this  consideration  of  the  individual  abilities  of  the  child  which  appears to be the most important determinant in the choice of administration mode. For the orally educated child, spoken communication should be employed. For  children  administration  who is  use  TC in  considered  school  appropriate  or  who  (Sullivan  know and  sign  language,  Vernon,  1979;  a TC Sattler,  1988). The choice of administration mode becomes a judgement on the part of the examiner. What is most imperative is to choose the method most likely to decrease the incidence of misinterpretation between examiner and child. For this study, a form of Total Communication using Pidgin Sign English was chosen to be the administrative mode used for both tests. Stokoe (1970) found that there were  different  varieties  of  sign  language  within  the  deaf  community.  Stokoe  described the language varieties in terms of a continuum ranging from American Sign Language (ASL) to English. used  the  Woodward (1973) developed this concept and  term Pidgin Sign English  (PSE) to describe  the  in-between  varieties  along the ASL-English continuum. He felt PSE was used as a bridge between the  two  languages.  communication that  As  such,  the  term  PSE  is  used  to  describe  signed  consists of some grammar from ASL and some grammar  from English. There is some empirical evidence to suggest that PSE is used by deaf students in the classroom (Woodward 1973;  Livingston, 1983). PSE could  28 well be the communication form used by many parents and educators. Not all parents  and  educators  are  equally  proficient  in  the  use  of  ASL or Signed  English. It is quite likely that there is a continuum of signing skills occurring with most communication being in the PSE form. The Signing Consultant employed  at the  school from which the  sample  population of this study was drawn felt that a PSE format would be the most familiar  form  procedure  was  of  communication  developed  that  procedures of each intelligence  for closely  these  children. Therefore,  followed  the  a careful PSE  standardized  administration  measure. For each of these tests, the procedures  followed were identical for each child. In the judgement of the Sign Consultant, the PSE test administration procedures of the WISC-R Performance Scale and the SB:FE  were  continuum.  consistent  in  their  location  along  the  ASL  — English  sign  C H A P T E R III. METHODOLOGY This study proposed to evaluate the appropriateness of selected subtests of the new  Stanford-Binet Intelligence  impaired  children whose  evaluation  Scale: Fourth  educational  included correlations  Edition  methodology  with  the  is  WISC-R  for use  Total  with hearing  Communication. The  Performance  Scale  and  the  academic measures of the SAT-HI. Chapter III outlines the procedures of this study, describes  the sampling  and instrumentation used, and briefly describes the methods for data analysis.  A. P R O C E D U R E For  each  subject  who  had parental permission  to  be  stud}', audiological information, medical information, and results  included in  this  of the previous  year's SAT-HI academic measures were collected from the school files. Each subject was and  administered both the  selected subtests of the  SB:FE  the WISC-R Performance Scale. Subjects were randomly assigned to one of  two groups, and the order of test administration was counterbalanced for the two groups. Because  of the effects of fatigue,  each  both  child  for  administrations  test  administrations.  for any child was  the time of day was The  between two  time  interval  consistent for between  test  and three weeks. Testing was  completed between April and June, 1988. Administration Procedures: The standardized administration procedures were followed with modifications made in the communication mode. The communication mode  chosen  for  test  administration  was  a  form  of  Total  Communication  combining the use of sign, fingerspelling and speech. The sign system used was a form of Pidgin Sign English. The signed administration format was carefully  29  30 developed with the assistance of the Signing Consultant emploj'ed by the school. Local signs were incorporated where  appropriate and the formats of both tests  were  the  judged  to  be  consistent  by  Consultant.  conducted by the researcher, a graduate student  Test  administration  was  in Educational Psychology and  Special Education trained in the use of both assessment instruments. In addition, this researcher holds a Diploma in the Education of the Hearing Impaired and has been a classroom teacher of hearing impaired children.  B. SAMPLING Students  between  the  ages  of  8.0  and  16.11  years  enrolled  at  a  Provincial School for the Deaf had been invited to participate. The 38 subjects were a volunteer sample for whom parental permission had been obtained. They were  enrolled  in  a  residential/day  Communication  as  the  educational  School  for  methodology.  the  Deaf  which  The  subjects  had  uses  Total  severe  to  profound sensorineural hearing losses. There were 23 boys and 15 girls with an age  range  ascertained  of that  8.0 the  to  16.11  subjects  years  included  had neither  in  severe  the motor  study.  Medical  disabilities  records  nor visual  impairments. Etiologies were as follows: Etiology  _n_  unknown  19  heredity  14  rubella  4  menningitis  1  This sample had a high proportion of children whose deafness can be attributed to causes of heredity.  31 With  the  prelingually  deaf  student's  age  exception with  of  of  one  onset before  onset  indicated no observable  was  4 1/2  differences  student,  all  2 years  of age.  years. of his  the  Analysis  subjects The of  considered  1 postlingually  this  scores from the  were  individual's  sample  deaf data  group mean  scores. Audiological  information  gathered  from  the  medical  records  yielded  the  PIQ administration with  the  following sample description: Better Ear Average (BEA) Mean  lOOdB  Standard Deviation'  9.4dB  Range of Hearing Loss No  student in this  study  had  a  prior  75dB to 113dB WISC-R  exception of a single student, who had undergone testing two years prior to this data collection.  C. INSTRUMENTATION  1. WISC-R Performance Scale There  are  five  subtests  in  this  scale  plus  one  supplementary  subtest.  Scaled scores with a mean of 10 and a standard deviation of 3 are obtained on each  subtest. Together,  these subtests form the  Performance IQ which has  a  mean of 100 and a standard deviation of 15. For this study the 1974 Weschler norms were used.  Subtests Picture Completion This subtest consists of drawings of common objects. Each drawing lacks a  32 single important feature.  The drawings are presented  one  at a time  and the  child is required to point to the missing feature within a 20-second time limit. There are 26 drawings and the items are scored pass or fail. The subtest is discontinued after four consecutive failures. Picture Arrangement In this subtest the child is required to rearrange a set of pictures into a story sequence. progressively  There are  through the  12 sets of the picture stories. sets  and  fast  responses  Time limits increase  receive  bonus  points.  The  subtest is discontinued after three consecutive failures. Block Design Using  a  set  of patterned blocks,  the  child  is  required to  duplicate a  geometric design displayed on a card. There are 11 items of increasing difficulty. Time limits increase progressively and fast responses receive bonus points. The subtest is discontinued after two consecutive failures. Object Assembly For this subtest the child is required to assemble jigsaw puzzle pieces to form a common object. Each item has a time limit and fast responses receive bonus points. If the puzzle is incomplete after the time limit has expired, partial credit may be awarded. All children receive all 4 puzzle items. Coding This is a paper and pencil task. A Code Key is displayed at the top of the page, and the child is required to fill in as many code items as possible in the blank boxes below as possible within two minutes. Mazes This is a supplementary test and is usually not included in the composite  33  score. In this paper and pencil task, the child is required to accurately trace a path from the middle of a maze out to the exit. All items are timed and the number of errors made determines the child's score.  2. Stanford-Binet Intelligence Scale: Fourth Edition—Selected Subtests The  subtests  performance  type  selected  tests.  for  The  this  Verbal  study  Reasoning  because in this initial investigation  of the  estimate  be  of  intelligence  would  not  were  considered  Area  SB:FE,  supported  to  subtests  their  be  nonverbal  were  excluded  inclusion to form an  (Vernon  and  Brown,  1964;  Sullivan and Vernon, 1979). Neither was the Quantitative subtest included in this study  because of the  language  content of the  word problems  as  well as  the  reading requirement. It is well supported in the literature that reading levels in the hearing impaired population are significantly lower than in the non-exceptional population (King and Quiglej', 1985).  Order of Administration The booklet.  order of administration followed  the  For this study, in order to gather  as  sequence in the  test response  much data as possible,  it was  decided that each child would attempt each subtest. This is a departure from the organization of the SB:FE. Normally, the Vocabulary subtest is used as a routing test to determine, together with age, which subtests will be included in the test batterj\ subtests.  The  Vocabulary subtest  The entry  descriptions  listed  levels  below.  used  The  also  determines  in this  instructions  study for  the will be  the  entry  level  noted  adaptive  into  in the  testing  these subtest  procedures  described in the administration manual were followed in this study. If the basal level (four consecutive correct responses) could not be achieved at the entry level,  34 then testing would progress backwards until a basal level was achieved. For all subtests, test administration was discontinued when three out of four items were failed in two consecutive levels, as in standardized instruction.  