UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Relationships between teacher ratings and the Gordon diagnostic system in the early identification of… Bergant, Lydia Bernadette 1990

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

Item Metadata

Download

Media
831-UBC_1991_A8 B47.pdf [ 8.17MB ]
Metadata
JSON: 831-1.0098724.json
JSON-LD: 831-1.0098724-ld.json
RDF/XML (Pretty): 831-1.0098724-rdf.xml
RDF/JSON: 831-1.0098724-rdf.json
Turtle: 831-1.0098724-turtle.txt
N-Triples: 831-1.0098724-rdf-ntriples.txt
Original Record: 831-1.0098724-source.json
Full Text
831-1.0098724-fulltext.txt
Citation
831-1.0098724.ris

Full Text

RELATIONSHIPS BETWEEN TEACHER RATINGS AND THE GORDON DIAGNOSTIC SYSTEM IN THE EARLY IDENTIFICATION OF ACADEMICALLY AT-RISK KINDERGARTEN CHILDREN by LYDIA BERNADETTE BERGANT B.Sc. (Biol.), The University of British Columbia, 1979 B.Sc. (Psych.), University of Calgary, 1981 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES Educational Psychology and Special Education We accept this thesis as conforming, to the required standard THE UNIVERSITY OF BRITISH COLUMBIA 16 December 1990 c LYDIA BERNADETTE BERGANT, 1990 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. The University of British Columbia Vancouver, Canada Department DE-6 (2/88) ABSTRACT This exploratory study investigated the similarities and differences between two assessment measures — the Kindergarten School Learning Profile teacher ratings and the Gordon Diagnostic System — in identifying children who would likely be at-risk for experiencing school failure as a result of attentional/impulse control deficits displayed in kindergarten. As attentional skills are believed to influence memory, visual memory was also investigated in relation to attention and impulse control. Twenty-eight teacher-nominated "high risk" kindergarten students were identified as functioning within the lowest 10% for overall school readiness. Computerized systematic random selection procedures were used to identify 30 control students. Teacher ratings of attentional and impulse control abilities manifested both within and outside of the classroom were obtained for all children and compared to their vigilant and impulse control performances on the Gordon Diagnostic System (GDS). Visual memory abilities were examined through use of the Bead Memory subtest of the Stanford-Binet Intelligence Scale: Fourth Edition. The results obtained reveal that normally achieving students were assigned higher qualitative ratings of attention and impulse control by their teachers than were the "high risk" students. Significant relationships between impulsivity (as measured by the GDS) and teacher ratings were unsubstantiated by the data obtained. Only the "high risk" group displayed few significant correlations between teachers' ratings of attentional skills and students' vigilant performances on the GDS. Normally achieving students were found to display significantly better ii vigilant and impulse control skills on the GDS compared to the poorly achieving "high risk" group. Significant performance deterioration over time was evident on the Vigilance Task but not on the Delay Task of the GDS. Few significant differences between boys and girls in both impulse control and sustained attentions] skills were displayed. iii TABLE OF CONTENTS ABSTRACT ii LIST OF TABLES vi LIST OF FIGURES ...„.....„...;„H.......... vii ACKNOWLEDGEMENTS .. ix I. INTRODUCTION 1 A. Research Problem . 5 B. Purpose of the Study 7 C. Definition of Terms 7 1. Attention Deficit Disorder 7 2. Impulsivity 8 3. Vigilance . 9 4. Continuous Performance Task (CPT) 10 5. Delayed Reaction Time Test (DRTT) .. 10 6. Differential Reinforcement of Low Rate Responding (DEL) ... 11 7. At-Risk, "High Risk", and Normally Achieving 12 D. Research Questions 13 II. LITERATURE REVIEW 14 A. Aspects of Attention 14 1. Sustained Attention 15 2. Impulsivity 19 3. Gender Differences 22 B. Early Identification of Attentional Dysfunction 24 1. Early Risk Factors and Initial Manifestations 27 2. Age at Onset 28 3. Classroom Predictors of Later Academic Adjustment 29 C. Assessment of Attention 34 1. Rating Scales 34 2. Computerized CPT 35 3. Gordon Diagnostic System (GDS) 37 D. Memory and Attentional Deficits 38 E. Hypotheses 41 HI. METHOD 43 A. Subjects :.?.r.*.?ici7J^£i.\ 43 1. Selection Procedure 43 2. Sample Composition 44 B. Instrumentation 45 1. Ratings 45 2. Standardized Measures 46 C. Procedure 49 D. Limitations 50 E. Data Analysis 52 i v IV. RESULTS , 53 A. Descriptive Statistics 53 B. Hypotheses 55 1. Hypothesis 1 55 2. Hypothesis 2 55 3. Hypothesis 3 62 4. Hypothesis 4 62 5. Hypothesis 5 . 69 6. Hypothesis 6 70 7. Hypothesis 7 79 8. Hypothesis 8 82 9. Hypothesis 9 . 90 10. Hypothesis 10 91 C. Summary . 92 V. DISCUSSION AND CONCLUSIONS 95 A. Summary 96 B. Discussion of Results 97 1. Teacher Ratings of School Readiness and Attention 97 2. The Gordon Diagnostic System 104 3. Bead Memory - SBFE 118 C. Conclusions 120 1. Hypothesis 1 120 2. Hypothesis 2 120 3. Hypothesis 3 122 4. Hypothesis 4 122 5. Hypothesis 5 123 6. Hypothesis 6 124 7. Hypothesis 7 124 8. Hypothesis 8 125 9. Hypothesis 9 126 10. Hypothesis 10 127 D. Limitations 127 E. Implications and Future Research 129 REFERENCES 135 APPENDIX A: LETTER TO TEACHERS INVITING THEIR PARTICIPATION IN THE STUDY 143 APPENDIX B: KINDERGARTEN SCHOOL LEARNING PROFILE (TEACHER RATING SCALE) 145 APPENDIX C: PARENT PERMISSION LETTER 149 APPENDIX D: THE GORDON DIAGNOSTIC SYSTEM (MODD7TED RECORDING SHEETS) 153 v APPENDIX E: INTERCORRELATIONS FOR CONTROL AND HIGH RISK STUDENTS ON GORDON DIAGNOSTIC SYSTEM TASKS; MEANS AND STANDARD DEVIATIONS OF GENDER PERFORMANCE DIFFERENCES ON THE SLP, GDS AND BEAD MEMORY 156 vi LIST OF TABLES Table 1: School Learning Profile: Sample Frequency Distribution 54 Table 2: Means (Standard Deviations) by Group on the School Learning Profile with Analysis of Variance for Group Effect 56 Table 3: Means (Standard Deviations) by Group on the Delay Task of the Gordon Diagnostic System and Bead Memory (SBFE) 57 Table 4: Means (Standard Deviations) by Group on the Vigilance Task of the Gordon Diagnostic System 58 Table 5: Intercorrelations of School Learning Profile (SLP) Ratings and GDS Performance for the Total Sample 60 Table 6: Intercorrelations of SLP 2 (Attention Span/Distractibility) and SLP 13 (Behaviour) Ratings for the Sample 64 Table 7: Repeated Measures Analysis of Variance (ANOVAR) Summary Table for GDS Delay Task Scores 71 Table 8: Repeated Measures Analysis of Variance (ANOVAR) Summary Table for GDS Vigilance Task Scores 74 Table 9: Intercorrelations of Sustained Attention and Impulsivity Among Groups 79 vii LIST OF FIGURES Figure 1: Intercorrelations of the SLP 2 (Attention Span/Distractibility) and SLP 13 (Behaviour) Ratings for the Sample 65 Figure 2: Mean Rating Values of SLP 13 (Behaviour) for the "High Risk" and Control Students 68 Figure 3: Mean GDS Delay Task Performance Over Time for "High Risk" and Control Students 72 Figure 4: Mean GDS Vigilance Task Performance Over Time of "High Risk" and Control Students 76 viii ACKNOWLEDGEMENTS Sincere appreciation is expressed to my advisor, Dr. David Kendall, for his academic guidance and support throughout my graduate program. I would like to thank my additional committee members, Dr. Marion Porath and Dr. Julianne Conry, for their constructive comments and direction in the preparation of this thesis. Dr. Nand Kishor's invaluable assistance with the statistical data analyses was also most appreciated. I would like to thank my colleague, John Carter, for allowing me to participate in his doctoral research study, from which interest in the present study was generated. Gratitude is also extended to the kindergarten teachers and students of School District No. 71 (Courtenay) without whose cooperation this study would not have been possible. A special note of recognition and thanks is expressed to my family, particularly to my mother, whose encouragement and unconditional support during my studies will always be cherished. ix I. I N T R O D U C T I O N Attention deficits in school children are gaining increasing recognition as common and devastating contributors to academic underachievement and behavioural maladaptation. A child's inability to orient, focus and organize his/her attention on a specific task, to sustain an effectual concentration span and to keep impulsive responding under control can cause serious delays in acquiring the readiness skills necessary for learning in school. Research suggests that in a large number of children, precursors of attentional problems are present and identifiable during the preschool years (Palfrey, Levine, Walker & Sullivan, 1985; Simner, 1983, 1987) and are important in predicting kindergarten outcomes (Chamberlin, 1977). Children experiencing attentional difficulties, coupled with hyperactivity, have been estimated to be at two to three times greater risk for school failure than normal children, with many being retained at least one grade before entry into the intermediate or junior high school years (Barkley, 1981). Thus, early identification of these attentional difficulties may assist educators at the preschool/kindergarten level in understanding a child's learning style so that curriculum materials may be customized to accommodate his/her distractibility, poor self-monitoring, and impulsive tendencies, thus reducing the risk of subsequent grade retention (McKinney, 1989). Accurate identification of these failure-prone, or "at-risk", children is a complex process which has generated considerable concern. In recent years, traditional approaches to readiness assessment have been increasingly challenged (Leigh and Riley, 1982), mainly due to their questionable reliability and validity. Traditionally, a combination of standardized psychometric tests and 1 INTRODUCTION / 2 behavioural/classroom observations have been almost exclusively used in diagnosing attentional deficit disorders (Barkley, 1981). The DSM III-R, a classification system of behavioural disorders stemming from both organic and/or non-organic origins, identifies children with deficits in attention (particularly sustained attention), differentiating those children exhibiting additional difficulties with hyperactivity and impulsivity (attentional-deficit hyperactivity disorder — ADHD) from those exhibiting normal activity levels and impulse control abilities (undifferentiated attentional deficit disorder — ADD). As the literature does not appear to adopt any consistent use of these terms, the author has chosen to use these terms interchangeably, (i.e., ADHD and hyperactive; ADD and nonhyperactive). In reporting specific research studies, the terminology used by their authors was adopted. Many ADHD/ADD children are felt to have their most significant problems in sustaining attention to task-relevant stimuli while inhibiting response to stimuli irrelevant to the task, (i.e., controlling impulses), (Douglas, 1972; Richman, 1987). It is these two aspects of attention that have been explored in the research and have been found to be most deficient in children experiencing attentional problems, especially in hyperactive children. Tasks commonly used to assess sustained attention are vigilance or Continuous Performance Tasks (CPT) where the subject is asked to detect relatively rare signals against a repetitive background of "noise". This component of attention has been related to academic difficulties. Keogh and Margolis (1976) found that educationally handicapped boys (in special class placements for remediation of learning problems) made significantly more errors than did their normally achieving peers on a CPT requiring the maintenance of sustained attention. These authors contend that ~" INTRODUCTION / 3 . . . sustained attention . . . may be interpreted in a number of ways, but importantly, is associated significantly with school success" (Keogh & Margolis, 1976, p. 283). Moderate correlations between CPTs and other measures, such as, direct observations of "off-task" classroom behaviour (Kupietz & Richardson, 1978; Richman, 1987) and teacher ratings of attentional difficulties (Charles, Schain, Zelniker, & Guthrie, 1979) have been documented. Delayed Reaction Time Tests (DRTT), based on a Differential Reinforcement of Low Rates (DRL) schedule together with CPTs, have long been used to assess impulse control, requiring the subject to inhibit responding in order to gain a reward. Teacher rating scales, in general, have continued to enjoy widespread use in screening children "at-risk" for experiencing later learning problems, despite heavy criticism regarding rater bias, the possibility of "halo" effects, inaccurate teacher judgements (Fletcher & Satz, 1982; Stevenson, Parker, Wilkinson, Hegion, & Fish, 1976; Whalen, 1983) and the relatively poor differentiation of the quantitative ratings involved (Ross & Ross, 1982). Some researchers believe that teacher rating scales are equally, if not more, effective than many standardized psychometric instruments in identifying individual academically "at-risk" children, and thus, have strongly recommended that they form part of the screening process (Barkley, 1981; Day & Peters, 1989; Gresham, Reschly, & Carey, 1987; Lindsay & Wedell, 1982; Simner, 1986). All in all, laboratory measures tend to be more objective than ratings, more sensitive than behavioural observations and perhaps most comparable from study to study. Based on modifications of the CPT and the DRT, Gordon (1986) developed a laboratory measure designed to assess sustained attention, impulsivity and distractibility. The Gordon Diagnostic System (GDS) is a microprocessor-based INTRODUCTION / 4 portable instrument claiming to accurately discriminate between hyperactive and nonhyperactive children selected from an outpatient clinic population (Gordon, 1979; Gordon & McClure, 1983; McClure & Gordon, 1984). Despite certain advantages, (i.e., portability, ease of administration, and ability to store multiple data points (Gordon, 1986c)), the GDS has been criticized for its questionable reliability and validity together with its insensitivity to low-dose stimulant drug effects (Barkley, Fischer, Newby, & Breen, 1988). The link between observable classroom behaviours and more "precise" objective measures of attentional and impulse control deficits has important clinical implications, (i.e., questionable validity of results collected in this manner). Do results from well-defined laboratory measures truly represent a child's ability to sustain attention and exercise self-control strategies within his/her natural classroom environment? Could the more favourable laboratory environment, for example, the one-to-one situation with the examiner, high-interest level of task, fewer distractions from environment and peers, actually mask those students who are truly "at-risk" for attentional/impulsivity problems, (i.e., inflated false negative results)? In order to obtain results that could be validly interpreted, the laboratory testing situation would need to be constructed in such a way as to precisely mimic the child's school learning situation. The likelihood of being able to achieve this is remote. Most learning tasks require children to perceive and attend to visual and auditory stimuli (Douglas, 1972; Richman, 1987). Hyperactive children have been found to perform poorer than nonhyperactives on both visual and auditory vigilance tasks (Sykes, Douglas, & Morgenstern, 1973). Amongst the hyperactive children in this study, greater performance efficiency was displayed on visual INTRODUCTION / 5 rather than on auditory vigilance tasks (Sykes et al., 1973). Extensive research in learning disabilities has demonstrated that learning disabled children experience memory deficits for both visual and auditory stimuli. Torgeson and Hall (1980) have suggested that learning disabled children are deficient in their use of learning strategies and may have poor language skills, resulting in a general memory deficiency. Three basic processes are involved in most memory tasks: encoding, storage and retrieval (Torgesen, 1981). In researching the relationship between attention and memory, attention is believed to be most critical during the encoding phase of the performance, (i.e., placing information into memory). Although there appears to be a paucity of research on memory and ADD (with or without hyperactivity), some research has suggested that poor, memory performance reflects inefficiencies in controlled information processing (Torgesen, 1981; van der Meere & Sergeant, 1988). Few studies have examined the relationship between vigilance/impulsivity and visual memory. This study will attempt to investigate whether significant relationships do exist in this area. A. RESEARCH PROBLEM It has been found that children with ADHD/Undifferentiated ADD experience rates of learning difficulties well above those in the general population (Richards, Samuels, Turnure, & Ysseldyke, 1990). Amongst the ADHD group, estimates of the proportion of children likely to experience learning problems in school range from 25% to 60% (Barkley, 1981). Children identified as experiencing difficulties with attention, impulse control and/or overactivity in the preschool/kindergarten years frequently continue to exhibit such problems when INTRODUCTION / 6 they reach the primary grades in elementary school (Campbell, Endman, & Bernfeld, 1977; Chamberlin, 1977; Fletcher & Satz, 1984), with these problems persisting into their later school years. Since adequate attention is a basic prerequisite for effective learning, and thus, for academic achievement, the need for establishing an effective means of identifying children with ADHD/Undifferentiated ADD early is essential. Classroom observations and teacher ratings have been strongly recommended (Gresham et al., 1987) for use in this early identification process. Computerized measures of attention, such as the GDS, are becoming increasingly popular for research. Such measures appear to be promising tools in facilitating the early identification process, providing useful information on specific types of intervention required as well as on the efficacy of these methods of intervention. The lack of sound standardization data of these methods has prevented them from becoming more popular in educational settings. If teacher ratings are indeed as effective in identifying children who may later encounter academic difficulties as the literature suggests (Barkley, 1981; Gresham et al., 1987), and if such computerized measures were found to be equally as effective, then use of instruments such as the GDS in the kindergarten assessment/screening process may be more desirable as this latter measure obtains objective data on sustained attention and impulse control abilities, compared with the more subjective information attained through teacher ratings. INTRODUCTION / 7 B. PURPOSE OF THE STUDY This is an exploratory study which compares teacher ratings of school readiness and classroom behaviour of a kindergarten population with sustained visual attention and impulse control performances on computerized CPT and DRT tasks (Gordon Diagnostic System, 1986a). The relationship between visual memory and sustained visual attention and impulse control performances will also be explored. C. DEFINITION OF TERMS Terms commonly associated with the research topic of this study are defined below. 1. Attention Deficit Disorder Attention Deficit Disorder is a clinical diagnostic term/label assigned by the psychiatric profession in the identification of individuals exhibiting attentional deficits, poor impulse control and overactivity. Although the identification of these behavioural symptoms and their categorization into diagnostic terms is not new, the American Psychiatric Association recognized the necessity for the inclusion of a category for these behavioural characteristics of attentional problems in the 1980 revision of the Diagnostic and Statistical Manual of Mental Disorders — DSM HI. The DSM III differentiated between two aspects of the Disorder based on the presence or absence of displayed motoric overactivity, (i.e., Attention Deficit Disorder with or without hyperactivity, ADDH and ADD, respectively). Since the publication of the DSM III, a recent (1987) further revision (DSM HI-R) has been found necessary in re-classifying the Attention Deficit INTRODUCTION / 8 Disorder based on the presence or absence of not only an exhibited hyperactive state, but also on the presence or absence of difficulties with impulse control. The most recent terminology used, therefore, in the field of Attention Deficit Disorders is as follows: 1. Attention-deficit Hyperactivity Disorder (ADHD), including problems in sustained attention, impulse control, and overactivity/hyperactivity. 2. Undifferentiated Attention-deficit Disorder (ADD), including problems in sustained attention only. Consistent use of the above-mentioned terminology/abbreviations in the literature does not appear to exist such that outdated references to the Disorder, (e.g., hyperactivity, ADDH, ADD [referring to the Disorder in general and not specifically to problems in sustained attention only]), continue to be frequently used interchangeably, potentially leading to some confusion for the reader. For the purposes of this study, the author has chosen to use the current terminology/abbreviations as set forth by the DSM DJ-R in her writing of this thesis. When references are made to specific research studies, however, the author felt it necessary to adopt the same terminology/abbreviations as that of the research authors so as to avoid reporting any distortions in the studies' purposes, identified populations, findings, conclusions, etcetera, as related to attentional deficits. 2. Impulsivity Impulsivity can be defined as an individual's consistent tendency to respond quickly in problem-solving situations where several possible solutions are available simultaneously, and the individual must evaluate the differential INTRODUCTION / 9 adequacy of each possibility. Children who demonstrate poor impulse control respond quickly and tend to make more mistakes on problem-solving tasks than children with good impulse control skills (Cairns & Cammock, 1978). Hyperactive children have been found to exhibit substantially higher rates of impulsive responding errors on Continuous Performance Tasks and Delayed Reaction Time Tests than have nonhyperactive controls (Sykes, Douglas, & Morgenstern, 1973). Measuring a child's cognitive response style is most beneficial in obtaining information on his/her general approach to solving academic problems and in preparing effective remedial programs. 3. Vigilance The vigilance paradigm involves the prolonged monitoring and detection of unpredictable and infrequently presented signals (visual or auditory) that are interspersed among background noise. Tasks requiring such maintenance of attention over time are especially difficult for children with learning problems (Day & Peters, 1989). Children experiencing difficulties maintaining attention are described as readily losing interest in an ongoing task and changing activities frequently, usually before task completion. Although they may be enthusiastic and competent "starters", their attention tends to deteriorate more rapidly than that of their peers. INTRODUCTION / 10 4. Continuous Performance Task (CPT) The Continuous Performance Task (Rosvold, Mirsky, Sarason, Bransome, & Beck, 1956) has been used to examine an individual's ability to remain vigilant over long periods of time, as measured through the effect of time on task. The CPT requires children to respond to a specific target letter/number, (e.g., A/1), or pattern of letters/numbers, (e.g., an X/5 immediately following a(n) A/3), presented within a series of nontarget letters/numbers to which the child must refrain from responding. The CPT provides two kinds of scores: omission errors, reflecting problems of inattention, and commission errors, indicative of problems of impulsivity. The CPT was originally designed to investigate brain-damaged children and adults who consistently exhibited inferior performances on this task. CPTs have frequently been used in the investigation of ADHD or Undifferentiated ADD. Children classified as such display a tendency to make fewer correct responses and have slower mean reaction times for correct responses on the CPT than do normal controls (Sykes, Douglas, & Morgenstern, 1973). The Vigilance Task of the Gordon Diagnostic System, used in this study to examine a child's ability to sustain attention to a task over time, is a version of the CPT. 5. Delayed Reaction Time Test (DRTT) Delayed Reaction Time Tests (Cohen, Douglas, & Morganstern, 1972) examine impulsive response styles, (i.e., an individual's ability to delay responding until a certain time interval has elapsed). DRTTs are based upon a Differential Reinforcement of Low Rate Responding behavioural schedule. Hyperactive children INTRODUCTION / 11 experience profound difficulties dealing successfully with the time delay component in a DRTT (Gordon & McClure, 1983). The Delay Task of the Gordon Diagnostic System is an example of a DRTT. 6. Differential Reinforcement of Low Rate Responding (DRL) Derived from research in operant conditioning (Skinner, 1938), the Differential Reinforcement of Low Rate Responding schedule is a useful means of measuring impulsive response styles. Reinforcement of a response is contingent upon responding after a specified time interval has elapsed. If a response occurs prior to the end of this time interval, the response is not reinforced and serves to reset the timer controlling the time interval. For example, a DRL 6-second schedule provides the subject with a reinforcement for each response emitted following a 6 second interval. Responses emitted prior to the lapse of 6 seconds from the previous response would not be reinforced, and the subject would then have to wait another 6 seconds so as to be reinforced for the next response emitted. Three measures of performance can be obtained from the DRL task: total number of responses, number of earned reinforcements, and a total efficiency score reflecting the percentage of reinforced responses. This latter score, total efficiency score, is considered to be the most sensitive indicator of impulse control performance (Gordon, 1979). The Delay Task of the Gordon Diagnostic System used in this study is based upon such a DRL behaviour schedule. INTRODUCTION / 12 7. At-Risk, "High Risk", and Normally Achieving Definitions of at-risk have usually taken on a broad perspective, subsuming a variety of problems and allowing for a range of severity of problems. Keogh, Wilcoxen, and Bernheimer (1983) provide a definition of at-risk children as those ". . . with a higher than average probability of problems in development. Negative developmental outcomes are . . . viewed as ranging from life threatening or handicapping conditions to school-failure." (p. 1). The need for establishing precisely delineated categories or types of risk is essential in understanding their early identification. Early identification of educationally at-risk children is particularly complex in regard to psychosocial and school-related problems as these two areas are multidimensional and interactive, thus giving rise to criticisms of predictive validity of early detection measures (Lindsay & Wedell, 1982). For the purposes of this study, at-risk refers to those children whose academic difficulties are unrelated to any other compounded risk conditions, (e.g., physical disabilities, visual and/or hearing impairments, emotional disturbances, etc.), with the exclusion of attentional deficits. "High risk", in this study, refers specifically to those at-risk kindergarten students whose teachers consider them to be functioning within the lowest 10% of students for overall ability to learn school materials by the Spring of their kindergarten year. The normally achieving students comprise the control group in this study. These students are considered by their teachers to be academically achieving within the average to above average ranges by the Spring of their kindergarten year. The terms normally achieving and controls have been used interchangeably INTRODUCTION / 13 in this research study. D. RESEARCH QUESTIONS 1. Do vigilance and impulsivity performances of "high risk", determined by teacher nomination of those students functioning within the lowest 10% for overall learning ability, and normally achieving kindergarten students relate to teacher ratings of attention in the school environment both within and outside of the classroom situation? 2. What is the relationship between teacher ratings of attention and their ratings of school readiness in a kindergarten population? 3. Are there significant differences between "high risk" and normally achieving kindergarten students on measures of sustained attention and impulsivity? Do "high risk" and normally achieving students differ in their performances on a vigilance task over time? 4. Is there a relationship between vigilance and impulsivity? That is, is there a significant correlation between these two variables amongst all kindergarten students or just within one group? 5. Are there significantly more inattentive boys than girls? 