Selected Subtests Abstract/Visual Reasoning Area Pattern Analysis Using  a  set  of patterned  blocks,  the  child is  required to  duplicate  a  geometric pattern which is initially modelled by the examiner and on later items is displayed on a card. The pattern must be correctly completed within a certain time  limit.  Entry  level  was  determined  by  the  basal  level  (four  consecutive  correct responses) attained in the preceding Bead Memory subtest. Copying In  this  subtest  the  child  is  required  to  reproduce  geometric  designs.  Depending on entry level, the nature of the task varies. Initially, young children are required to reproduce with unpatterned blocks, a model demonstrated by the examiner. The task then changes to a paper and pencil task copies a design  where  the child  from a displayed line drawing. Entry level for each child was  the first item of the paper and pencil task. Matrices Each item in this subtest consists of a configuration three  figures  complete  the  and  one  blank  cell.  The  matrix. In each item, the  child  missing figure  some logical relationship to the other figures. the first item.  selects  the  of cells containing best  alternative  to  for the blank cell  has  For all children entry level  was  35 Paper Folding and Cutting For each item, the child first looks at a sequence of drawings  showing  the stages of a piece of paper as it is folded and cut. The child then selects the  best  alternative  to  represent  how  the  folded  and  cut  paper  would look  unfolded. For all children entry level was the first item. Quantitative Reasoning Area Number Series The child is shown a sequence of numbers arranged according to a certain rule, and the child must indicate which two numbers would come next in the series. For each child entr}' level was the first item. Equation Building Each item consists of a series of numbers followed  by operation signs.  The numbers and signs must be rearranged to make a true number sentence. For each child entry level was the first item. Short-Term Memory Area Bead Memory Using  beads  of  various  shapes  and  colours,  the  child  is  required  to  reproduce from memory, a picture of beads arranged on a stick. The picture is displayed  for  five  seconds.  For each  child entry  level  was  initially item 11,  which is the first item using the beads on a stick format. Memory for Objects The child views pictures of objects presented  in a specific  order at one  second intervals. Then from a picture containing those objects plus several other distracting objects, the child must identify  those objects displayed in the correct  sequence. For each child entry level was the first item.  . 36 3.  Stanford  Achievement  Test  (7th  Ed.),  Special  Edition  for  Hearing  Impaired Students (Gardner et al., 1982) The SAT-HI is a norm-referenced achievement test. Separate norms have been  provided  for  Hearing Impaired  determined by grade levels. Tests  students.  E n t r j r  into  recommended for use  different  tests  is  with hearing impaired  students are: Reading Comprehension A sentence completion test with a multiple choice format. Spelling For each item, a group of phrases  are presented,  each one having one  underlined word. Students are to identify the incorrectly spelled underlined word. Language For each item,  a group of words are presented.  The student  has four  options: identify the group as (1) a complete sentence, (2) a run-on sentence, or (3) and (4) which are alternate choices to complete the sentence. Word Reading For each item, students match a picture with the correct word. Concepts of Number Numerical  concept  questions  are  provided  with  multiple-choice  answer  format. Math Computations Computation questions are presented with a multiple choice answer format. Math Applications Word problems are presented with a multiple choice answer format.  37 D.  A N A L Y S I S  All subject.  analyses for the  For younger  SB:FE  were computed  children who  obtained  norms, the Standard Age Score was table an  method with the  score  all subtests for each  in subtests beyond  their  obtained from the next highest age norm  available. For comparison purposes,  alternate  a  using  each subject's test was  Bead Memory basal level  rescored using  determining the entoy  level for the remaining tests, as suggested in the SB:FE Examiner's Handbook (Delaney  & Hopkins,  1987),  for  proficiency or are non-language Bl  testing  examinees  who  have  limited English  proficient. This comparison is displayed in Table  of Appendix B. For both test instruments,  the published norm tables were  used to obtain the Scaled Scores or Standard Age Scores and their composites. Correlations  between  the  Scale  and the  academic  SPSS-X  Data  Performance using  the  significance  selected  Analysis  was set at p = 0.05  subtests  of  the  measures of the  System,  (SPSS  SB:FE,  SAT-HI  Inc.,  the  WISC-R  were calculated  1988).  The  level  of  for this study. An exception was made for the  calculations involving the Language measure of the SAT-HI which had less than 10  subjects.  Therefore,  to  minimize  the  risk  of  inflating  Type  II  error to  SPSS-X  computer  untenable levels, alpha was relaxed to p = 0.10. The  t-test  for  significance  was  calculated  program to determine systematic differences  using  the  between the selected subtests of the  SB:FE and the WISC-R Performance Scale. The SB:FE  patterns  and the  of subtest intercorrelations for the  WISC-R Performance Scale  program MULTICORR (Steiger, 1979).  selected subtests of the  were compared using the computer  38 Item analysis on the selected subtests of the SB:FE was performed using the computer program LERTAP (Galan and Nelson, 1986).  C H A P T E R IV. RESULTS Chapter IV will present the results of the study and describe the results in relationship to the hypotheses tested. Hypothesis  1 stated that for the hearing impaired sample in this study,  the pattern of correlations between the WISC-R Performance Scale I.Q. (PIQ) and the  Abstract/Visual, Quantitative  and  Short-Term Memory Area  scores of  the  Stanford-Binet: Fourth Edition will not be significantly different from that reported for the non-exceptional sample. In order to test this hypothesis, the  Pearson R coefficients were obtained for  hearing impaired group. Table 3 displays  also the data reported for the non-exceptional  these correlation coefficients and sample. To determine whether the  pattern of correlations for the hearing impaired group was  significantly  from the  a test for the  pattern reported for the  non-exceptional  sample,  different null  hypothesis regarding the difference between independent correlation coefficients was employed. This test was carried out according to procedures described by Glass and Hopkins (1984, p. 307). The and with  difference  the non-exceptional the  impaired  group  the  is  significantly  sample.  for the  hearing impaired group  sample is significant for the SB:FE Abstract/Visual Area  PIQ. The correlation  non-exceptional and  in correlation coefficients  of  the  higher  two  test  instruments  than  the  correlation  The correlation between the  Short-Term Memory and  PIQ for  the  for  reported  Quantitative  hearing  the  hearing for  the  and the PIQ,  impaired group  was  computed to be not significantly different from the non-exceptional  data. Because  the  different from  pattern of correlations  for the  hearing impaired group was  that reported for the non-exceptional sample, Hypothesis 1 was rejected.  39  40 Table 3 Correlations Between Area Scores of the Stanford-Binet: Fourth Edition and the WISC-R PIQ Stanford-Binet Intelligence Scale (Fourth Edition) Abstract/Visual r  Quantitative r  Short-Term Memory r  82 n = 38  74 n = 35  72 n = 38  67  63  63  Hearing Impaired WISC-R PIQ Non-exceptional WISC-R PIQ  Note: All entries rounded to 2 significant figures, decimals omitted. All coefficients are significantly different from zero at a = .05. However, when comparing the hearing impaired coefficents to the non-exceptional coefficients, only the coefficients for the Abstract/Visual Scale (82 versus 67) are significantly different at a = .05.  Hypothesis 2 stated that for the hearing impaired group, the pattern of subtest intercorrelations of the SB:FE will not be significantly different from that reported for the non-exceptional sample. Hypothesis 3 stated that for the hearing impaired group in this study, the pattern of subtest intercorrelations of the WISC-R PIQ will not be significantly different from that reported for the non-exceptional sample. To 1979)  was  test these hypotheses, used.  Results  of this  the  computer  analysis  program MULTICORR  are summarized in Table  (Steiger, 4. These  results indicate that the pattern of intercorrelations of the SB:FE subtests for the hearing impaired sample is not significantly different than for the non-exceptional sample. Hypothesis 2 can therefore  be accepted.  For the  WISC-R Performance  Scale subtests, the pattern of intercorrelations for the hearing impaired sample is  41 not significantly  different than for the non-exceptional sample. Hypothesis 3 can  therefore be accepted.  Table 4 Subtest Pattern Intercorrelations Between the Hearing Impaired Sample and the Non-exceptional Sample Hypothesis  Chi Square  df  P  2 (SB:FE)  18.41  21  .62  3 (WISC-R PIQ)  21.17  15  .13  Hearing Impaired Sample: n = 38 Exploratory research questions were addressed as part of this study. 1.  To investigate  instruments, (Galan  item  item analysis was  and Nelson,  characteristics  of  the  various  subtests  of both  conducted using the computer program LERTAP  1986). The exploratory  item  analysis  provided the  results  displayed in Table 5. For a more comprehensive display of the item analysis of each subtest, the reader is referred to Appendix A. Examination of the subtest analysis  revealed  that  the  only  remarkable  finding occurred in  the  Equation  Building subtest. All subjects failed the first item in this subtest. As the SB:FE has  been  constructed  to  reflect  a  stepwise progression  failure of the first item is unexpected.  