6. What is the relationship between vigilance and impulsivity performances of "high risk" and normally achieving kindergarten students and their performances on a visual memory task? II. LITERATURE REVIEW The past two decades have produced increasingly more research acknowledging that attentional problems present during the preschool/primary years elevate the risk for poorer academic achievement later, coupled with subsequent social/emotional adjustment problems (Barkley, 1981; Campbell, Endman, & Bernfeld, 1977; Douglas, 1972; McKinney, 1989; Palfrey et al., 1985; Richman, 1987; Ross & Ross, 1982; Whalen, 1983). Without extensively addressing, the entire field of attention and attentional deficits, this chapter attempts to present a critical review of the literature as it relates to the early identification of attentional deficits, teacher rating scales and computerized continuous performance tests (namely, the Gordon Diagnostic System) in the identification of attentional deficits, and the relationship of visual memory and attentional deficits. A. ASPECTS OF ATTENTION Attentional difficulties can occur in many forms. Problems in orientation and detection of stimuli or responding to incorrect/inappropriate stimuli can create serious impairments in a child's learning process. Although a wide variety of significant attentional/behavioural problems can be suggestive of ADD, two fairly consistent and prominent deficits appear to be commonly associated with this disorder in children: (1) difficulty sustaining attention to task-relevant stimuli; (2) difficulty with impulse control, or impulsivity (Barkley, 1981; Douglas, 1972). Douglas (1972) has demonstrated that, as a group, hyperactive children are not significantly different from normals with respect to language abilities, comprehension, conceptual thinking, or short term memory abilities. She does 14 LITERATURE REVIEW / 15 contend that, while hyperactive children are more physically active than normals, most of their behaviour is goal-oriented. Douglas suggests that this demonstrated overactivity is not the most critical aspect of hyperactive children, for as these children grow older excess physical activity diminishes; however, impulsivity and inability to attend remain problematic. This link between attention disturbances and poor impulse control was recognized by the American Psychiatric Association, and led to the categorization of attention deficit disorder with hyperactivity (ADD-H) or without hyperactivity (ADD) (DSM in, 1980). A recent revision of the DSM HI (DSM IH-R, 1987) has recategorized these disorders such that: (1) Attention-deficit Hyperactivity Disorder (ADHD) includes problems in sustained attention, impulse control, and hyperactivity; and (2) Undifferentiated Attention-deficit Disorder (ADD) includes problems in sustained attention only. 1. Sustained Attention Sustained attention is an aspect of attention involving " . . . the ability to remain vigilant over long periods of time (as measured through the effect of time on task) and the ability to prepare and maintain readiness for response (as measured through the effect of warning signals on reaction time)" (Schachar, Logan, Wachsmuth, & Chajczk, 1988, p. 362). Successful learning and academic performance is dependent on a child's ability to sustain attention to a critical stimulus. In a classroom situation, therefore, a child needs to stay focused on his/her reading or math for at least several minutes in order to learn the lesson at hand. Teachers and parents often comment that children who are not doing well LITERATURE REVIEW / 16 in their school performance do not pay attention. Research studies support these observations, suggesting that for many children a failure to sustain attention may be a primary component in their inability to learn. Douglas (1972) has strongly argued that the main disability in learning disabled and hyperactive children is their inability to sustain attention and to control impulsive responding on tasks requiring focused, reflective, organized and self-controlled effort. Most learning tasks, according to Douglas, appear to require adequate perception of and attention to visual and auditory stimuli followed by processing, thinking and deciding about the task, and finally responding verbally or motorically. A wide variety of measures have been used to assess sustained attention. One fairly objective approach, less prone to subjective interpretaion than traditional rating scales, has been provided by tasks known as vigilance tasks. The study of vigilance performance originated in the early 1940s with the investigations of declined detection accuracy of radio operators during prolonged watch periods (Mackworth [cited in Whalen, 1986]). Since then, all vigilance studies have essentially simulated this "watch keeping" task. A well-known and commonly used vigilance task is the Continuous Performance Test (CPT) (Rosvold, Mursky, Sarason, Bransome, & Beck, 1956), designed originally to detect and study brain damage in both children and adults. The task required the subject to respond to a target signal (usually the letter "X" or the letter "X" which follows the letter "A") presented simultaneously with other nontarget signals in a random sequence. Consistently inferior performances on this task were found among the brain-damaged group, as compared with controls. Rosvold et al. (1956) suggested that this inferior performance was due to a decreased alertness within the brain-damaged group. In a later study LITERATURE REVIEW / 17 (Mirsky & Rosvold [cited in Douglas, 1972]), hyperactive children were found to identify fewer correct stimuli and to respond more frequently to incorrect ones, in which both visual and auditory stimuli (the letter "X" when preceded by the letter "A") were presented on a screen or through earphones over a 15-minute interval. Additonally, a more severe deterioration in the performance of the hyperactives over time was observed compared with the performance exhibited by the control group. Many researchers have since recognized the existence of a relationship between poor vigilance performance and attention deficits and have reported using vigilance, or continuous performance, tasks in examining sustained attention abilities in both children and adults (Barkley et al., 1988; Douglas, 1972; Gordon, 1979, 1986; Gordon & McClure, 1983; Klee & Garfinkel, 1983; Kupietz & Richardson, 1978; Murphy-Berman, Rosell, & Wright, 1986; Richman, 1987; Sykes, Douglas, & Morgenstern, 1973). Although it has been argued that vigilance tasks of the CPT type involve monotonous routines which do not reflect typical classroom behaviour or real-life demands, considerable validational support for these assessment procedures has been documented. Sykes, Douglas and Morgenstern (1973) found that comparisons in vigilance performance on an AX version of the CPT for normal and ADD children (age range 5 years 10 months to 11 years 5 months) revealed less accuracy in task performance and a greater deterioration with time on task for the latter group. Michael, Klorman, Salzman, Borgstedt, & Dainer [cited in Schachar et al., 1988] presented contradictory results to those of Sykes et al. (1973). Michael et al. found that hyperactives in their study did not demonstrate a greater deterioration LITERATURE REVIEW / 18 in vigilance performance with time on task than did the controls. Sykes et al. (1973) also suggested that generalized behavioural problems, specifically inattentiveness and distractability, were apparent in other settings, apart from the confines of a structured laboratory setting. Kupietz and Richardson (1978) conducted a comparative study on visual and auditory vigilance performance and teacher ratings of "off-task" classroom behaviour of 7 to 12 year olds attending remedial reading classes. They found that vigilance errors showed a moderately positive correlation with teacher-rated "off-task" behaviours, which provides support for the hypothesis that sustained attention relates to a child's ability to remain attentive and "on-task" in the classroom. Simon (1982) found that preschool performance scores on a visual vigilance task involving the detection of a change in a stationary picture over a 30-minute vigilance period was significantly related to Metropolitan Readiness Scores. Swanson (1980) tested the hypothesis that learning disabled children exhibit an attention deficit related to reading performance. In this study, Swanson used two CPTs — visual and auditory — for approximately 5 and 10 minutes. He found that learning disabled students with reading deficits made significantly fewer correct decisions than did the normal group. Learning disabled children, therefore, clearly appear to exhibit deficits in sustained attention, as evaluated through use of vigilance tasks in numerous investigations. Continuous performance tasks, although very useful in examining sustained attention, have received criticisms. Krupski (1981) argues that, since attention is a construct, it cannot be directly measured. Thus, ascertaining whether or not a LITERATURE REVIEW / 19 child is attentive to a particular task can only be based on inference, which can lead to serious problems of interpretation. Krupski suggests, therefore, that these inferences be made with caution. It has also been asserted that high error rates on CPT measures need not necessarily be indicative of attentional deficits, but instead may result from noncompliance or a lack of subject involvement in the task. Murphy-Berman, Rosell and Wright (1986) contend that the CPT is not useful in providing information regarding developmental differences in attentional skills since it is more suitable for children of certain ages than others. For example, a particular rate of signal presentation may not be equally challenging for both younger and older children (Murphy-Berman et al., 1986). Schachar et al. (1988) recognize that performance differences between hyperactive and normal children on the CPT ". . . are not necessarily attributable to inability to sustain attention" (p. 362). They go on to point out that . . methods used in studies to date do not permit the rejection of alternative explanations for poor performance (e.g., deficits in coordination, vision, understanding, or cooperation) that could affect the overall level of performance on an attention-demanding task" (p. 362). 2. Impulsivity Impairment of the decision-making process involved in problem solving situations has frequently been implicated in learning problems. Problem solving requires sustained attention coupled with a decision-making style that relies upon intact impulse control. Problems with impulse inhibition, therefore, are frequently responsible for hasty, inaccurate responses, particularly noticeable amongst LITERATURE REVIEW / 20 hyperactive children both within and outside of the classroom situation. Research has shown that there is a tendency for more errors to be made when children respond quickly (impulsive responding) than when they respond slowly (reflective responding) (Cairns and Cammock, 1978). According to Douglas (1972), the hyperactive child is characterized by his/her inability to "stop, look and listen." Hyperactive children have been found to exhibit substantially higher rates of impulsive errors on Continuous Performance Tasks (Sykes, Douglas & Morgenstern, 1973) as well as on Delayed Reaction Time Tests (Parry & Douglas [cited in McClure & Gordon, 1984]). Douglas has concluded that hyperactivity is not necessarily typified by the amount of overactive behaviour, but instead by displayed significant deficits in organization and impulse control. Together with her colleagues, Douglas contends that problems with attention and impulsivity occur together as a single attentional-impulsivity deficit interfering with the learning process. Day and Peters (1989) support Douglas' viewpoint: " . . . the inability to organize and sustain attention and inhibit impulsive responding are the qualities that most clearly differentiate between hyperactive and nonhyperactive children" (p. 357). Measures of impulse control have taken many forms. Adaptations of the Bender Gestalt (Koppitz, 1964) and selected subtests of the WISC-R have been used but have been shown to inadequately assess impulsivity as they assess abilities/traits that are predominantly unrelated to impulsivity and do not accurately discriminate between hyperactive and nonhyperactive children (Douglas, 1972). The most commonly used measure of impulsivity has been the Matching Familiar Figures Test (MFFT) which yields two scores — latency to first LITERATURE REVIEW / 21 response and number of errors. Using this measure, it was found that low achievers responded more impulsively than reflectively, with the opposite being true amongst the high achievers. Additionally, low-achieving boys demonstrated significantly greater errors on the MFFT than did the high-achieving boys (Hallahan, Kauffman, & Ball [cited in Day & Peters, 1989]). On a reading comprehension task, Peters and Rath (1983) found that impulsive children read stories more quickly, making greater errors than reflective students. Teacher ratings have also been found to be very useful in evaluating children's impulsive nature within the classroom. The most commonly used scale has been the Conners' Teacher Rating Scale (Conners, 1969). The inattentive-passive factor of this Scale has been found to significantly differentiate impulsive and reflective groups of Grade 3 students, with the former group demonstrating more inattentive-passive classroom behaviour than the latter group (Peters & Rath, [cited in Day & Peters, 1989]). Day and Peters (1989) found this same factor to be the best discriminator between normally achieving and underachieving Grade 3 and 4 students; however, it did not correlate with measures of cognitive impulsivity (MFFT), selective and sustained attention (The Jumbled Numbers Game) or with the reading comprehension task used. While the underachievers were rated as being distractible and inattentive in class when given a specific problem to solve, they were able to maintain their level of attention and focus on relevant aspects of the task while discarding task-irrelevant diversions as competently as their normally achieving peers. Day and Peters concluded that inattentive-passive classroom behaviour ratings correlated strongly with academic underachievement and that teacher ratings of disruptive classroom behaviour (using the Self-Control LITERATURE REVIEW / 22 Rating Scale) indicated that underachievers demonstrate less cognitive-behavioural self-control than normally achieving students. Among other assessment measures available, the CPT and Delayed Reaction Time Test (DRTT) (Cohen, Douglas, & Morgenstern [cited in Gordon & McClure, 1983]) are reportedly most effective in differentiating hyperactive from nonhyperactive children. Sykes, Douglas and Morgenstern (1973) found that, on a CPT, hyperactive children produced significantly more anticipatory responses coupled with a greater total number of responses than did the controls. As Douglas (1972) has argued that impulsivity and attention are related aspects of the same process, error scores obtained from the CPT can provide valuable information on impulse control in addition to vigilance skills. Derived from research in operant conditioning and the DRTT, the Differential Reinforcement of Low Rate Responding (DRL) schedule is a useful means of measuring impulsive styles. The Delay Task of the GDS (Gordon, 1986) is based on such a DRL schedule. An efficiency ratio (ER) can be calculated from trial, or Block, scores, (i.e., total number of correct responses divided by the total number of responses), reflecting the level of impulsivity a child exhibits on the task. This ER score has been found to be a very sensitive measure in differentiating "accurately" between hyperactive and nonhyperactive children (Gordon & McClure, 1983). 3. Gender Differences While minimal performance differences between the sexes on ability, achievement and aptitude measures have been reported (Salvia & Ysseldyke, 1985), there is consistent evidence that attentional deficits are significantly more LITERATURE REVIEW / 23 prevalent in boys than in girls (Vogel, 1990), with a ratio of 5:1 suggested (Ross & Ross, 1982). McKinney (1975) found that teacher ratings on a classroom behaviour inventory indicated that impulsive boys were rated to be more impulsive and distractible than girls. In their longitudinal study on the emergence of attention deficits in early childhood, Palfrey, Levine, Walker and Sullivan (1985) reported that boys exhibited the most persistent attentional problems (i.e., apparent early [at 2 1/2 years of age] and persisted 2 1/2 years later into kindergarten). Palfrey et al. used the following criteria to identify attention deficits in this study: " . . . chronic inattention, distractibility, disorganization, poor self-monitoring (as evidenced by impulsivity), and overactivity" (p. 346) which they consider to justify early suspicion of later attentional problems. Murphy-Berman, Rosell and Wright (1986) used a microcomputer-based test of sustained attention and distractibility with children in grades K to 9 to examine variances in children's performance due to age and sex. First, they found that overall performance on the attention task continually improved up through the fifth grade, at which time performance began to deteriorate through to the ninth grade. Secondly, significant gender differences in performance on the attention task were observed and remained consistent across all grade levels. Boys displayed higher overall false alarm rates, indicative of a more impulsive style of responding, than girls. During the task, boys were more talkative and tended to be more restless in their seats than girls. These gender differences in performance were found to correlate significantly with scores on an abbreviated Conners' Teacher Rating Scale, that is, children who displayed greater restlessness in the classroom also exhibited less ability to sustain attention on LITERATURE REVIEW / 24 the computerized test. Murphy-Berman et al. concluded that the data do not suggest that girls possess superior attentional abilities, but that differences in attentional styles or a proneness to making different kinds of attention errors may well exist between the sexes, and that these observations correlate significantly with teacher ratings of in-class behaviour. Cooper and Farran (1988) conducted a study in which kindergarten teachers were asked to provide behavioural ratings of interpersonal and work-related behaviours on their students in an attempt to identify those behaviours critical to success. It was found that teachers were less tolerant of "off-task" behaviours and inattentiveness to instructions/routines than they were of disobedience or poor peer interactions. Teachers generally rated girls more positively than boys, where mean scores for girls " . . . were at least half a rating point higher than means for boys on two-thirds of the items" (p. 9). B. E A R L Y IDENTIF ICATION OF A T T E N T I O N A L DYSFUNCT ION Despite growing appreciation of the clinical spectrum of attentional deficits, relatively little investigation has been undertaken on the early identification of attentional problems within the preschool/kindergarten population and its ability to predict later academic achievement. Retrospective and longitudinal studies, although few in number have suggested that children demonstrating greater degrees of problematic behaviour associated with Attention Deficit Disorder (i.e., overactivity, limited attention span, and impulsivity) in the preschool years are more likely to continue to exhibit these behaviours at school-age, coupled with an elevated risk for poor academic progress and adjustment (Campbell et al., 1977; Chamberlin, 1977; McKinney, 1989). LITERATURE REVIEW / 25 Early correlational studies examining the relationship between classroom behaviour and academic achievement have suggested that many of the behaviours characteristic of learning disabled children are also associated generally with academic achievement (Bryan, 1974; Hoge & Luce, 1979; McKinney et al., 1975). Academic success was more likely attained amongst students who were independent, attentive and who interacted in a task-oriented manner during instruction rather than amongst those who displayed dependence, distractibility, poor task orientation (Hoge & Luce, 1979), poor concentration and behavioural disorganization (Palfrey et al., 1985). McKinney's (1989) recent longitudinal study on the behavioural characteristics of learning disabled children, spanning a 3-year period beginning with the first and second grades, revealed that these students, as a group, exhibited persistent patterns of maladaptive behaviour in the classroom which distinguished them from average achieving peers. Continued academic underachievement throughout the elementary school period was strongly associated with this behaviour. McKinney classified these LD children into more specific subgroups according to their patterns of behavioural strength and weakness, and found that children in the attention and conduct/classroom management problem subtypes experienced poorer academic outcomes 3 years later compared to those categorized as exhibiting withdrawn-dependent and normal types of behaviour. Interestingly, Garfinkel (1986) found that, in a clinic population of children referred for attentional deficits, many teacher and parent ratings of these children identify inattention and conduct behaviours as the two main factors associated within this referred population. Szumowski et al. (1987) support these findings, yet provide an additional LITERATURE REVIEW / 26 perspective which is not found in other research literature. They contend that dramatic developmental changes in behaviour during the preschool years make it difficult to distinguish between normal and deviant behaviours. Since many typical behaviours can be most annoying to parents/teachers, it becomes even more difficult to distinguish them from truly deviant behaviours. While some of these extreme forms of annoying behaviour may change into appropriate, mature behaviours during the developmental process, it is sometimes unclear as to when impulsivity, distractibility, inattentiveness and excess motor activity may well be early signs of Attention Deficit Disorder with or without hyperactivity. Szumowski et al. (1987) report on a longitudinal study by Fischer et al. (1984) in which parent ratings of behaviours of children were obtained at ages 2 to 6 and again at 9 to 15. Severe behavioural problems exhibited in children in the preschool years were significantly associated with ratings within the range of clinical disturbance at follow-up. Although these results showed a modest degree of stability for this high risk group of children, the authors reported that " discontinuity was more typical on an individual basis rather than continuity" (p. 87). Explanations of marked individual differences in achievement outcomes for children with LD are quite rare (Kavale, 1988), however, longitudinal research suggests that personal and social competencies can develop amongst some high risk children, making them more resilient to stress and responsive to intervention (Farran & McKinney, 1986). Palfrey et al. (1985) found that many preschool children, followed from 2 weeks of age to kindergarten entry, demonstrated a reduction in attentional deficit symptoms exhibited over time and appeared to be able to develop strategies to minimize early tendencies toward distractibility, lack of focus, or overactivity. Good cognitive LITERATURE REVIEW / 27 and language skills coupled with keen perceptual awareness appeared to aid in organizing their environment. A key factor in optimum responsiveness to developmental/behavioural intervention programs was lack of familial stress. It is the early identification of these adaptive characteristics that would aid in improving diagnostic, prognostic and remedial practices, and perhaps reduce later persistent school failure. 1. Early Risk Factors and Initial Manifestations Whether or not attention deficits emerge in early childhood remains questionable. The research literature does not appear to present a set of foolproof criteria that reliably distinguish extremes of typical behaviour from precursors of attention deficit disorders in preschool/kindergarten children. Instead, specific behavioural traits suggestive of attentional deficits commonly recur within the literature, that is, chronic inattention, poor self-monitoring (depicted by impulsive behaviour), frequent overactivity (Campbell et al., 1977; Douglas, 1972, 1983; Garfinkel, 1986; Sykes et al., 1973; Szumowski et al., 1987; van der Meere et al., 1988), disorganization, distractibility (Palfrey et al., 1985) and conduct disorders (McKinney, 1989). Within the psychiatric profession, diagnostic criteria of the ADHD/Undifferentiated ADD syndromes include inability to orient, focus, and organize one's attention on a specific task such that concentration is sustained, and impulsiveness (DSM HI-R, 1987). LITERATURE REVIEW / 28 2. Age at Onset Pin-pointing a specific age of onset for the early detection of attentional dysfunction has been very difficult. The DSM III-R asserts that manifestations of ADHD occur prior to the age of four in approximately fifty percent of diagnosed cases, with frequent lack of detection until school entry. Age of onset is not clearly specified for the Undifferentiated ADD category. Palfrey et al. (1985) conducted a longitudinal study which followed children (N=224) from 2 weeks of age to kindergarten entry to investigate the emergence of attention deficits. All children participated in an intensive educational and diagnostic program for the duration of this investigation. Very few concerns about attention were detected in infancy. The investigators suggest several possibilities for this finding: 1. Attention deficits are not problematic at this early stage of life. 2. The nature of the attention deficit symptom complex may differ at this age compared with older youngsters. 3. An infant characteristically does not encounter experiences and task demands that would bring out these symptoms. 4. Current measures used to elicit attentional weakness lack sophistication for this age group. The toddler period revealed significantly greater attentional concerns, with the 30-to 42-month age group showing the greatest proportion of children who were considered to meet criteria for definite attentional concerns some time before the age of five. Forty-one percent satisfied criteria for both definite and possible concerns within this period, as supported by the DSM III-R criteria for age of LITERATURE REVIEW / 29 onset. Thus, the results of this longitudinal study suggest that peak age of onset of attentional dysfunction lies between the ages of 3 and 4, implying that recognition of a sizeable portion of children destined to have attentional weakness could be possible prior to kindergarten entry. 3. Classroom Predictors of Later Academic Adjustment Identification of critical and non critical behaviours in the classroom has been directly associated with teachers' expectations of appropriate classroom behaviour (Rist, 1970). Rist indicated that teachers vary considerably in their expectations and that children, upon school entry, vary widely in their attainment of academic and interpersonal skills. According to Rist, it is this interaction between the variations in teacher expectations and in children's development that determine which classroom behaviours will serve to classify a child as being at-risk for later academic problems. Rubin and Balow (1978) reported a range of estimates of the prevalence of behavioural abnormalities, as identified by elementary classroom teachers, from 1.5% to greater than 30%. These researchers were not convinced that this sizeable range reflected the true prevalence of problem behaviours within schools, but instead attributed these fmdings to represent differences in teachers' tolerance limits for some problem behaviours. Rolf and Garmezy (1974) found that teachers showed great intolerability toward those children who did not apply their intellectual ability to school work. Spivack and Swift (1977), however, concluded that the earliest risk behaviours were inattentiveness and high external reliance (e.g. not following directions, lack of demonstrated independence, failure to use available resources). Cooper arid Farran (1988) found that kindergarten teachers LITERATURE REVIEW 7 30 could apparently tolerate disobedience and poor interaction with peers more than behaviours associated with being off-task and inattentive to routines, instructions and completion of assignments. From their results, Cooper and Farran proposed that kindergarten teachers consider themselves to play a primary role in teaching the skills necessary to learn and achieve, instead of focusing predominantly on the acquisition of appropriate socialization skills. Some studies have shown that preschool teachers frequently indicate uncertainty in listing characteristics that would best predict later academic success/failure (Becker & Snider, 1979; Keogh, Tehir & Windeguth-Behn, 1974). It has been suggested that teachers have difficulty identifying classroom behaviours that are representative of later poor achievement, but are better able to predict good achievement (Feshbach et al., 1974; Fletcher & Satz, 1984). Fletcher and Satz (1984) investigated the accuracy of teachers' global judgement of a preschooler's learning potential and found that preschool teachers