of item  difficulty,  the  42 Table 5 Item Analysis Characteristics for the SB:FE Hoyt Estimate  Raw Score Mean  Standard Deviation  Standard Error of Measure  Pattern Analysis  .92  31.92  6.44  1.84  Copying  .78  20.21  3.37  1.56  Matrices  .93  8.68  5.93  1.56  Paper Folding & Cutting n = 29  .91  2.76  3.59  1.07  Number Series n = 35  .90  8.79  4.84  1.53  Equation Building n=15  .85  1.16  2.14  0.81  Bead Memory  .86  21.53  5.03  1.84  Memory for Objects  .73  5.97  1.99  1.00  Subtest  1  1  No comparable values are given in manual; only transform score means provided.  Note: For all subtests: Cronbachs Alpha for Composite = n = 38 unless otherwise indicated.  .88  2. For the hearing impaired group in this study, systematically WISC-R  will either  instrument  yield greater Standard Age Scores? To conduct this analysis,  Performance  Scale  IQ was  converted  to  the  SB:FE  scale  using  the the  following formula: Z = [ ((WISC-R PIQ - 100)/15) X 16 ] + A correlated groups t-test was  100  performed between the  converted WISC-R  PIQ scores and each Standard Age Score of the SB:FE. Table 6 displays the results of this analysis. The WISC-R Performance Scale IQ scores are almost a  43  Table 6 Comparison of Standard Age Area Scores of the SB:FE and Converted WISC-R PIQ Scores for the Hearing Impaired n Test  Mean  Standard Deviation  Mean (Difference)  t  p  WISC-R PIQ SB:FE Partial Composite  38  99.89 85.32  21.27 17.58  14.57  7.27  <0.01  WISC-R PIQ SB:FE Abstract/Visual  38  99.89 87.34  21.27 17.49  12.55  6.27  <0.01  WISC-R PIQ SB:FE Quantitative  35  100.21 89.97  21.16 14.80  11.24  4.46  <0.01  WISC-R PIQ SB:FE Short-Term Memory  38  99.89 84.47  21.27 18.56  15.42  6.26  <0.01  a = .05 standard deviation higher than any of the SB:FE Area Scores. It can be stated that there is a significant for  systematic  difference  this hearing impaired group. The alternate  between these two  instruments  scoring method for the  SB:FE,  displayed in Table B l , produced a group partial composite mean of 85.24. The different scoring methods did not yield significantly different group means. 3. Performance  The  pattern  Scale  and  of the  subtest SB:FE  LERTAP (Galan and Nelson, 1986)  intercorrelations was  investigated.  between The  the  WISC-R  computer program  was used in this data analysis. The results  of this investigation are displayed in Table 7. The median intercorrelations of the various subtests of the WISC-R and the SB:FE produced coefficients  ranging from Picture Completion (WISC-R) at .64  44 Table 7 Subtest and Area Intercorrelations Between the SB:FE and the WISC-R PIQ  WISC-R PIQ  SB:FE  PT  CP  MX  PF  AV  NS  EB  QA  BM  MO  ST  PO  SM  PC  64  63  70  58  78  60  62  65  65  65  62  74  64  PA  56  58  67  32  66  68  79  72  64  54  67  73  61  BD  78  75  55  57  79  70  54  71  68  49  64  76  63  OA  55  65  49  38  63  57  54  59  56  30  50  60  55  CD  46  38  41  35  50  42  39  45  49  46  59  55  42  MZ  24  23  10  12  21  31  -5  32  22  26  27  29  33  PS  73  73  68  53  81  71  74  74  72  53  72  81  SM  56  61  52  37  59  54  60  48  Note: All entries rounded to two significant figures, decimals omitted, boldface coefficients significant at a = .05 n = 38 unless otherwise indicated below PC: PA: BD: OA: CD: MZ: PS: SM: PT: CP: MX: PF: AV: NS: EB: QA: BM: MO: ST: PO:  Picture Completion Picture Arrangement Block Design Object Assembly Coding Mazes Performance Scale I.Q. Subtest Median Pattern Analysis Copying Matrices Paper Folding & Cutting (n = 29) Abstract/Visual Area Number Series (n = 35) Equation Building (n=15) Quantitative Area (n=35) Bead Memory Memory for Objects Short-Term Memory Area Partial Composite  45 to Mazes (WISC-R) at  .33.  The following list displa3's  in rank order the  six  highest median subtest intercorrelations:  Subtest  The  r  Picture Completion (WISC-R)  .64  Block Design (WISC-R)  .63  Picture Arrangement (WISC-R)  .61  Copying (WISC-R)  .61  Bead Memory (SB-.FE)  .60  Number Series (SB:FE)  .59  highest single subtest correlational coefficients  were as follows: Block  Design (WISC-R) correlated .79 with the Abstract/Visual Area (SB:FE) and .78 with Pattern Analysis (SB:FE). Picture Arrangement (WISC-R) correlated .79 with Equation  Building  (SB:FE).  Picture  Completion  correlated  .78  with  the  Abstract/Visual Area score of the SB:FE. The highest correlation coefficients  of .81 were computed between both the  WISC-R Performance IQ and the Abstract/Visual Area and between the WISC-R Performance Scale IQ and the Partial Composite of the SB:FE. 4.  Both the WISC-R Performance Scale  and the  SB:FE  were correlated  with academic measures from the Stanford Achievement Test — Hearing Impaired (SAT-HI). For this analysis the Pearson R correlational coefficient  was computed  for each measure. Of the  various WISC-R Performance Scale subtests, Picture Arrangement  produced the most and the highest significant correlations. Object Assembly had no significant correlations with anj' of the academic measures.  Of the SAT-HI measures,  Number Concepts and Reading produced more  significant correlations (see Table 8). Language and Word Recognition produced no significant correlations with any of the WISC-R Performance Scale results.  Table 8 Correlations of the WISC-R PIQ and the SAT-HI SAT-HI Word Reading  Number Concepts  Math Comp  Math Applic  28 n=7 P = .269  35 n = 18 P == .075  48  n = 32 P == .003  27 n = 32 P == .005  n = 27 P == .002  29 n=7 P = .266  01 n = 18 P == .488  53  43  46  n = 32 n = 20 = .001 P= P == .019  n = 32 P == .001  n = 32 P == .007  n = 27 P == .007  Block Design  38 31 n = 32 n = 20 P == .017 P == .094  -29 n=7 P = .266  21 n = 18 P == .200  32  n = 32 P == .037  17 n = 32 P == .171  n = 27 P == .104  Object Assembly  23 31 n = 32 n = 20 P == .105 P == .089  -35 n=7 P = .221  -05 n = 18 P == .416  24 n = 32 P == .096  20 n = 32 P == .137  24 n = 27 P == .119  Coding  31 23 n = 32 n = 20 P == .042 P"= .161  20 n=7 P = .334  35 n = 18 P == .076  30  n = 32 P == .046  23 n = 32 P == .099  n = 27 P == .046  41 28 n = 32 n = 20 P == .063 P"= .035  -52 n= 7 P = .117  07 n = 18 P == .386  n = 32 P == .017  -01 n=7 P = .489  20 n = 18 P == .219  44  32  43  n = 32 P == .005  n = 32 P == .037  n = 27 P == .012  WISC-R PS Picture Completion  Picture Arrangement  Mazes  Performance I.Q.  Reading Comp Spelling Language 49  39  n = 32 n = 20 = .002 P= P == .043 54  46  47  42  n = 32 n = 20 = .004 P-= .032 P=  37  ,  37  n = 32 P == .019  54  25  33  31 n = 27 P == .056  Note: Boldfaced coefficients significant at a = .05, except where n<10 then a = .10. All entries rounded to two significant figures, decimals omitted.  Table 9 displays  the results of the correlations  between the SB:FE and  the SAT-HI. Among the SAT-HI measures, Reading Comprehension, Language and Number  Concepts  coefficients.  produced  Language  the  produced  the  greatest  number  of  highest coefficients.  significant Of the  correlational  various  subtests, the Short-Term Memor}' Area produced the most significant  SB:FE  correlations.  The Paper Folding and Cutting subtest yielded no significant correlations.  Table 9 Correlations of the SB:FE and the SAT-HI SAT-HI Reading Comp Spelling Language  Word Reading  Number Concepts  Math Comp  Math Applic  Pattern Analysis  21 40 n = 32 n = 20 P == .011 P == .186  19 n=7 P = .341  24 n = 18 P == .170  40 n = 32 P == .012  18 n = 32 P == .158  29 n = 27 P == .072  Copying  32 20 n = 32 n = 20 P == .037 P == .197  -17 n=7 P = .360  21 n = 18 P == .202  24 n = 32 P == .096  07 n = 32 P == .352  24 n = 27 P == .113  Matrices  49 18 n = 32 n = 20 P == .002 P == .229  52 n=7 P = .144  21 n = 18 P == .204  38 n = 32 P == .017  27 n = 32 P == .069  32 n = 27 P == .053  Paper Folding & Cutting  -07 10 n = 26 n = 17 P = .314 P == .391  27 n=7 P = .283  47 n = 12 P == .064  05 n = 26 P == .413  -25 n = 26 P == .107  -02 n = 22 P == .463  Abstract/Visual Reasoning 42 22 Area n = 32 n = 20 P == .008 P == ,182  32 n=7 P = .241  31 n = 18 P == .107  37 n = 32 P == .020  16 n = 32 P == .193  31 n = 27 P == .061  63 n=7 P = .064  12 n = 17 P == .326  29 n = 31 P == .060  09 n = 31 P == .312  27 n = 26 P == .091  SB:FE  Number Series  :  39 24 n = 31 n = 19 P == .016 P == .164  48 Table 9 continued  SAT-HI Reading Comp Spelling Language  Word Reading  Number Concepts  Math Comp  Math Applic  33 n=14 p = .126  83 n=8 p = .003  40 n=5 p = .254  -79 n=4 p = .105  36 n=14 p = .103  -03 n=14 p = .462  29 n=10 p = .212  41 n = 31 p = .011  26 n=19 p = .138  57 n=7 p = .093  16 n=17 p = .264  36 n = 31 p=.024  12 n = 31 p = .253  31 n = 26 p = .064  Bead Memory  50 n = 32 p = .002  23 n = 20 p = .162  63 n=7 p = .066  15 n=18 p = .274  27 n = 32 p = .067  14 n = 32 p = .225  24 n = 27 p = .110  Memory for Objects  43 n = 32 p=.007  11 n = 20 p = .323  64 n=7 p = .059  01 n=18 p=.480  33 n = 32 p=.031  21 n = 32 p = .119  22 . n = 27 p=.132  52 n = 32 p = .001  21 n = 20 p = .186  65 n=7 p = .059  09 n=18 p = .355  33 n=32 p = .031  19 n=32 p = .145  27 n = 27 p = .085  48 n = 32 p = .003  24 n=20 p = .155  55 n=7 p = .100  21 n=18 p = .204  37 n=32 p = .018  17 n=32 p = .176  31 n = 27 p = .059  SB:FE Equation Building Quantitative Reasoning Area  Short-Term Memory Area Partial Composite  Note: Boldfaced coefficients significant at a = .05, except where n<10 then a = .10. All entries rounded to two significant figures, decimals omitted.  C H A P T E R V. SUMMARY A N D DISCUSSION In  this final chapter, the purposes, procedures, and results of the study  are summarized. The major findings and their implications are then discussed and an orientation for further research is suggested.  A. SUMMARY The  present  study  proposed to  evaluate  subtests of the new Stanford-Binet Intelligence for  use  Scale  appropriateness  of selected  — Fourth Edition (SB:FE)  with hearing impaired children whose educational methodology  Communication. were  the  similar  It was for  reasoned  both  the  that if the  hearing  test characteristics  impaired  group  and  is Total  of the  the  SB:FE  non-exceptional  standardization sample then the test could be used for its intended purpose. In the  field  of education,  tests  of intelligence  are  traditionally used  to estimate  general cognitive ability and to predict potential levels of academic achievement. The for  the  subjects used in this stud}' were enrolled in a residential/day School  Deaf  whose  educational  methodology  was  Total  Communication. The  subjects were tested on both the selected nonverbal subtests of the SB:FE and the  Performance  Scale  of the  WISC-R.  