correctly identified only 20% of students who later experienced substantial learning difficulties. Some controversy has existed as to whether teacher judgements are as accurate as standardized tests in identifying children at-risk for academic failure. Gerber and Semmel (1984) proposed that " . . . teachers are fallible judges and that their referrals constitute suspicions, as opposed to conclusions, of potential learning problems" (p. 148). Other researchers feel that teachers are in an excellent position to evaluate classroom behaviour, due to their familiarity with normative levels of children's behaviour during different developmental periods. Gresham, Reschly and Carey (1987) recognize the vast amount of time teachers spend in the classroom with their students daily, having the opportunity to observe individual behaviour LITERATURE REVIEW / 31 in many different circumstances to a much greater degree than an examiner who may spend one to two hours administering standardized tests to a child. They affirm, therefore, that teacher judgements " . . . are based on a much wider and comprehensive sampling of the content domain of achievement and classroom behaviour than standardized tests" (p. 544). Other researchers have supported that teacher ratings are representative of a vast array of characteristics/qualities that a child exhibits (e.g. work habits, cooperation, creativity) that standardized measures do not control for in their objective scores (Hardman & Oldridge, 1985). Gresham et al. (1987) further propose that " . . . regular classroom teachers are defendable "tests" of student achievement and perhaps should be used as one of the criteria by which psychoeducational tests are validated" (p. 551). Considering that validation of the original Binet scales relied on and showed correspondence to teacher judgements of intelligence and achievement, Gresham et al.'s proposal is well taken. Apart from parents, teachers have the greatest daily contact with children and " . . . their input is valuable particularly when juxtaposed with the information that can be gathered using traditional, individually administered tests" (Day & Peters, 1989, p. 357). These researchers found that teacher ratings of poor classroom behaviour and greater problems with peer relationships significantly discriminated normally achieving and underachieving Grade 3 and 4 students, based on language and mathematical skills. In general, therefore, research conducted over the last two decades largely supports teachers as being highly reliable and accurate judges in identifying those students likely to experience learning difficulties in later school years. Following an extensive review of research findings on both standardized LITERATURE REVIEW / 32 tests and teacher rating scales, Simner (1983) proposed five specific warning signs as being most effective in identifying kindergarten children truly "at-risk" for later school failure: (1) in-class attention span, distractibility, or memory span, (2) in-class verbal fluency, (3) in-class interest and participation, (4) letter or number identification skills, and (5) printing errors. Three of these signs, namely attention/distractibility/memory span, verbal fluency and letter/number identification skills, have been found to be the best overall in-class indicators of future academic success when considered in relation to other potential in-class warning signs, i.e. general coordination, intellectual independence, peer acceptance, adaptability to new situations and willingness to work hard (Dykstra, 1967; Stevenson et al., 1976). The data for the first two warning signs, attention/ distractibility/memory span and verbal fluency, were derived primarily from teacher rating scales (5-point scales) completed usually around the spring semester of the kindergarten year. Correlations between the presence of Simner's five specific warning signs assessed in kindergarten and later school achievement reveal coefficients approximating 0.50. These statistical data are significant since correlations of this magnitude have been known to demonstrate good predictive efficiency with many screening devices, correctly identifying approximately 70 to 80 percent of kindergarten children who later perform at or near the bottom of their class in the elementary grades (Mercer, Algozzine & Trifiletti, 1979; Simner, 1982 [as cited in Simner, 1983]). In a recent study, Simner (1987) found that teachers' accuracy in identifying "at-risk" preschoolers improved four-fold (82%), compared with Fletcher and Satz's (1984) results (20%), when "at-risk" status was determined from daily observations of Simner's five specific warning signs. LITERATURE REVIEW / 33 Barkley (1977) and Douglas (1972) have identified the standard cognitive factors typically examined by standardized measures in the evaluation of ADHD/Undifferentiated ADD in children and adolescents: (1) impaired concentration, (2) high rate of impulsive responses, (3) decreased reflection, (4) poorly organized responses, (5) decreased alertness, (6) failure to earn appropriate reward contingencies, (7) various visual motor integration problems, and (8) increased restlessness or purposeless motor activity. Prediction studies using intelligence tests have not proven particularly effective. Research studies have reported frequently used and reputable standardized tests such as the Wechsler Preschool and Primary Scale of Intelligence or the Metropolitan Readiness Tests, to rarely achieve correlations with later academic performance that surpass the range of correlations found for Simner's five effective warning signs (Dykstra, 1967; Flynn & Flynn, 1978). Fletcher and Satz (1984) suggested that intelligence tests are predictive only to the extent that valid precursor skills are measured. Adaptations of the Bender Gestalt test and selected subtests of the WISC-R have been used to assess impulsivity and attentional deficits; however, their ability to discriminate between hyperactive and nonhyperactive children has received criticism. Douglas (1972) asserted that these measures assess a wide variety of rather loosely defined cognitive abilities that may or may not be affected by impulsivity or difficulties in sustaining attention to tasks. LITERATURE REVIEW / 34 C. ASSESSMENT OF ATTENTION 1. Rating Scales Rating scales have long been used in evaluating the more global aspects of behaviour that may be difficult to measure through direct standardized measures. Since attention is believed to consist of an assortment of interactional processes dependent on the situational circumstances in which it is measured (Douglas, 1972; Parasuraman, 1983), behavioural rating scales have enjoyed widespread use with hyperactive children in assessing attentional/conduct behaviours within both the home and the classroom. Low correlations between parent and teacher behaviour ratings have demonstrated discrepancies in the identification of children at-risk for later academic failure (Chamberlin, 1977). Many children identified as malfunctioning at home have been considered to be performing satisfactorily at school. When given rating scales or checklists to evaluate similar behaviours, it would appear that teachers' and parents' ratings reflect those attentional behaviours they deem to be necessary within their respective environments. Although some controversy does exist, teacher rating scales have been found to be valid and reliable measures of classroom behaviour and indicators of potential learning problems (Gresham et al., 1987). Of the teacher rating scales, the Conners' Teacher Rating Scale (CTRS) has enjoyed the most popularity and use in evaluating - attentional deficits within the school situation. This scale has consistently proven to be drug sensitive and to correlate with classroom behaviour. Teacher rating scales, however, have received considerable criticism in their use with hyperactive children. They have been shown to have poor interrater LITERATURE REVIEW / 35 reliability and poor correlation between rated and objectively observed behaviours (Barkley, 1981). Their subjectivity has also been criticized as it allows for potential rater bias and the "halo" effect (Ross & Ross, 1976). Despite these drawbacks in teacher rating scales, they continue to be frequently used within the educational system and are considered to provide valid, reliable and valuable information in identifying children with attentional difficulties (Gresham et al., 1987). 2. Computerized CPT Within the last decade, microcomputers have replaced laboratory equipment in providing more precise data on impulsivity and sustained attention performance tasks. These microcomputers are frequently portable and allow data to be collected within a more natural setting (e.g., within a child's school or classroom), and are, therefore, believed to elicit a more realistic representation of a child's impulsive and vigilant behaviour that would interfere with learning. Microcomputers are capable of presenting stimuli at clearly defined intervals with simultaneous recording of subject responses. Both errors of omission, reflecting inattentiveness, and errors of commission, reflecting impulsivity, can be measured on computerized CPT's. In a study examining the usefulness of a computerized CPT (a PASCAL program on an Apple II computer) as a descriptive measure for attentional problems, Klee and Garfinkel (1983) found that CPT performance was significantly correlated with psychometric (MFFT, WISC-R Arithmetic and Coding subtests) and behavioural ratings (Conners' Teacher Rating Scales, [CTRS]) of inattention and hyperactivity. Additionally, the CPT performance was reported to LITERATURE REVIEW / 36 provide slightly greater sensitivity in identifying ADD and Conduct Disorder within a clinic population of children (mean age: 12.5 years) compared with the CTRS. Klee and Garfinkel suggested that the computerized CPT is a useful classroom screening device with similar application to the CTRS. Since a hospitalized psychiatric population was used in this study, this latter suggestion would require further investigation with a non-clinic-referred sample to see if similar conclusions could be drawn. As Murphy-Berman et al. (1986) found that teacher ratings (CTRS - abbreviated version) of motor restlessness correlated significantly with performance on a structured computerized attention test, they suggested that " . . . using this type of a computer game format to assess attention skills may be useful in predicting the children's behaviour in other situations requiring attentional ability" (p. 27). Gordon (1983) developed a computerized behavioural measure of ADD, the Gordon Diagnostic System (GDS), which he considers to be " . . . a precise, valid and efficient technique for the diagnosis of attention disorders" (p. 1). In an earlier study, Gordon (1979) proposed that the GDS could accurately discriminate between hyperactive and nonhyperactive children in samples derived from an outpatient clinic for severely emotionally disturbed children and a school-referred population. A later study investigating the diagnostic use of the GDS on a school-referred population alone yielded results which suggest that the GDS can distinguish ADD children from reading disabled, overanxious and normal children (Gordon & McClure, 1983). The GDS is discussed in more detail below. LITERATURE REVIEW / 37 3. Gordon Diagnostic System (GDS) The GDS is a microcomputer-based portable unit which permits the examination of three areas of attention related to ADHD and Undifferentiated ADD — impulsivity, sustained attention and distractibility — through easily administered game-like tasks. (As this research study involved a kindergarten population, the Preschool versions of the GDS tasks are described below. For more information on the versions pertaining to older children, the reader is referred to the GDS Instruction Manual (1986a)). The Delay Task examines a child's ability to refrain from responding, (i.e., button pressing), for at least four seconds from the previous response so as to gain a reward, with the reward points being cumulatively recorded on the display counter. Four separate trials are administered for this Task. The Vigilance Task examines a child's ability both to control impulsive responding and to sustain attention over a 6-minute interval. This Task requires the child to press the button whenever a "1" flashes on the front display amidst a series of randomly presented single-digit numbers. No performance reinforcement is given during the three Task trials. The Distractibility Task, a version of the Vigilance Task, examines the effects of distraction on a child's ability to sustain attention. It is used with children over the age of five. The task remains the same (standard setting), therefore, except two additional columns of flashing numbers on either side of the "hot" digit are introduced to distract the subject from the simultaneous presentation of the flashing "hot" numbers. LITERATURE REVIEW / 38 a. Research Use of the GDS Although Gordon (1986b) has reported the frequent use of the GDS in research studies investigating drug responsivity and pharmacotherapeutic success with brain damaged children, and in the clinical evaluation of ADD (with or without hyperactivity) by school systems and pediatricians, there remains a paucity of available research providing feedback on the effectiveness of the GDS in these areas. D. M E M O R Y A N D A T T E N T I O N A L D E F I C I T S Intact memory abilities have been considered to be strongly related to academic achievement (Swanson, Cochran, & Ewers, 1990). Although a fan-amount of research has addressed memory functions of learning disabled children, there is a paucity of literature available on the relationship of visual memory and attention deficit disorders. This is surprising since it has been shown that more individuals tend to be visual learners than auditory learners, with younger children being less efficient selectors of visual information than older children and adults (Enns & Akhtar, 1989). It has been reported that ADHD children experience greater difficulties attending to auditory than to visual information (Sykes, Douglas & Morgenstern, 1973). In examining performance differences in sustained attention between hyperactive and nonhyperactive children, Sykes et al. found that both of the research groups exhibited greater efficiency in attending to the visual than to the auditory form of the CPT. Although these two tasks were not comparable on all of the relevant dimensions, (i.e. intensity or distinctiveness of stimuli, etc.), these researchers proposed that information can be more effectively communicated with LITERATURE REVIEW / 39 young hyperactive children (mean CA. of hyperactive subjects in this study: 8 years 2 months; range of ages: 5 years 10 months to 11 years 5 months) through the visual modality than through the auditory modality. Using a slightly modified version of the CPT that was used in the Sykes et al. (1973) study, Swanson (1980) found that the visual modality elicited a greater rate of incorrect responding, as applied to reading deficits, compared to the auditory modality amongst both LD and control groups (mean CA. for LD group: 12.62 years). These results contradict those obtained by Sykes et al. (1973). Swanson suggested that although ". . . the visual modality may very well be a more effective medium for sustaining attention in young hyperactives, the inverse is suggested for older normoactives [learning disabled subjects]" (p. 76). Few studies of memory function in children with ATJHD/Undifferentiated ADD have been reported. Little is known about the contributory role of memory problems to clearly established cognitive deficits demonstrated by ADHD children (Douglas, 1983). Deficits in attention, particularly in divided attention, focused attention and sustained attention, found amongst ADHD children have been felt to reflect controlled information processing deficits (Fisk & Schneider, 1981). Fisk and Schneider (1981) proposed a model of attention which suggests that any interference in the rate at which information can be processed in working memory will result in attentional deficits. Relatively novel and specific task demands would require intact attentional skills so as to facilitate information to be processed slowly, serially and with effort. Once task demands have been well learned, information becomes automatically processed. Fisk and Schneider have shown that deficits in sustained attention are more apparent when tasks LITERATURE REVIEW / 40 require controlled processing instead of relatively automatic processing. These researchers found it necessary to make a distinction between controlled and automatic processing as they concluded that sustained attention deficits in ADHD require the use of controlled processing tasks. Van der Meere and Sergeant (1988) have attempted to clarify this proposed relationship between vigilance and controlled processing through explaining how memory operations are utilized in the CPT. In one CPT paradigm . . . the subject is given a constant target to retain in working memory and is instructed to press a button when the letter (or number) defined as the target appears in a train of stimuli. The subject retains the target in working memory, each stimulus is encoded, and a comparison occurs between the presented stimulus and the target. A decision is made, followed by a response, (p. 643) In this particular paradigm, similar to the GDS Vigilance Task (Gordon, 1986), the visual task can be easily learned, provided that sufficient time and trials are given, facilitating the development of relatively automatic processing. A second CPT paradigm is presented in which the constant target is not defined. Button pressing commences . . . when the previously appearing stimulus is the same as the current stimulus. The subject encodes a stimulus, retains it in working memory, encodes the succeeding stimulus, and compares the current with the prior stimulus. If a match occurs between the two in working memory, a decision is reached and a response is made. This paradigm requires more controlled processing than the former since the subject is continuously required to update working memory by defining new relations that do not remain constant over time, (van der Meere & Sergeant, 1988, p. 643) LITERATURE REVIEW / 41 E. HYPOTHESES From reviewing the literature, the following research hypotheses were generated: • There will be no significant (p<.05) relationship (correlation) between test scores of impulsivity (GDS) and teacher ratings on the 13 variables of the School Learning Profile (SLP) for normally achieving and "high risk" kindergarten students. • There will be no significant (p<.05) relationship (correlation) between test scores of vigilance (GDS) and teacher ratings on the 13 variables of the School Learning Profile (SLP) for normally achieving and "high risk" kindergarten students. • There will be no significant (p<.05) relationship (correlation) between teacher ratings of attention and school readiness (overall ability to learn school materials) for each of the normally achieving and "high risk" groups. • There will be no significant difference (p<.05) between control and "high risk" groups on teacher ratings of attention and school readiness (overall ability to learn school materials). • There will be no significant (p<.05) difference between the normally achieving and "high risk" groups on the GDS measures of impulsivity (Delay Task ~ Total ER score) and sustained attention (Vigilance Task -- Total Correct Responses and Total Commission Errors scores). • There will be no significant (p<.05) difference between the normally achieving and "high risk" groups on their Vigilance Task performance over time, (i.e., Blocks 1, 2 and 3). LITERATURE REVIEW / 42 There will be no significant (p<.05) relationship (correlation) between the GDS measures of sustained attention (Vigilance Task ~ Total Correct Responses and Total Commission Errors scores) and impulsivity (Delay Task -- Total Efficiency Ratio score) for the normally achieving and "high risk" groups. There will be no significant (p<.05) difference between boys and girls on test scores of the GDS Vigilance Task (inattentiveness). There will be no significant (p<.05) relationship (correlation) between test scores of impulsivity (GDS) and visual memory (Bead Memory — Stanford-Binet Intelligence Scale: Fourth Edition) for normally achieving and "high risk" kindergarten students. There will be no significant (p<.05) relationship (correlation) between test scores of sustained attention (GDS) and visual memory (Bead Memory — Stanford-Binet Intelligence Scale: Fourth Edition) for normally achieving and "high risk" kindergarten students. The following chapter presents the methodology used in this ploratory study in testing each of these hypotheses. III. METHOD Based on the premise that teacher judgements about a student's ability to learn have good content validity (Gresham, Reschly & Carey, 1987), this study investigates what kinds of similarities and differences exist between subjective teacher ratings and objective Gordon Diagnostic System (GDS) standardized scores in identifying those kindergarten children who manifest attentional and/or impulse control difficulties and are likely to experience later learning problems. This research study developed out of a larger study. For the purposes of the present study, the researcher selected specific subjective and objective measures, from those established by a larger study (J. Carter, doctoral dissertation in progress), from which to collect and analyze data. A. SUBJECTS 1. Selection Procedure Two groups of kindergarten students from 12 elementary schools within School District No. 71 (Courtenay) on Vancouver Island were selected for participation in this research project. All District kindergarten teachers were sent a letter describing the study (see Appendix A) and several copies of the Kindergarten School Learning Profile (Carter & Conry, 1988). (see Appendix B) Participating teachers were asked to identify those students whom they considered, based on their teaching experience, to be functioning within the lowest ten percent of the normal curve, as illustrated in item 1 of the Kindergarten 43 METHOD / 44 School Learning Profile (SLP), in their overall ability to learn school material. Once teachers had nominated "high risk" students, computerized systematic random selection of the remaining District kindergarten students (with the exclusion of multiply handicapped special needs students, those enrolled in the French immersion and cadre programs, and one non-participating regular kindergarten class) yielded the normally achieving students that comprised the control group, in proportion to the number of referred boys and girls in the "high risk" group. Teachers were then given SLPs to complete on these normally achieving control group students as well. Parental consent to collect data was obtained for all children, both computer-selected and teacher-nominated, participating in the study (see Appendix C - Parent Permission Letter). Excluding multiply handicapped special needs students, those enrolled in the French immersion and cadre programs, and those students of a class in which the teacher declined participation, the school district had a total of 396 registered kindergarten students. Obtaining a sample of 29 control and 31 "high risk" subjects each represented approximately seven percent of the total District population participating in the study. 2. Sample Composition Of the 60 kindergarten children selected, a complete data file was obtained on 58 students, (i.e., one child from each of the control and "high risk" groups was eliminated as complete data on the Gordon Diagnostic System measure was lacking). At the time the GDS was administered, the mean chronological age of the entire sample was 71 months, with a standard deviation of 4.41. The METHOD / 45 teacher-nominated "high risk" group consisted of 30 kindergarten students — 19 boys (63%) and 11 girls (37%) — ranging in age from 64 months to 82 months (mean age: 71 months, standard deviation: 4.55). The control group was comprised of 28 students (15 boys (54%) and 13 girls (46%)), ranging in age from 59 months to 76 months (mean age: 70 months, standard deviation: 4.28). B. INSTRUMENTATION 1. Ratings Kindergarten School Learning Profile (SLP) Completion of the SLP for each student participating in the study provided teacher ratings of academic and behavioural performances, both within and outside of the classroom setting. The SLP was used to initially identify the teacher-nominated "high risk" sample, considered to be functioning within the lowest 10% for overall learning ability of school material. This rating scale is comprised of thirteen items, the last having 10 subitems, each rated on a five-point scale. Several different rating response categories exist. With an illustration of the normal curve, the first item requests teachers to identify the percentage range for overall learning ability within which each student was believed to be functioning at the time, (i.e., ratings of 1 = lowest 10%, 2 = lowest 30% but not lowest 10%, 3 = middle 40%, 4 = upper 30% but not highest 10%, and 5 = highest 10%). All students considered to be "high risk" candidates for school learning ability, as definition in this study, were assigned the rating of 1, (i.e., lowest 10%). On the remaining twelve SLP items, each point on the five-point rating METHOD / 46 scale corresponds to response categories ranging from "Poor" to "Very Good", "Poor" to "Perfect printing", "None" to "All correct", etc. Item 13 is comprised of ten subitems which examine exhibited behavioural characteristics, both within and outside of the classroom situation, believed to influence a child's learning abilities. These subitems were derived from the diagnostic criteria of ADHD and Undifferentiated ADD set forth by the American Psychiatric Association in its revised third edition (1987) of the Diagnostic and Statistical Manual of Mental Disorders (DSM ITI-R). On the SLP, a five-point rating scale for these ten subitems identifies approximate frequencies of these exhibited behaviours with response categories ranging from "Never" to "Constant". 2. Standardized Measures Select standardized measures, from those outlined in the doctoral study, were pertinent to the present study and used for data collection and analysis. The following psychometric instruments were individually administered to control and "high risk" students: • The Gordon Diagnostic System — Delay Task (Preschool Setting) — Vigilance Task (Preschool Setting) • The Stanford-Binet Intelligence Scale: Fourth Edition — Bead Memory subtest data only METHOD / 47 a. Gordon Diagnostic System (GDS) The GDS is a microprocessor-based portable unit which utilizes three game-like tasks to objectively measure vigilance, impulsivity and distractibility, purportedly without interference from other factors such as intelligence or visual-motor skills. Only two of the subtests were administered for this research study. They are as follows: (i) Delay Task (Preschool Setting) The Delay Task investigates the child's ability to refrain from impulsive responding in order to obtain a reward. Based upon a Differential Reinforcement of Low Rates (DRL) behaviour schedule, this task requires the child to press a button, wait, and then press again. If a minimum of four seconds elapse between presses, a red light flashes and the front display visually keeps track of the accumulated rewards, or points. Responding prior to the elapse of this four-second delay interval results in the time being reset and no reward points recorded. The Task lasts for 6 minutes, with four 90-second trials, or Blocks. Three scores are obtained: the number of responses (button presses), the number of correct responses, and the Efficiency Ratio (ER), which represents the percentage of correct responses. This latter score, Total ER, represents the child's level of impulsivity. (ii) Vigilance Task (Preschool Setting) This task is a version of the Continuous Performance Task (Rosvold et al., 1956) and examines the child's ability to sustain attention to a task over time. A series of digits flash at 0.2 second intervals on the front display. The child is required to press the button every time a "1" appears. This task lasts 6 minutes, with three 2-minute trials, or Blocks. Three scores are obtained: the METHOD / 48 number of correct presses (reflecting the ability to achieve and maintain alert, vigilant responding), the number of omission errors, (i.e., the number of times the child fails to press the button when the "1" appears), and the number of commission errors, (i.e., extraneous button presses), reflecting the degree of impulsive responding and not lapses in sustained attention. Modified recording sheets for the GDS were used (See Appendix D). Standardization of the GDS The GDS Delay and Vigilance Tasks were standardized on 700 children, aged 3 to 16 years, randomly selected from nine public and private schools in Virginia and New York State. Normative data for the Distractibility Task were derived from only 362 children from the same schools. To qualify for the standardization sample, teacher ratings — either Teacher Rating Form (Edelbrock & Achenbach, 1984) or Teacher Rating Scale (Conners, 1969) — of all children had to indicate a nonhyperactive population with no history of grade