The order of test administration  was  counterbalanced and a consistent Total Communication format was used for both tests. To examine test characteristics of the SB:FE, the data gathered from the group of hearing impaired children were compared with the standardization data for  the  non-exceptional  comparisons  sample  involved correlations  as  reported  with  the  in  the  Performance  technical Scale  manual. of  the  These  WISC-R,  patterns of subtest intercorrelations, as well as exploratory item analysis of the  49  50 SB:FE.  Correlations were  measures  of academic  also  computed  achievement.  between  each  cognitive  measure  In addition, the Area Scores of the  were compared to the Performance IQ of the WISC-R to detect any  and  SB:FE  systematic  differences between these two instruments.  B. DISCUSSION OF RESULTS Hypothesis  1 stated that for the hearing impaired sample in this study,  the pattern of correlations between the Performance Scale of the  WISC-R and  the Abstract/Visual, Quantitative, and the Short-Term Memor}' Area scores of the SB:FE  would  non-exceptional  not  be  sample.  significantlj' It is  different  recalled that  correlations were reported as being .63 and  from  for the  that  reported  non-exceptional  between the  for  the  sample,  the  Short-Term Memory Area  the WISC-R PIQ, .63 between Quantitative and the PIQ, and .67 between  Abstract/Visual and the PIQ. The pattern is that of similar correlations between the  Short-Term  relative^ results  higher  indicate  Memory  and  the  Quantitative  correlation between that  this  general  the  Areas  with  the  Abstract/Visual Area  pattern is  PIQ, with and  mirrored in the  a  PIQ. The  hearing impaired  sample of this study. The correlation between Short-Term Memory and the PIQ was  .72, between Quantitative and the PIQ it was .74, and between Abstract/  Visual  and the  impaired  sample  PIQ it and the  were no significant  was  .82.  Comparing  non-exceptional  differences  the  sample  between either  the  correlations  yielded  the  of result  the  hearing  that  there  Quantitative Reasoning Area  and  the WISC-R Performance IQ or between the Short-Term Memory Area and  the  WISC-R Performance IQ. In this  PIQ  correlation coefficient  was  sample  however,  the Abstract/Visual and  great enough to be significantly  higher than its  51 counterpart in the non-exceptional sample. This pattern for both the non-exceptional sample and the hearing impaired sample is logical considering the nature of the subtests from either instrument. The  Performance Scale of the WISC-R does not have a separate  with visual short-term memory as does the Scale  have  a  specific  subtest  dealing  SB:FE.  with  subtest dealing  Nor does the Performance  numerical  concepts  as  does  the  Quantitative Area of the SB:FE. The Arithmetic subtest of the WISC-R is to be found in its Verbal Scale rather than in the Performance Scale. The Abstract/ Visual  Reasoning Area of the  SB:FE  appears to have  more content similarity  with the WISC-R Performance Scale. For example, Block Design of the WISC-R and  Pattern Analysis of the  designs  with  intelligence  patterned  measures  SB:FE  blocks.  produced  The a  both require the reproduction of geometric subtest  coefficient  intercorrelations  of  .78  for  the  for  these  two  hearing impaired  children of this sample, confirming the suggestion of apparent content similarity. Sattler (1988) has suggested that the Performance Scale of the WISC-R might be considered to be an index of nonverbal and fluid intelligence involving immediate problem solving ability. In most subtests the stimuli are presented visually and may be described as perceptual organization skills. This description coincides with the  information  provided  describes the three-level  in  the  Administrative  Manual  of  the  SB:FE  hierarchical model of the structure of cognitive  that  abilities  upon which the SB:FE has been based. According to this model, the Abstract/ Visual Reasoning subtests are from the second level ability termed Fluid/Analytic Abilities (Thorndike, Hagen and Sattler,  1986a). These abilities  are thought to  require cognitive skills necessary for solving new problems that involve figural or other nonverbal stimuli. It is apparent that the intent of these areas of both  52 instruments is to assess the construct of fluid intelligence.  Both tests also use  predominantly nonverbal performance-type tasks using visual perceptual abilities to tap  this  dimension  of intelligence.  Considering these  similarities  of intent and  means for the WISC-R Performance Scale and the Abstract/Visual Area of the SB:FE, the results indicating higher correlations for these areas would reasonably be expected. Hypothesis 2 stated that for the hearing impaired sample in this study, the pattern of subtest intercorrelations among the Stanford-Binet: Fourth Edition will not be different from that reported for the standardization sample. Results of this study found that there were no significant differences found for the hearing impaired group. This suggests that the dimensions of intelligence being tapped by the SB:FE are operating in a similar manner for both populations. This in turn suggests  the  subtest  specificity  is  similar  for  both  populations.  The hearing  impaired children do not seem to be substituting an}' cognitive abilities or skills that are different from those used by the standardization sample. The  exploratory  item  similar subtest specificity. of  item difficulty  satisfactory  analysis  results  confirm this  This exploratory investigation  general  evidence  of  revealed the progression  and test reliability throughout each subtest was  found to be  with the exception of the Equation Building subtest. A discussion of  the possible reasons for this finding is deferred to a later analysis of the group subtest means profile. The  similarity of patterns of subtest intercorrelations for both populations,  as well as the acceptable item characteristics of most of the subtests indicates thus  far that  this  instrument operates in a similar manner for the  impaired sample as it does for the standardization sample.  hearing  53 Hypothesis  3  compared  the  subtest  intercorrelations  of  the  Performance Scale with the standardization sample. The study results significant  difference  for  the  hearing  impaired  group.  The  WISC-R found no  dimensions  of  intelligence tapped by the WISC-R Performance Scale appear to be operating in a similar  manner  for  both  populations.  administration mode for the testing  It  will  be  recalled  that  the  choice  of  conducted in this study was based on the  work of Sullivan (1982) who found increased mean scores and reduced subtest differences when using a TC communication mode with children whose educational methodology was TC. The results of this analj'sis and the group subtest profile displayed in Figure 1 seem to support Sullivan's findings. As well, factor analytic research has revealed similar "principal factors extracted from deaf and hearing samples" on the WISC-R (Braden, 1985, p. 378). It seems reasonable to conclude that  the  choice  of  administration  mode  was  correct  and that  this particular  sample of hearing impaired children can be considered to be consistently The exploratory research questions compared these two cognitive in  several  ways.  The  significant differences, addition,  each  measures. regarding  the  of  the  total  test  scores  were  measures  compared  for  the subtests of both measures were intercorrelated and, in  cognitive  The  means  average.  results  measure of  these  appropriateness  of  was  correlated  investigations the  SB:FE  with  have with  the  SAT-HI  produced this  some  population  academic questions  of hearing  impaired children. The  most  significant  comparison  comparison. When the test results converted consistently  to the  same scale  was  the  total  test  of the hearing impaired study  and tested for  significance,  the  score  means  sample were  differences  were  almost 15 points which was significant at the .001 level. The scores  54  Picture Completion  Picture Arrangement  Block Design  Object Assembly  Coding  Mazes  19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1  Figure 1 WISC-R PIQ Subtest Profile  1  1  Mean of 10, standard deviation of 3  from the Stanford-Binet were a standard deviation  lower than the scores from  the WISC-R Performance IQ. Such a significant difference for the non-exceptional Technical  Manual  population. As a basis for comparison, it is noted in the  of the  Stanford-Binet  mean WISC-R Performance IQ was Visual  Reasoning  has not been reported  Area  was  98.9,  that for a non-exceptional  105.3 the  and from the Quantitative  SB-.FE,  Area  was  sample  the  the Abstract/ 102.1,  and  Short-Term Memory was 102.6. Thus the score discrepancies found between these two instruments in this hearing impaired group are striking. A difference of this  55 magnitude  between  two  cognitive  measures  certainly  precludes  their  use  as  interchangeable instruments. Estimates of cognitive ability are often an important factor in educational placement decisions. A child classified according to one test would most likely receive a different classification  if judged by the other test.  This would result in considerable confusion and lack of consistency in placement decisions  for hearing impaired children. This is especially  recommendation  of  Vernon  who  (1979),  plurality  of  advocated  assessment that  one  measures  true because  made  test of cognitive  by  of the  Sullivan and  ability be  used  to  confirm the results of another. This is deemed necessary in the hearing impaired populations because  of the unique problems associated with testing the cognitive  abilities of hearing impaired children. The conclusion that these two instruments cannot be used interchangeably for  this  hearing impaired sample  regarding  the  interchangeability  supports, of  the  in part, Sattler's WISC-R  and  the  (1988) conclusion SB:FE  for  the  non-exceptional population. He cites the correlations as being .66 to .83 between the WISC-R Full Scale and the SB:FE Composite. Sattler concludes that the two tests yield scores that are approximately equal but they are not interchangeable. It is therefore concluded that the values of the correlations produced from this study  are  similarly not high enough  to  permit recommendation for their  use  together as interchangeable instruments of cognitive ability. The  marked  mean  score  discrepancy  between  these  instruments  adds  further support to this conclusion. It is again noted that these large mean score differences  were  not  found  in  the  correlational studies  in  the standardization  sample. At this point in the discussion, the possible reasons for a discrepancy of this magnitude in the hearing impaired sample should be suggested. The question  56 to be addressed  then is, what are the differences  between these two  measures for the hearing impaired sample? It seems reasonable the differences  cognitive  to suspect that  lie in the nature of the subtest task demands. It will be recalled  that for this study, sources of external contamination were anticipated and steps were undertaken to minimize their effects. Subjects with identified motor or visual impairments  were excluded  from the  study.  It was  also  anticipated  that  the  WISC-R PIQ results would reveal an}' unusually unique population characteristics for this study's hearing impaired sample. The administration procedures were also identified  as being a possible  source  of contamination. It was  decided that for  this study, a consistent Total Communication format would be used for both test administrations. As explained in Chapter II, Sullivan's (1982) work was influential in  this  differences  decision. when  Sullivan a  administration of the  had  Total  found  increased  Communication  scores  format  and  was  minimized  used  in  subtest  the  test  WISC-R PIQ with hearing impaired children using Total  Communication as the educational methodology.  The results of this study concur  with Sullivan's findings. The group subtest profile displayed in Figure 1 shows this group to be average in all subtests compared to the standardization sample. This finding, in addition to supporting the Total Communication format issue, also provides information about this group's unique characteristics.  The results of the  WISC-R  group  PIQ  administration  have  impaired children to be consistently  revealed  this  particular  of  hearing  average in the measured abilities. The mean  Performance IQ for this group was close to the mean Performance IQ for the norming  sample  as  reported  in  the  manual.  The  standard  deviation  was  considerably greater for this hearing impaired group. This is consistent with the findings from other studies using the WISC-R PIQ which found greater standard  57 deviations  for  their  hearing impaired samples  (Ritter,  1976;  Hirshoren et al.,  1977; Anderson & Sisco, 1977; Brooks & Riggs, 1980; Watson & Goldgar, 1985; Phelps & Ensor, 1986). These larger standard deviations values may be reflecting the  greater  possibility  of  additional  handicaps  due  to  the  different  etiologies  responsible for the hearing impaired. The standard deviation found in this study was  even  larger  than  those  previously  reported,  which  may  also  be  a  characteristic of this particular sample. It can be concluded that in the cognitive abilities tested by the WISC-R PIQ, the hearing impaired group in this study was  similar to  their non-exceptional  age  peers.  This  finding supports  previous  research reviewed in Chapter II that suggests hearing impaired children are in the average range on measures of nonverbal intelligence (Vernon 1968). For group  the purposes of this study it can be concluded that this particular  of  hearing  impaired children appear  to  be  average  in  their  cognitive  abilities as measured by the WISC-R PIQ. It can further be suggested that the Total Communication format appeared to be appropriate. Since the TC formats for both instruments were judged as being consistent by the Sign Consultant, It seems justifiable  to assume that the TC format for the SB:FE  was similarily  appropriate. The cognitive  selected subtests of the  SB:FE  may be measuring a dimension of  abilitj' that this hearing impaired sample  are relatively low in when  compared to their age appropriate peers in the standardization sample. Referring to Table 6 it will be remembered that the  mean score differences  are fairly  consistent across all areas. For the hearing impaired sample the scores were: 87 in Abstract/Visual reasoning, 89 in Quantitative reasoning, and 85 in Short-Term Memory.  This uniformity of score  results  seems to indicate there  may be an  58 underlying characteristic present in these selected subtests of the SB:FE that is responsible for the generalty depressed scores. It is also noted that the standard deviation was reduced as well for this hearing impaired sample in comparison to the WISC-R PIQ results. The  group mean subtest profile displayed in Figure 2 permits  a more  detailed picture of the results. The two lowest scoring subtests; Equation Building and  Copying,  may  have  had  interfering  external  factors  operating.  In  the  Equation Building subtest, item anatysis revealed all subjects in the study group failed the first item. This is interpreted as reflecting a difference in curriculum content  for  contains  items consisting  rearranged  this  to  arrangement  particular group  make  which  a  of  hearing  impaired children. This  of a series of numbers and symbols true  needed  number  operations  number sentence 5 + 2 = 3 + 4 is  sentence. on  both  an example  The sides  first of  of this  that are to be  item  the  subtest  required  equal  arrangement.  an  sign. The The example  items for this subtest did not follow this format, but rather had only one digit following the been  so  equal sign. It could be that the  accustomed  to  performing  hearing impaired children have  mathematical  operations  in  the  format  demonstrated in the example items that an alternate format did not appear as a possibility. The Copying subtest was also observed to reflect a possible difference in curriculum emphasis for the hearing impaired children. In the administration of this  subtest, the  examiner  noted  a consistent attitude  of disregard for precise  duplication of the geometric designs. The scoring criteria of this subtest are very strict and designs may be failed, for example, for incomplete closure of a circle, or for a tilt in excess of 10 degrees of a design on the page. As a group, the hearing impaired students  seemed inattentive  to  these fine  details.  A possible  59  PA  CP  MX  PF  EB  NS  74 72 70 68 66 64 62 60 58 56 54 52 50 48 46 44 42 40 38 36 34 32 30 28 26  Figure 2 SB:FE Subtest Profile ^ean  of 50, standard deviation of 8  Note: PT: CP: MX: PF: NS: EB: BM: MO:  Pattern Analysis Copying Matrices Paper Folding & Cutting Number Series Equation Building Bead Memory Memory for Objects  1  BM  MO  60 reason for this is that there is such an emphasis on language acquisition and development that attention to details such as neatness of work presentation may not be accorded the same emphasis There  is  also  the  possibility  as in a class of non-exceptional  that,  although  the  students.  studj' group did not include  children with obvious motor disabilities, there may have been children with minor fine motor disabilities. The research literature does indicate that the population of hearing impaired children contains a higher incidence of motor dysfunctions due to the etiologies of the hearing impairment (Vernon, 1968). For  the  underlying  remaining  abilities  that  subtests,  account  the  possibility  of  the  depressed  scores  for  there  being  still  common  remains. The  results of the exploratory research questions investigating the correlations between both  these  intelligence  measures  and  academic  measures  provide  the  first  indication of what this underlying characteristic may be. Both instruments had significant  correlations with the Reading and Number Concepts of the  SAT-HI.  The WISC-R PIQ correlated significantly with more of the SAT-HI measures than did  the  SB:FE.  instruments the  However  most  striking  and their correlations with the  Language measures  between the Composite  the  SB:FE  of  the  of the  and the SB:FE  SAT-HI.  difference  academic measures  Language  of the  produced  a  these  was  The largest coefficients  Language measures  and  between  two  found with  were produced  SAT-HI.  correlation  The Partial of  .55.  The  Quantitative Area correlated .57 with Language, The Short-Term Memory Area yielded  the  highest  correlation of  .64  with  Language.  This  particular finding  correlating short-term memory skills and language supports work done by Watson et al.  (1982)  who  suggested  that  visual  short-term  English language learning in hearing impaired children.  memory  is  important in  These Findings for the SB:FE are contrasted with the complete absence of significant correlations between this Language measure PIQ  subtests.  instruments  This  for  may  this  be  hearing  an  indication  impaired  of  sample.  and any of the  differences The  between  WISC-R  WISC-R  these  two  does  not  PIQ  correlate with Language whereas the SB:FE does. The question that now arises is how is language involved on the SB:FE? To  answer  instruments  this question the nature of the task demands  should  be  examined.  The  Examiner's  Handbook  of these two  of  the  SB:FE  (Delaney and Hopkins, 1987) provides a description of the abilities and influences for  each test. While these are considered arbitrary designations  is made by the approaches  when  and no attempt  authors to claim that all examinees will employ all potential solving  a  given  task,  these  listed  abilities  and  influences  nevertheless reveal information about the nature of the subtests. The abilities and influences analysis,  mentioned visual  for  imagery,  the  Abstract/Visual  spatial  subtests  visualization,  involve  planning  skills  ability  in  and  visual  inductive  reasoning. As well, performance may be subject to the influence of flexibility. For the  Quantitative subtests, performance is thought  to reflect  skills in numerical  fluency, math concepts and computation, and inductive reasoning. Performance is also thought to be influenced by flexibilty. For the Short-Term Memory subtests, performance is thought to reflect skills in visual memory; visual perception; visual analysis;  sequencing,  chunking, or  clustering  strategies;  and  a  verbal labeling  memory strategy. These abilities and influences are relevant for a non-exceptional population. Whether they are also relevant to a hearing impaired population is not  3'et known.  