retention. Three ranges of scores for these tasks were determined in the identification of significant attentional deficits from acceptable attentional difficulties in performance exhibited — ABNORMAL: score falls at or below the 5th percentile; BORDERLINE: score falls between the 6th and 25th percentiles; NORMAL: score falls above the 25th percentile. Those children typically identified as hyperactive have been found to perform in the ABNORMAL range of the GDS scores (Gordon, 1984). In examining the comparative diagnostic use of traditional assessment techniques, (i.e., WISC-R, Bender Gestalt Designs, PIAT, Sentence Completion, Visual Aural Digit Span, Children's Personality Questionnaire, and METHOD / 49 Draw-A-Person), and the GDS, Gordon (1984) found that school psychologists expended a great deal of time in their assessments and that they tended to overdiagnose ADD when relying on traditional assessment measures alone. b. Stanford-Binet Intelligence Scale: Fourth Edition (SBFE) The SBFE is a commonly used standardized instrument assessing intellectual functioning, (i) Bead Memory Subtest This is a visual memory and gross/Fine motor task requiring the subject to reproduce a specific bead sequence of varying shapes and colours onto a vertical standing rod. The subject places the beads on the rod following a five-second pictorial presentation of the stimulus bead sequence. Successful execution of this task relies minimally on language competency (comprehension of instructions). C. PROCEDURE Prior to the administration of any of the standardized psychometric instruments, teachers completed the SLP for each student in the spring (March and April) of 1988. According to the methodological criteria established by the doctoral study, tests were administered in a counter balanced order during the latter months (May and June) of the 1987-88 school year. A minimum of two testing sessions over a two-day period were required with each child to complete data collection. The GDS and the SBFE were usually administered within the same session. The counter balanced test administration schedule, pertinent to the doctoral study only, ensured that approximately half of the control group and METHOD / 50 half of the "high risk" group received the GDS and the SBFE in May 1988. Following a one month interval, the remainder of the two groups was tested. All children were tested in their elementary schools during kindergarten hours (morning or afternoon) attended in their respective schools. Test administrators included a university school psychology professor, two district special education counsellors and four graduate students all trained in the administration and scoring of the tests used in this study. D. LIMITATIONS Although it is not a standardized, psychometrically sound instrument, the SLP was designed for use in this study to (1) identify the teacher-nominated "high risk" group and (2) compare subjective teacher ratings/evaluations of kindergarten students' behaviours with an objective behaviour-based standardized measure (GDS) that examines a child's ability to sustain attention and control impulsive responding. While no empirical data supporting use of the SLP exists, the construct of teacher ratings alone does bear considerable credibility in the assessment of children's behaviour. Teacher rating scales have enjoyed widespread use for many years, both within educational settings and for research purposes, having been found to possess good psychometric reliability and validity in assessing that which they purport to assess (Gresham et al, 1987; Hardman & Oldridge, 1985). Use of a teacher rating scale which is based on items found in commonly used standardized teacher rating scales, (e.g., Conners' Teacher Rating Scale) therefore, is considered to be justifiable for the purposes of this study. Additionally, the latter 10 behavioural items of the SLP, (i.e., item #13), are directly derived from the DSM IH-R's diagnostic criteria for METHOD / 51 ADHD/Undifferentiated ADD. As these diagnostic criteria continue to be used by the psychiatric profession and are considered to be reliable and valid diagnostic indicators of these Disorders, incorporation of these behavioural criteria in the SLP is once again considered to be justifiable for the purposes of this study. It is important to recognize, however, that the DSM III-R provides a classification system as a useful and convenient conceptual framework in describing syndromes. This classification system, however, is not static and is prone to frequent changes in the specification of the criteria associated with the Disorders, (e.g., both the diagnostic label and criteria associated with attentional and impulse control difficulties in children has changed considerably from the DSM II edition to the DSM HI-R edition, currently used). Apart from its questionable psychometric credibility, an additional limitation of the SLP is the unknown extent to which the items achieve any discriminatory power. For example, teachers may find it difficult to discriminate between rating a child's attention span, as related to following classroom instructions (item #2) and his/her ability to listen (item #13G). As no intention to achieve any degree of discriminatory power between the items existed in the development of this opinion-based instrument (personal communication with J. Carter, 1990), this information is unavailable and does present some limitations in the interpretation of the data, and thus, prevents any definitive, conclusive statements/arguments from being made. Related to this latter limitation is an additional one. While some items appear to assess more global aspects of attention/impulse control, (e.g., item #1 — overall ability to learn school material; item #2 — attention span, as related to following classroom instructions), others appear to address more specific aspects METHOD / 52 (e.g., item #13J — blurts out answers to questions before they have been answered; item #13G — difficulty listening). It is not known to what extent each item evaluates intact attentional and impulse control skills. In summary, interpretive limitations in the obtained SLP data do unquestionably exist, however, use of the SLP in this study is justifiable as the construct of teacher ratings in obtaining teachers' opinions of students' school behaviours has been found to be both valid and reliable (Gresham et al., 1987; Sattler, 1988) Additionally, teachers' opinions generated by the SLP in this study were based on an 8-month observation period, (i.e., SLP's were completed in March/April of the kindergarten year, commencing in September of the previous year), lending greater reliability to teachers' judgements. E. DATA ANALYSIS In testing the hypotheses of this study, data analyses included analysis of of variance, repeated measures analysis of variance, and correlational (Pearson product-moment correlation coefficients) statistics. The results of these analyses are presented in Chapter 4. IV. RESULTS This chapter presents a description of the sample population and the results of data analyses pertaining to each hypothesis being investigated in this research study. The level of significance for acceptance or rejection of the hypotheses was selected as pS.05. This p value has been adjusted using the Bonferroni adjustment (Glass & Hopkins, 1984) where necessary, so as to reduce the possibility of making a Type I error. A. DESCRIPTIVE STATISTICS Sixty kindergarten students conformed to the criteria specified in the previous chapter for subject selection for participation in this study. A complete data file was obtained on only 58 of these students — 28 control and 30 "high risk" subjects. Interpretations of the obtained data for the sexes of each sample group need to take into consideration each group's sample size. While approximately equal numbers of control boys (n=15) and girls (n=13) were present, the "high risk" group was characterized by 19 boys and only 11 girls. The first item of the School Learning Profile (SLP) — Overall Learning Ability for School Material — was used to obtain the teacher-nominated "high risk" sample. Table 1 illustrates the SLP frequency distribution of rating values by group and by sex. By definition, "high risk" subjects alone received a rating score of 1 (lowest 10% for overall learning ability of school material). Approximately 27% more boys than girls were nominated by their teachers as being at-risk for overall learning ability. Within the control group, 4 was the 53 Jable 1 School Learning Erode Sample Frequency Distribution Entire Sample (n=58) Boys (n=34) Girls (n=24) Category Rating Value Control High Risk Control High Risk Control High Risk Lowest 10%' 1 0 30 (52%) 0 19 0 11 Lower 30% but not lowest 10% 2 1 (2%) 0 1 0 0 0 Middle 40% 3 6 (10%) 0 4 0 2 0 Upper 30% but not highest 10% 4 15 (26%) 0 8 0 7 0 Highest 10% 5 6 (10%) 0 2 0 4 0 Totals 28 (48%) 30 (52%) 15 19 13 11 'All "high risk" subjects, by definition, were assigned to the lowest 10% category for overall ability to learn i school material. CO <=} H CO RESULTS / 55 most frequently assigned rating value, (i.e., upper 30% but not highest 10%), with the rating value of 2, (i.e., lower 30% but not lowest 10%), being assigned least frequently. Descriptive and analysis of variance statistics for the School Learning Profile, the Gordon Diagnostic System and the Bead Memory subtest of the Stanford-Binet Intelligence Scale: Fourth Edition are reported in Tables 2, 3 and 4. Analyses of variance (ANOVA) procedures were used on the School Learning Profile and the Bead Memory subtest while repeated measures analyses of variance (ANOVAR) procedures were used on the Delay and Vigilance Tasks of the Gordon Diagnostic System in examining the effect on performance over time, (i.e., task trials, or Blocks). The Bonferroni adjustment was utilized for the analyses of variance performed on the SLP and GDS Tasks to reduce the possibility of making a Type I error. B. HYPOTHESES 1. Hypothesis 1 There will be no significant (p<.05) relationship (correlation) between test scores of impulsivity (GDS) and teacher ratings on the 13 variables of the SLP for normally achieving and "high risk" kindergarten students. 2. Hypothesis 2 There will be no significant (p<.05) relationship (correlation) between test scores of vigilance (GDS) and teacher ratings on the 13 variables of the SLP for normally achieving and "high risk" kindergarten students. RESULTS / 56 Table 2_ Means (Standard Deviations) by Group on the School Learning Profile with Analysis of Variance for Group Effect Item - Controls1 "High Risk"2 MSe F(l,56) 1. Overall Learning Ability 3.929 (0.766) 1.000 (0.000) 0.28 438.66* 2. Attention Span & Distractibility 4.107 (1.031) 1.967 (0.850) 0.89 74.85* 13 A. Fidgets 1.214 (0.832) 2.345 (1.203) 1.08 16.90* B. Difficulty Staying Seated 0.929 (0.979) 2.100 (1.269) 1.29 15.34* C. Difficulty Waiting Turn 0.893 (1.066) 1.633 (1.402) 1.56 5.07 D. Easily Distracted 1.286 (1.013) 3.000 (0.983) 0.99 42.78* E. Defiant & Uncooperative 0.286 (0.600) 0.900 (1.213) 0.94 5.84 F. Temper Tantrums 0.143 (0.448) 0.267 (0.691) 0.35 0.64 G. Difficulty Listening 0.893 (1.066) 2.700 (1.119) 1.20 39.54* H. Difficulty Playing Quietly 0.821 (0.819) 1.500 (1.253) 1.13 5.87 I. Short Attention Span 0.679 (0.905) 2.333 (1.155) 1.08 36.54* J. Blurts Out Answers 0.786 (0.995) 1.267 (1.285) 1.33 2.51 Total Behaviour Score 7.929 (7.086) 17.967 (8.487) 61.52 23.72* 1 n = 28 2 n = 30 Note: *pS0.0038, with Bonferroni adjustment RESULTS / 57 Table _3_ Means (Standard Deviations) by Group on the Delay Task of the Gordon  Diagnostic System and Bead Memory (SBFE) Variable Controls1 "High Risk"2 Total Correct Responses 44.786 (11.746) 34.333 (10.060) Total Responses 68.607 (21.226) 71.400 (28.657) Block 1 Correct Responses 10.893 16.929 ( 3.685) ( 6.588) 8.667 16.767 ( 3.231) ( 8.881) Block 2 Correct Responses 11.321 17.179 ( 3.560) ( 5.831) 8.833 19.433 ( 2.878) (11.358) Block 3 Correct Responses 11.107 17.393 ( 3.270) ( 6.839) 8.300 17.800 ( 3.153) ( 8.829) Block 4 Correct Responses 11.464 17.107 (• 3.305) ( 5.593) 8.533 17.400 ( 3.340) ( 8.177) Total Efficiency Ratio (ER) 0.693 ( 0.182) 0.588 ( 0.236) ER Block Variability 0.121 ( 0.059) 0.145 ( 0.088) Slope Scope 0.001 ( 0.164) -0.038 ( 0.243) Block 1 ER 0.708 ( 0.227) 0.645 ( 0.278) Block 2 ER 0.702 ( 0.206) 0.585 ( 0.269) Block 3 ER 0.706 ( 0.227) 0.570 ( 0.270) Block 4 ER 0.717 ( 0.211) 0.578 ( 0.245) Bead Memory (SBFE) 49.429 ( 7.346) 39.167 ( 5.113) 1 n = 28 2"r7=30 RESULTS / 58 Table _4 Means (Standard Deviations) by Group on the Vigilance Task of the Gordon Diagnostic System GDS Variable Controls1 "High Risk"2 Total Correct Responses 26.643 ( 2.641) 21.567 ( 5.618) Block 1 Correct 11.357 ( 0.989) 10.167 ( 2.365) Omission Errors 0.643 ( 0.989) 1.833 ( 2.365) Commission Errors 2.464 ( 3.305) 7.200 ( 8.849) Block 2 Correct 8.357 ( 1.193) 6.233 ( 2.144) Omission Errors 0.643 ( 1.193) 2.767 ( 2.144) Commission Errors 0.929 ( 1.762) 4.267 ( 8.674) Block 3 Correct 6.929 ( 1.412) 5.167 ( 2.151) Omission Errors 1.071 ( 1.412) 2.833 ( 2.151) Commission Errors 0.857 ( 1.976) 4.200 ( 7.911) Total Omission Errors 2.357 ( 2.642) 7.433 ( 5.618) Total Commission Errors 4.250 ( 5.873) 15.867 ( 24.093) Commissions Block Variability 1.34179 (1.58480) 2.75133 (2.49286) 1n_=28 2 n = 30 For the purposes of this study, the School Learning Profile in its entirety was unsuitable for data analysis. Specific variables, namely items 1, 2 and 13 (A to J), were selected in the testing of the above research hypotheses. a. GDS Delay TasklSLP Correlations In examining the correlation coefficients for the control and "high risk" groups between the Total Efficiency Ratio score and the specifically selected SLP variables mentioned above, coefficients are predominantly negative, low, and are RESULTS / 59 all insignificant, using the Bonferroni adjustment of the p value, (i.e., 0.05/13). (see Table 5) These results suggest that a significant relationship does not exist between kindergarten students' ability to refrain from impulsive responding on a microcomputer delay task and teachers' ratings of attention and impulse control, as believed to affect their ability to learn school materials. The negative directionality of this correlation reflects an inverse relationship between these two variables within both sample groups due to the reversal in the quality of rating values between SLP items 1 and 2 and all of item 13, (i.e., low scores on items 1 and 2 denote difficulties whereas on item 13, they represent no difficulty). b. GDS Vigilance TasklSLP Correlations A. Total Correct Responses Score Once again, correlation coefficients for the entire sample appear to be predominantly low (range: -0.503 to 0.257), displaying a negative