pattern  of  However  subtest  the  results  intercorrelations  from  this  (Hypothesis  study's 2)  investigation  indicate  that  the  into  the  SB:FE  62 subtests are interrelating in a similar manner for both populations. It appears that  the  hearing  impaired children were  not  substituting  discernably  different  strategies. Considering the  descriptions  of the  abilities  required by the  subtests of  these two instruments leads to a suggestion that while there are similar abilities involved in the visual perceptual area, the  SB:FE  thinking skills involving planning ability, strategy skills and flexibility.  also seems to involve more formation, inductive  reasoning  The time factor element may be an important distinction  between these two tests. While both tests require visual perceptual skills and problem solving abilities, the time element may suggest different problem solving strategies must be employed for either test. The WISC-R Performance Scale is designed  to reward immediate  problem solving skills. The SB:FE,  power test by its authors, places  more emphasis  considered a  on problem solving strategies  employing more reflective thinking skills. The inductive reasoning required by the SB:FE  seems to  cautioned  however,  require a that  sequential  ordering of  because of the  information. The reader  limited sample  size  available  is  for the  SAT-HI Language measures, rather than being regarded as empirical support, the results have  served to initiate speculation regarding the association  of language  and the SB:FE. This  suspected  association  found  further  support  with  the  WISC-R  Performance Scale and SB:FE subtest intercorrelations. Drawing on the results of the  exploratory research questions  involving the  subtest intercorrelation of both  the instruments provides another clue to the possible nature of these underlying abilities.  The subtest intercorrelations  analysis  found  Picture  Completion, Block  Design, and Picture Arrangement, to yield the highest median correlations of .64,  63 .63, and .61, respective^, indicates  with the Stanford-Binet: Fourth Edition subtests. This  that these WISC-R Performance subtests share the  most commonality  with the Stanford-Binet: Fourth Edition subtests. The correlational analysis with the  SAT-HI  academic  Arrangement  were  measures  most  indicated  highly  that  correlated  Picture Completion and Picture  with  academic  achievement.  It  is  interesting to note that in a review of factor analysis of the WISC-R for the non-exceptional  population,  Sattler  (1988)  stated  that  Picture  Completion and  Picture Arrangement had moderate loadings on the Verbal Comprehension factor. This  was  mediation  interpreted to to  a  greater  suggest that degree  than  provides an insight for a possible  these  two  do the  subtests  other  explanation  may  require verbal  Performance  subtests. This  for the  commonality  these two subtests, the Stanford-Binet subtests, and higher language  shared by achievement  in the hearing impaired group. The use of language as a cognitive tool may be the  most  common  indications  appear  factor to  shared  suggest  in  that  these  these  different  selected  cognitive  nonverbal  tasks.  These  performance  type  subtests of the SB:FE reflect a model of cognition that links cognitive functioning with language use. That cognition may be facilitated by language is certainly not a novel idea in the study of the cognition.  Vygotsky  (1962)  has  nature of the relationship of language and produced  a  framework  for  the  study  of  internalized, private speech and its role in directing cognition. Vygotsky sought to trace the development of speech from its initial interpersonal use to its eventual intrapersonal use.  Luria  (1961) also believed  that internalized speech comes to  control thought and action. From Luria's research it seems apparent that people who develop a verbal language also develop verbal language mediation strategies to  facilitate  cognition.  The use  of  language  as  a  cognitive  tool  in  devising  64 memory strategies or in the inductive reasoning process may well be the common underlying factor necessary for successful execution of these subtests. This is not to suggest that hearing impaired individuals do not use language as a cognitive tool. Indeed, the nature of the systems used by the hearing impaired population for symbolic mediation is an active area of research. A discussion of this field is deferred to the directions for further research section. important  issue  is  that  the  standardization  At this point the most  population  has  access  to verbal  language based cognitive tools that may not be equally available to most deaf children. This leads to the suggestion that the level of language development may have a strong influence in these particular subtests. It could be that cognition and  linguistic levels may not be independent in these subtests. It is  accepted  that most hearing impaired children may be considered delayed in their language development (Quigley and Paul, in measures  where language  1984; Kretschmer and Kretschmer, 1978). Thus  and cognition are thought to interact, the hearing  impaired children would display a lower performance level when compared with their non-exceptional may  age  peers  where  standard levels of language  be more readily assumed. This suggestion  SB:FE  uses  to  determine  utilizes  a multistage  entry  level  adaptive testing  into  development  is reinforced by the method the  the  different  subtests.  The  SB:FE  pattern. In the first stage, the examiner  gives the Vocabulary Test which serves as a routing test. The Vocabulary Test level attained and the chronological age are used to determine the entry level for all other subtests. Within the organization and structure of the SB:FE there is in effect a screening process whereby a child is determined to have a certain level of language before attempting certain subtests. For  the  purposes  of  this  study,  it  is  thought  that  there  is  enough  65 evidence  to  suspect  that  these  nonverbal  selected  subtests  of  the  SB:FE  nevertheless require a facility with language as a cognitive tool. For the hearing impaired sample, the tests results from this instrument may therefore be more a reflection of language  and/or symbolic mediation abilities than of the underlying  cognitive structures. The low scores reported for this hearing impaired group were obtained despite very conservative selection of SB:FE subtests. The battery chosen for this study did not include several of the subtests recommended in the SB:FE Examiner's Handbook for a deaf examinee.  Several recommended subtests were  excluded  reading  due  to  their  overt  language  or  requirements.  This  study's  battery more closely resembled the battery recommended for examinees considered to have  Limited English Proficiency or who  are Non-Language Proficient. This  finding therefore, would suggest even greater concern over the suitability of the recommended battery  for Deaf Examinees  and may  also have  implications for  other populations of exceptional examinees.  C.  CONCLUSIONS The  results of this study suggest the following conclusions regarding the  appropriateness  of  the  nonverbal  selected  subtests  of  the  SB:FE  for  this  population of severe to profoundly hearing impaired children whose educational methodology is Total Communication. 1. Results from the selected subtests of the SB:FE and from the WISC-R Performance differences  Scale  cannot  between the  be used  interchangeably.  scores of these two  The very  instruments  significant  mean  for this population of  hearing impaired children preclude their use together as confirmatory estimates of cognitive ability.  66 2. There are indications that the selected subtests of the SB:FE may not measure  cognitive  research  using  structures  the  general cognitive  SB:FE  independently  is  necessary  from  before  language.  it  is  used  Therefore, as  further  an indicator of  ability for hearing impaired children whose levels of language  development may be delayed.  D. LIMITATIONS OF T H E STUDY The primary limitation of this study may be considered typical of research with special low incidence populations. A balance, or compromise must be made between the number of population variables chosen for control, and the number of subjects available for the research study. The population variables selected for control in this  study,  important for the thought  to  attending  be a  as  explained in Chapter  nature of this  representative  of most  day/residential  Communication.  The results  study.  III,  were  considered the  The population used in this  most  stud}' is  populations of hearing impaired children  school  whose  educational  methodology  of this  study  can therefore  only be  is  Total  regarded as  having implications for similar populations. Another procedures  limitation  with  of  standardized  the  study  tests.  is  This  the  use  issue  of modified administration  represents  another  necessar3'  compromise in research populations with special needs. The effects of the modified procedures were minimized by conferring with the suitability  of  Nevertheless,  the  signs  used  as  well  as  the  school's  consistency  Sign Consultant for across  both tests.  such test administration modifications would introduce an unknown  degree of variation. Yet  another  limitation  of  this  study  is  the  very  small  sample  size  67 available to calculate the SAT-HI Language correlations. The seven students who completed this test were the older students in the stud} . This would certainly 7  limit the generalizability of these results.  E . DIRECTIONS One SB:FE  of the findings of this study indicated the selected subtests of the  may  have  comparatively academic  F O R FURTHER R E S E A R C H  predictive  high  validity  correlations  achievement  measures  for  this  hearing  produced  between  needs  be  to  impaired  this  verified.  sample.  instrument If  subsequent  and  The the  research  confirms this finding, then an understanding of how language is involved in the execution  of these subtests may provide information that might be incorporated  into educational practices. Certainly, language learning is a central focus in the education of deaf children. The SB:FE seems to require the use of language-based cognitive strategies. Whether these strategies are based in verbal language (English) or whether these cognitive  strategies may be based  investigated.  in a form of Sign Language needs to be  The high correlation between English language and the SB:FE need  not imply a direct causal relationship. The work of Cummins (1984) may have implications for this investigation.  Cummins proposed a model that suggested the  level of language proficiency in the first language is directly related to proficiency in  the  second  language.  The  applicability  of  Cummins' framework  could be  verified using carefulfy defined subgroups of hearing impaired children. The most important  variable  development. cognitive  in  this  investigation  would  be  level  of  first  language  If it is found that a hearing impaired subgroup apparently uses  strategies  as  effectively  as  their  age  appropriate  peers  in  the  68 non-exceptional population, further investigations the cognitive strategies. a  verbal language  could probe the exact nature of  It is possible that some cognitive strategies may employ  based  system  in the  Short-Term  Memory  subtests. Conrad  (1979) has conducted studies indicating that some deaf people use speech coding in short term memory tasks. It is also conceivable that there are other cognitive  strategies necessar}'  for the execution of the subtests requiring certain problem solving skills. It has been  noted  that  several  of the  subtests require inductive  reasoning.  As  this  reasoning process seems to require the deliberate, sequential progression of logical statements,  it  could  be  that  a  system  of  symbolic  mediation  is  employed.  Research in this area is growing. King and Quigley (1985) have concluded that the many studies in this area demonstrate a growing belief that a gestural form of language, such as American Sign Language, is probably an efficient thoughtmediating  system.  The  suggestion  that  a  reflective,  sequential  progression  of  logical thoughts characterize the inductive reasoning processes required by several of the SB:FE subtests leads to yet another interesting speculation. Perhaps the effects of the dela}' of language  development is related to the development of  successive cognitive processes as suggested by the model of cognitive functioning proposed by Das, Kirby, and Jarman (1979). An academic  age and  appropriate internalized behavioural  consequences.  language It  is  system anticipated  may that  have  significant  these problem  solving skills maj' be of greater academic consequence in the higher grades and at the post secondary levels. As well, the ability to anticipate possible outcomes and modify or choose alternate behavioural responses may be facilitated by an internalized language-based mediation system.  69 To the extent that such an internalized mediation system is important in acquiring  certain  problem  solving  skills,  the  the  important in providing an estimate of these abilities.  SB:FE  may  prove  to  be  REFERENCES  Allen,  W., Abraham, S., & Stoker, R. (1988). Providers and practices in psychoeducational assessment of the hearing impaired in educational settings. Diagnostique, 14(1), 26-48.  Amoss, H. (1949). Ontario School Ability Examination. Toronto: Ryerson Press. Anastasi, A. (1982). Psychological testing (5th ed.). New York: Macmillan. Anderson, R.J. & Sisco, F . H . (1977). Standardization of the WISC-R Performance Scale for deaf children (Series T.No.l). Washington, D.C.: Gallaudet College Office of Demographic Studies. Arthur, G. (1955). Arthur Adaption of the Leiter International Performance Scale. New York: The Psychological Corporation. Baker, C. & Cokely, D. (1980). American Sign Language: A teacher's resource text on grammar and culture. Silver Spring, Maryland: T . J . Publishers. Braden,  J.P. (1985). WISC-R deaf Psychology, 23, 375-382.  norms  reconsidered.  Bragman, R. (1982). Review of research on test instructions American Annals of the Deaf, 127, 337-346. Brill,  Journal  of  School  for deaf children.  R.C. (1962). The relationship of Wechsler IQs to academic achievement among deaf students. Journal of Exceptional Children, 28, 315-321.  Brooks,  C.R., & Riggs, S.T. (1980). WISC-R, WISC, and reading achievement relationships among hearing-impaired children attending public schools. The Volta Review, 82(2), 96-102.  Burgemeister, B.B., Blum, L . H . , & Lorge, I. (1972). Columbia Mental Maturity Scale. New York: Harcourt Brace Jovanovich, Inc. Cummins, J . (1984). Bilingualism and special Press Inc.  education.  San Diego: College-Hill  Conrad, R. (1979). The deaf schoolchild. London: Harper & Row. Das,  J.P., Kirby, J.K., & Jarman, R.F. (1979). Simultaneous cognitive processes. New York: Academic Press.  and successive  Delaney, E.A. & Hopkins, T.F. (1987). Examiner's handbook: An expanded guide for fourth edition users. Chicago: The Riverside Publishing Co.  70  71 Evans,  L. (1980). WISC Performance Scale and Coloured Progressive Matrices with deaf children. British Journal of Educational Psychology, 50, 216-222.  Federal Register. (August 23, 1977). Public Law 94-142 (p. 42494). Furth,  H.  (1966). Thinking without language: deafness. New York: Free Press.  Psychological  implications  Galan,  A.R.N., & Nelson, L.R. (1986). LERTAP 3 technical reference North Vancouver, B.C.: Ditactics Services.  of  manual.  Gannon, J.R. (1981). Deaf heritage: A narrative history of deaf America. Silver Spring, MD: National Association of the Deaf. Gardener,  E . , Rudman, H., Karlsen, B., & Merwin, J . (1982). Stanford Achievement Test (7th ed.). San Antonio, TX: Psychological Corp.  Gerweck,  S., & Ysseldyke, J.F. (1975). Limitations of current psychological practices for the intellectual assessment of the hearing impaired: A response to the Levine study. The Volta Review, 77(4), 243-248.  Glass,  G.V., & Hopkins, K.D. (1984). Statistical methods in psychology (2nd ed.). New Jersey: Prentice-Hall, Inc.  education and  Goetzinger, M.R., & Houchins, R.R. (1969). The 1947 colored Raven's Progressive Matrices with deaf and hearing subjects. American Annals of the Deaf, 11(4), 95-101. Goodenough, F . L . , & Harris, D.B. (1963). Goodenough-Harris Drawing Test. New York: The Psychological Corporation. Graham,  Hirshoren,  E . E . , & Shapiro, E . (1953). Use of the Performance Scale of the Wechsler Intelligence Scale for Children with the deaf child. Journal of Consulting Psychology, 17(5), 396-398. A., Hurley, O.L. & Hunt, J.T. (1977). The WISC-R and the Hiskey-Nebraska test with deaf children. American Annals of the Deaf, 122(4), 392-394.  Hirshoren, A., Kavale, K., Hurley, O.L. & Hunt, J.T. (1977). The reliability of the WISC-R Performance Scale with deaf children. Psychology in the Schools, 14(4), 412-415. Hirshoren, A., Hurley, O.L., & Kavale, K. (1979). Psychometric characteristics of the WISC-R Performance Scale with deaf children. Journal of Speech and Hearing Disorders, 44(1), 73-79. Hiskey, M.S. (1966). Hiskey-Nebraska Test of Learning Aptitude. Lincoln, NB: College View Printers.  72 James,  P.R. (1984). A correlational analysis between the Raven's Matrices and WISC-R Performance Scales. The Volta Review, 86(7), 336-341.  Kaufman, A., & Kaufman, N. (1983). Kaufman Assessment Battery for Children. Circle Pines, MN: American Guidance Service. King, C M . & Quigley, S.P. (1985). Reading and deafness. San Diego: College-Hill Press. Kretschmer, R.R. & intervention Press. Kyle,  Kretschmer, L.W. (1978). Language developement and with the hearing impaired. Baltimore: University Park  J.G. (1980). Measuring the intelligence of deaf children. Bulletin of the British Psychological Society, 33, 54-57.  Lane, H. (1938). Measurement of the mental and educational ability of the deaf child. Journal of Exceptional Children, _4, 169-173. Leiter, R.G. (1948). Leiter International Performance Scale. Chicago: Stoteling Co. Levine,  E.S. (1981). The ecology of early deafness: Guides to fashioning environments and psychological assessments. New York: Columbia University Press.  Levine, E.S. (1974). Psychological tests and practices with the deaf: A survey of the state of the art. The Volta Review, 76, 298-319. Levine,  E.S. (1971). Mental assessment of the 73(2), 80-105.  deaf child. The Volta  Review,  Livingston, S. (1983). Levels of development in the language of deaf children: ASL grammatical processes, Signed English structures, semantic features. Sign Language Studies, 40, 193-286. Luria, A.R. (1961). The role of speech in the regulation of normal and abnormal behavior. New York: Liveright. McQuaid, M.F. & Alovisetti, M . (1981). School psychological services for hearing impaired children in the New York and New England area. American Annals of the Deaf, 126(1), 37-42. Moores,  D., Weiss, K., & Goodwin, M . (1976). Recommended policies and procedures: Preschool programs for hearing-impaired children. Minneapolis: University of Minnesota Research, Development and Demonstration Center in Education of Handicapped Children, Research Report No. 104.  73 Murphy, L . J . (1957). Tests of abilities and attainments. Pupils in schools for the deaf aged six to ten. In A.W.G. Ewing (Ed.), Educational guidance and the deaf child (pp. 213-251). Manchester, England: Manchester University Press. Myklebust,  H.R. (1960). Stratton.  The  psychology  of  deafness.  New  York:  Grune  &  Myklebust, H.R. (1966). The psychology of deafness, (2nd ed.). New York: Grune & Stratton. Neuhaus,  M. (1967). Modifications in the administration of the WISC-R Performance Subtests for children with profound hearing losses. Exceptional Children, 33, 573-574.  Phelps, L. & Ensor, A. (1986). Concurrent validity of the WISC-R using deaf norms and the Hiskey-Nebraska. Psychology in the Schools, 23(2), 138-141. Pintner, R., Eisenson, J . , & Stanton, M. (1941). The psychology of the physically handicapped. New York: Crofts. Porter,  Quigley,  L . J . & Kirby, E.A. (1986). Effects of two instructional sets on the validity of the Kaufman Assessment Battery for Children - nonverbal scale with a group of severely hearing impaired children. Psychology in the Schools, 23(1), 37-43. S.P. (1969). The influence of fingerspelling on the development of language, communication, and educational achievement in deaf children. Urbana, IL: Institute for Research in Exceptional Children.  Quigley, S.P. & Paul, P.V. (1984). Language and deafness. San Diego, California: College-Hill Press. Raven,  J.C. (1960). Company.  Standard  Progressive  Matrices.  London:  H.K. Lewis  &  Raven, J.C. (1965). The Coloured Progressive Matrices Test. London: H.K. Lewis & Company. Ray,  S.  (1979). An adaptation of the Wechsler Intelligence Scale for ChildrenRevised for the deaf. Natchitoches, LA: Northwestern State University of Louisianna.  Reed, M. (1970). Deaf and partially hearing children, in P. Mittler (Ed.), The psychological assessment of mental and physical handicaps. London: Methuen.  74 Ritter,  D.R. (1976). Intellectual estimates of hearing-impaired children: A comparison of three measures. Psychology in the Schools, 13(4), 397-399.  Rosenstein,  J . (1961). Perception, cognition, Exceptional Children, 27, 276-284.  Salvia,  & Ysseldyke, J . E . (1985). Assessment in education (3rd ed.). Boston: Houghton Mifflin.  J.  and  language  in  deaf  children.  special  and  remedial  Sattler, J.M. (1982). Assessment of children's intelligence and special abilities (2nd ed.). Boston: Allyn and Bacon. Sattler,  J.M. (1988). Assessment of Children (3rd ed.). Sattler.  San Diego: Jerome M.  Snijders, J.T., & Snijders-Oomen, N. (1970). Non-verbal Intelligence Test for Deaf and Hearing Subjects (4th ed.). Groningen, Holland: Wolters-Noordhoff. Spragins, A., & Gibbins, S. (1982, March). Interaction effect: School psychology and deafness in training and practice. Paper presented at the meeting of the National Association of School Psychologists, Toronto, Ontario. SPSS Inc. (1988). SPSS-X user's guide. Chicago: SPSS Inc. Steiger, J.H. (1979). MULTICORR: a computer program for fast, accurate, smallsample testing of correlational pattern hypotheses. Educational and Psychological Measurement, 39, 677-680. Stokoe, W.C. (1970). Sign language diglossia. Studies in Linguistics, 21, 27-41. Sullivan,  P.M. (1982). Administration modifications on the WISC-R Performance Scale with different categories of deaf children. American Annals of the Deaf, 127(6), 780-788.  Sullivan, P.M. & Vernon, M. (1979). Psychological assessment of hearing impaired children. School Psychology Digest, _8_(3), 271-290. Terman, L., & Merrill, M. (1973). Stanford-Binet Intelligence Scale, edition. Boston: Houghton Mifflin.  1972  norms  Thorndike, R.L., Hagen, E.P., & Sattler, J.M. (1986a).Guide for administering and scoring the SB:FE. Chicago: The Riverside Publishing Co. Thorndike, R.L., Hagen, E.P., & Sattler, J.M. (1986b). Stanford-Binet Intelligence Scale: Fourth Edition. Chicago: The Riverside Publishing Company. Tweney,  R.D., & Hoemann, H.W. (1976). Translation and sign language. In R.W. Brislen (Ed.), Translation: Applications and research. New York: Gardner Press, Inc.  75 Ulissi, S.M. & Gibbons, S. (1984). Use of the Leiter International Performance Scale and the Wechsler Intelligence Scale for Children - Revised with hearing impaired children. Diagnostique, _9_, 142-153. Vernon, M. (1976). Psychological evaluation of hearing impaired children. In L. Lk>3'd (Ed.), Communication assessment and intervention, (pp. 195-224). Baltimore: University Park Press. Vernon,  M. (1968). Fifty years of research on the intelligence of deaf and hard-of-hearing children: A review of literature and discussion of implications. Journal of Rehabilitation of the Deaf, _1_(4), 1-12.  Vernon, M. (1967). Relationship of language to the thinking process. Archives of General Psychiatry, 16, 325-333. Vernon, M . , & Brown, D.W. (1964). A guide to psychological tests and testing procedures in the evaluation of deaf and hard-of-hearing children. Journal of Speech and Hearing Disorders, 29(4), 414-423. Vygotsky, L.S. (1962). Thought and language. Cambridge, MA: MIT Press. Watson,  B.U. & Goldgar, D.E. (1985). A note on the use of the Hiskey-Nebraska Test of Learning Aptitude with deaf children. Language, Speech, and Hearing Services in Schools, 16, 53-57.  Watson, B., Goldgar, D.E., Kroese, J . , & Lotz, W. (1986). Nonverbal intelligence and academic achievement in the hearing impaired. The Volta Review, 88(3), 151-158. Watson,  B., Sullivan, P.M., Moeller, M.P., & Jensen, J.K. (1982). Nonverbal intelligence and English language ability in deaf children. Journal of Speech and Hearing Disorders, 47(2), 199-204.  Wechsler, D. (1949). Wechsler Intelligence Psychological Corporation. Wechsler, D. (1974). Wechsler Intelligence The Psychological Corporation.  Scale  for Children.  New  York: The  Scale for Children-Revised. New York:  Woodward, J . (1973). Some characteristics of Pidgin Sign English. Sign Language Studies, 3, 39-46. Zimmerman, I.L. & Woo-Sam, J . (1972). Research with the Wechsler Intelligence Scale for Children: 1960-1970. Psychology in the Schools, 9, 232-271.  A P P E N D I X  I T E M  A N A L Y S I S  C H A R A C T E R I S T I C S  O F  T H E  76  A  O F  S B : F E  S E L E C T E D  S U B T E S T S  77 Table A l Item Analysis Characteristics of SB:FE Subtest: Pattern Analysis Item  % Excluded  diff  discr.  13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42  0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.6 2.6 2.6 2.6 5.3 5.3 15.8 15.8 28.9 28.9 39.5 39.5 47.5 47.5 50.0 50.0 57.9 57.9 68.4 68.4  97.4 0.0 97.4 0.0 97.4 0.0 0.0 94.7 97.4 92.1 97.3 94.6 70.3 67.6 72.2 83.3 68.8 75.0 77.8 48.1 69.6 78.3 95.0 75.0 52.7 68.4 66.7 50.0 66.7 75.0  .13 .00 .31 .00 .13 .00 .00 .46 .33 .49 .37 .41 .54 .60 .51 .58 .69 .81 .76 .58 .73 .85 .86 .75 .72 .74 .77 .58 .66 .65  78 Table A2 Item Anatysis Characteristics of SB:FE Subtest: Copying Item  % Excluded  diff  discr.  14 15 16 17 18 19 20 21 22 23 24 25 26 27 28  0.0 0.0 0.0 0.0 0.0 7.9 7.9 21.1 21.1 26.3 26.3 52.6 52.6 68.4 68.4  94.8 65.8 92.1 55.3 68.4 71.4 60.0 53.3 66.7 39.3 42.9 38.9 55.6 33.3 41.7  .09 .43 .40 .57 .37 .51 .50 .44 .82 .59 .71 .54 .72 .52 .54  79 Table A3 Item Analysis Characteristics of SB:FE Subtest: Matrices Item  % Excluded  diff  discr.  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26  0.0 0.0 0.0 0.0 15.8 15.8 18.4 18.4 28.9 28.9 47.4 47.4 63.2 63.2 65.8 65.8 71.1 71.1 73.7 73.7 84.2 84.2 86.8 86.8 89.5 89.5  100.0 78.9 78.9 68.4 71.9 93.8 67.8 54.8 48.1 55.6 50.0 70.0 78.6 42.9 76.9 61.5 54.5 63.7 50.0 40.0 66.7 0.0 20.0 20.0 0.0 0.0  .00 .63 .62 .51 .66 .55 .78 .58 .81 .72 .68 .83 .84 .59 .81 .67 .71 .66 .66 .59 .'63 .00 .37 .26 .00 .00  80 Table A4 Item Analysis Characteristics  of SB:FE Subtest: Paper Folding & Cutting  Item  % Excluded  diff  discr.  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18  2.6 2.6 2.6 2.6 47.4 47.4 78.9 78.9 78.9 78.9 84.2 84.2 94.7 94.7 94.7 94.7 94.7 94.7  73.0 37.8 21.6 32.4 35.0 30.0 87.5 37.5 2.5 37.5 33.3 50.0 100.0 50.0 50.0 100.0 100.0 50.0  .42 .56 .82 .56 .86 .84 .87 .35 .15 .76 .75 .65 .75 .61 .61 .75 .75 .61  81 Table A5 Item Analysis Characteristics of SB:FE Subtest: Number Series Item  % Excluded  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26  0.0 0.0 0.0 0.0 10.5 10.5 15.8 15.8 15.8 15.8 31.6 31.6 55.3 55.3 60.5 60.5 63.5 65.8 86.8 86.8 97.4 97.4 97.4 97.4 97.4 97.4  diff 86.8 92.1 71.1 81.6 76.5 76.5 96.9 46.9 53.1 71.9 46.2 42.3 94.1 41.2 66.7 60.0 0.0 7.7 20.0 60.0 0.0 0.0 0.0 0.0 0.0 0.0  discr. .60 .54 .53 .63 .54 .60 .72 .69 .67 .78 .73 .68 .81 .60 .68 .67 .00 .28 .28 .46 .00 .00 .00 .00 .00 .00  82 Table A6 Item Analysis Characteristics  Item  % Excluded  1 2 3 4 5 6 7 8 9 10 11 12  18.4 18.4 18.4 18.4 84.2 84.2 86.8 86.8 86.8 86.8 92.1 92.1  of SB:FE Subtest: Equation Building  diff 0.0 51.6 6.5 22.6 66.7 83.3 80.0 60.0 0.0 60.0 0.0 0.0  discr. .00 .62 .76 .83 .91 .93 .87 .86 .00 .81 .00 .00  83 Table A 7 Item Analysis Characteristics of SE:FE Subtest: Bead Memory Item 1 2 3 4 5 6 7 8 ' 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40  % Excluded  diff  discr.  0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.6 2.6 2.6 2.6 2.6 2.6 2.6 2.6 7.9 7.9 23.7 23.7 44.7 44.7 57.9 57.9 68.4 68.4 78.9 78.9 84.2 84.2 92.1 92.1 92.1 92.1 94.7 94.7  100.0 97.4 100.0 100.0 100.0 100.0 97.4 100.0 94.7 97.3 89.5 73.7 86.5 86.5 86.5 86.5 81.1 83.8 78.4 59.5 54.3 59.5 31.0 56.6 71.4 38.1 31.3 56.3 66.7 41.7 87.5 25.0 33.3 16.7 66.7 100.0 33.3 0.0 0.0 0.0  .00 .48 .00 .00 .00 .00 .15 .00 .41 .48 .42 .28 .06 .24 .40 .33 .22 .38 .47 .55 .41 .49 .69 .58 .76 .52 .45 .59 .74 .68 .72 .52 .52 .31 .47 .62 .31 .00 .00 .00  84 Table A8 Item Analysis Characteristics of SB:FE Subtest: Memory for Objects Item  % Excluded  diff  discr.  1 2 3 4 5 6 7 8 9 10 11 12 13 14  0.0 0.0 0.0 0.0 0.0 0.0 7.9 7.9 34.2 34.2 86.8 86.8 92.1 92.1  100.0 100.0 100.0 76.3 73.7 68.4 22.9 20.0 24.0 16.0 60.0 40.0 0.0 0.0  .00 .00 .00 .59 .60 .60 .56 .63 .59 .57 .65 .60 .00 .00  APPENDIX COMPARISONS  B  OF SB:FE TEST SCORES USING A L T E R N A T E SCORING  85  86 Table B l Comparisons of SB:FE Scores Using Alternate Scoring* Student  Age  Sex  WISC-R PS  SB:FE (1)  diff  SB:FE (2)  diff  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38  11-1 11-8 13-1 10-8 13-6 14-6 11-10 16-7 9-3 9-3 16-8 16-8 9-1 12-8 16-7 15-0 16-9 12-8 14-7 12-5 14-4' 11-6 15-6 8-3 14-6 15-6 14-7 15-6 14-0 8-3 16-7 9-6 12-4 16-2 15-9 9-7 10-9 13-6  b b g b b b b g b b b g g b g g g g g b b b g b b g b b b b g b g g b g b b  112 88 77 104 106 112 91 90 133 138 101 46 96 92 80 106 85 112 112 146 100 81 73 129 78 118 105 128 80 98 87 95 120 108 96 87 95 90  94 74 78 101 75 94 83 78 123 122 79 54 91 75 68 87 71 115 111 124 69 77 74 104 69 87 88 82 64 93 71 67 92 89 72 91 86 64  -18 -14 1 -3 -31 -18 -8 -12 -10 -17 -22 8 -5 -17 -12 -19 -14 3 -1 -22 -31 -4 1 -25 -9 -31 -17 -46 -16 -5 -16 -28 -28 -19 -24 4 -9 -26  95 72 85 102 72 95 82 81 123 123 81 56 91 78 68 88 70 119 114 122 65 70 69 103 69 92 96 83 62 85 74 63 97 90 69 88 83 64  -17 -16 8 -2 -33 -17 -9 -9 -10 -16 -20 10 -5 -14 -12 -18 -15 7 2 -24 -35 -11 -4 -26 -9 -26 -9 -45 -18 -13 -13 -32 -23 -18 -27 1 -12 -26  * Alternate scoring as explained in Chapter 3.  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items