directionality (see Table 5). No statistically significant relationships (using the Bonferroni adjustment) between GDS Vigilance Task performance and teacher ratings of attention and impulsivity were apparent for the normally achieving group. The "high risk" group, however, yielded only one significant coefficient between Total Correct Responses and SLP item 131 (fails to finish things started [short attention span]), reflective of a strong relationship existing between these two variables. In summary, significant relationships between teachers' ratings of kindergarten students' attentional skills, both within and outside of the classroom situation, and these students' Total Correct Responses scores on the GDS Table i Intercorrelations oJ School Learning Profile {SLE) Eatings SOi GDS Performance £21 Ihe Dotal Sample GDS Delay Task GDS Vigilance Task SLP #1 Total E.R. Total Correct Responses Total Commission Errors Overall Learning Ability Student Learning Profile (SLP) Control High Risk Control High Risk Control High Risk Control High Risk1 Item #1 Overall Learning Ability -0.035 —1 -0.068 —1 -0.234 —1 _i Item #2 Attention span and distractibility 0.272 -0.021 0.069 0.257 -0.090 -0.308 0.667* Item #13 A. Fidgets -0.178 0.140 -0.284 0.141 0.072 0.284 -0.323 B. Difficulty staying seated -0.310 -0.069 -0.225 -0.173 0.048 0.360 -0.451 C. Difficulty waiting turn in games or group activity -0.065 -0.238 -0.146 -0.437 0.200 0.506* -0.372 D. Easily distracted -0.115 -0.059 -0.113 -0.187 0.112 0.334 -0.402 E. Defiant and uncooperative -0.114 -0.076 -0.050 -0.173 0.242 0.585* -0.357 F. Has temper tantrums -0.047 -0.227 -0.080 -0.457 -0.155 -0.041 -0.293 G. Difficulty listening -0.353 0.084 -0.172 -0.153 0.123 0.341 -0.463 H. Difficulty playing quietly 0.054 -0.148 -0.219 -0.321 0.079 0.236 -0.198 I. Fails to finish things started (short attention span) -0.234 -0.245 -0.019 -0.503* 0.239 0.106 -0.568* J. Blurts out answers to questions before they have been completed -0.100 -0.214 -0.326 -0.146 0.219 0.439 -0.118 Total Behaviour Score -0.192 -0.143 -0.211 -0.324 -0.090 0.453 -0.444 •p^ O.0038, with Bonferroni Adjustment 'Correlation coefficients could not be computed due to lack of variance in scores. Note: negative correlations are representative of direct relationships between variables. due to the artifact of the rating values. RESULTS / 61 Vigilance Task are not apparent from the data analysis conducted. B. Total Commission Errors Score Inspection of the nature of the relationship between the number of commission errors elicited on the GDS Vigilance Task and teacher ratings of impulsivity and attention on select SLP items suggests few differences between the control and "high risk" groups. Although the normally achieving control group yielded Pearson correlation coefficients that were very low (range: -0.159 to 0.242), predominantly positive, and, without exception, did not achieve significance (using the Bonferroni adjustment), the "high risk" group yielded somewhat higher positive coefficients (range: -0.308 to 0.585), with two achieving statistical significance. Teacher ratings on SLP items 13C (difficulty waiting turn in games and group activities) and 13E (defiant and uncooperative) both appear to bear a relatively strong relationship to the degree to which "high risk" students made errors of commission on the GDS Vigilance Task. In summary, correlational data analyses suggest that significant relationships between teachers' ratings of kindergarten students' attentional skills and Total Commission Errors scores on the GDS Vigilance Task are apparent for only few of the SLP items and only amongst the "high risk" group. The normally achieving group exhibited no significant correlations. Teacher ratings of 'defiant and uncooperative behaviour' amongst the "high risk" group correlated most highly, with 'difficulty waiting turns in games/activities' trailing closely, with the number of commission errors made by this group on the GDS Vigilance Task. Thus, comparing impulsivity and vigilance performances of kindergarten RESULTS / 62 students on objective and subjective assessment measures (GDS and SLP teacher ratings, respectively), reveals fewer significant relationships than originally anticipated. Correlational statistical data analyses indicate variability in the strength, directionality and significance of the obtained Pearson product-moment correlation coefficients amongst the performance scores on the above measures. 3. Hypothesis 3 There will be no significant (p<.05) relationship (correlation) between teacher ratings of attention and school readiness (overall ability to learn school materials) for each of the normally achieving and "high risk" groups. 4. Hypothesis 4 There will be no significant difference (p< .05) between normally achieving and "high risk" groups on teacher ratings of attention and school readiness (overall ability to learn school materials). a. Correlations Examination of the correlational data analyses first reveals noteworthy results between the normally achieving and "high risk" groups, (see Table 5) While the correlation coefficients for the normally achieving controls are generally low (range: -0.118 to 0.667) and predominantly negative with few significant p values, coefficients for the "high risk" group were unconfutable due to lack of variance in the SLP item #1 score for this group. By definition of "high risk" group membership, teacher-nominated students were all assigned the rating value of 1, (i.e., within the lowest 10% for overall ability to learn school material). RESULTS / 63 Inspecting only the correlational data analyses of the control group, therefore, indicates that SLP item #2 (attention span and distractibility) is the most highly correlated (r= 0.667) with SLP item#l (overall learning ability) with a p<0.001. Of the behavioural ratings for the SLP#13 subitems, 131 (fails to finish things started — short attention span) correlated most highly (r =-0.568) with item #1 (overall learning ability), achieving statistical significance. Although a negative Pearson r value was obtained, this correlation coefficient represents an inverse relationship between these two variables. These correlational results suggest, in fact, that teachers' ratings of control kindergarten students' attention span, based on their ability to complete tasks they have started, strongly relates to assigned ratings of their general school readiness, or overall learning ability, on SLP item #1. Low and statistically insignificant correlation coefficients arise on SLP items 13H (difficulty playing quietly) (r=-0.198, p>0.05) and 13J (blurts out answers to questions before they have been completed) (r = -0.118, p>0.05). Although moderate correlations were obtained between SLP item #1 and item 13B (difficulty staying seated) - r=-0.451, p = 0.008, item 13G (difficulty listening) - r=-0.463, p = 0.007, and the Total Behaviour score of SLP item 13 — r=-0^ 444, p=0.009, they were found to be only approaching significance, taking into consideration the Bonferroni adjustment. Intercorrelations between SLP#2 (attention span and distractibility — following directions in class) and SLP#13 subitem ratings yielded many highly significant and strongly correlated relationships, (see Table 6 and Figure 1) Once again, negative Pearson r values represent inverse relationships between variables since lower ratings on the SLP#13 subitems are indicative of less frequently RESULTS / 64 Table _6 Intercorrelations of SLP _2_ (Attention Span/Distractibility) and SLP 13 (Behaviour) Ratings for the Sample SLP 2 Attention Span/Distractibility (following directions in class) SLP 13 Controls r 1 "High Risk" r 2 A. Fidgets -0.675* -0.527* B. Difficulty Staying Seated -0.690* -0.476* C. Difficulty Waiting Turn -0.663* -0.474* D. Easily Distracted -0.775* -0.743* E. Defiant & Uncooperative -0.531* -0.338 F. Temper Tantrums -0.195 -0.219 G. Difficulty Listening -0.798* -0.808* H. Difficulty Playing Quietly -0.284 -0.340 I. Short Attention Span (fails to finish things started) -0.716* -0.550* J. Blurts Out Answers -0.266 -0.528* Total Behaviour Score -0.724* -0.674* *pS 0.0045, with Bonferroni adjustment 1n_=28 2n = 30 Note: negative correlation coefficients represent direct relationships between variables, due to the artifact of the rating values. RESULTS / 65 A ft i 2 * Irlni 1* ll 41 I F si «1 s i f 1 1 d i £ 5 4 1 i i f-CcS? i $ CoMTtKS " H , W ^ " Figure _1_ Intercorrelations of SLP _2_ (Attention Span/Distractibility)  and SLP 13 (Behaviour) Ratings for the Sample RESULTS / 66 displayed problematic behaviour. In general, somewhat stronger relationships between SLP#2 and the SLP#13 subitems are evident for the normally achieving group than for the "high risk" group, however, both sample groups yielded the highest statistically significant r values for the same three variables within their respective group, (i.e., 13G [difficulty listening], 13D [easily distracted], and 131 [short attention span — failure to complete things started]). Both sample groups yielded the lowest r values on subitem 13F (temper tantrums) representative of weak and statistically insignificant relationships with SLP#2 ratings. 'Difficulty playing quietly' (subitem 13H) revealed a low to moderate relationship, with an insignificant p value, with SLP#2 for both groups. While subitem 13J (blurts out answers to questions before they have been completed) correlated weakly to SLP#2 (r=-0.266) for the normally achieving students, a highly significant strong relationship (r = -0.528) between these two variables was obtained for the "high risk" group. Conversely, the control group yielded a statistically significant coefficient (r=-0.531) on subitem 13E (defiant and uncooperative), while an insignificant and lower coefficient (r=-0.338) was obtained for the "high risk" group. The difference between the strength of these two correlations was not found to be significant using the Fisher's z transformation of r (Glass & Hopkins, 1984). The Total Behaviour score, a cumulative score of all the SLP#13 subitem ratings combined, also yielded highly significant and high r values for both of the sample groups. RESULTS / 67 b. Means Means on the specific SLP items investigated in this study between the control and "high risk" groups were predominantly significantly different, as determined through analysis of variance, (see Table 2 and Figure 2) As expected, the teacher-nominated "high risk" group received a mean rating of 1.000 on school readiness, or' overall ability to learn school material' (SLP item #1), with no variance. The control students, on average, were rated at the upper end of the middle 40% range (x = 3.929, sd = 0.766) on this variable — a significant mean difference from the "high risk" group. Once again, significant mean differences in ratings of attention span for both groups on SLP items #2 and #131 suggest that teachers appear to be rating attentional abilities as they relate to specific skills, (i.e., 'following directions in class' and 'finishing things started both within and outside of the classroom situation', respectively). The mean . rating for the normally achieving control group on SLP item 2, (i.e., x = 4.107, sd= 1.031), approached a qualitative "very good attention" rating, (i.e., 5). This is supported by the low mean rating on SLP item 131 (x = 0.679. sd = 0.905) for this group suggesting that these children, on average, were not devoid in exhibiting a failure to complete things started, however, they did so less often than "rarely". These control group findings do illustrate some consistency between ratings of attention span on these two SLP items, while ratings for the "high risk" group display less consistency. While the SLP item #2 mean rating for the "high risk" group qualitatively approaches "poor attention", (i.e., x = 1.967, sd=0.850), close to the same degree to which control group ratings approach "very good attention", the mean rating on SLP item 131 (5 = 2.333, sd= 1.155) suggests that teachers felt that children RESULTS / 68 Ld 4 V or r fat I N e v t r O- —I 1 1 1 r C t> E F iff *I * * - i 1 1 r ~ S rt • r ^ T 6 - . S3 1* 3 3 3 «-> u ' i r . « £ f l f i i i t e l l r t i 31 M tx. ^ T» p* _ KEY CONTROLS 'WW te." Figure _2_ Mean Rating Values of SLP 13 (Behaviour) for the "High Risk" and Control Students RESULTS / 69 exhibited 'failure in finishing things started' a little more frequent than occasionally, and that teachers are not rating general attention span on these two variables for "high risk" students, but instead are rating the specific skill areas presented on the Profile that are dependent on adequate attentional skills. This may well be true for the normally achieving group, however, the qualitative rating values between these two variables appear to be more consistent for this group than for the "high risk" group. One-way analysis of variance procedures conducted indicate the existence of highly significant (p< 0.000) differences between group means on these two variables. Highly significant differences in mean teacher ratings between normally achieving control and "high risk" kindergarten students include items SLP#2 (attention span and distractibility), 13A (fidgets), 13B (difficulty staying seated), 13D (easily distracted), 13G (difficulty listening), 131 (short attention span — failure to complete things started) and the Total Behaviour score for the SLP#13 subitems. 5. Hypothesis 5 There will be no significant (p<.05) difference between the normally achieving and "high risk" groups on the GDS measures of impulsivity (Delay Task — Total ER score) and sustained attention (Vigilance Task — Total Correct Responses and Total Commission Errors scores). RESULTS / 70 6. Hypothesis 6 There will be no significant (p<.05) difference between normally achieving and "high risk" groups on their Vigilance Task performance over time, (i.e., Blocks 1, 2, and 3). a. Part 1: GDS Delay Task The Total Efficiency Ratio (Total E.R.) score on the GDS Delay Task is the best single indicator of the level of impulsive behaviour exhibited by a subject (Gordon, 1986a). This score is representative of the percentage of times the button is pressed after an appropriate time delay, and is thus, arithmetically derived from the ratio of Total Correct Responses to Total Responses. One-way analysis of variance procedures indicate that the mean scores for these latter two variables suggest that no significant differences between normally achieving and "high risk" students exist for the Total Responses elicited, (i.e., total number of button presses) — 5 = 68.61, sd=21.23 and 5 = 71.40, sd=28.66, respectively. Significant mean score differences (F =14.11, MSe= 112.17, p<.05) do (1,56) appear, however, for the Total Correct Responses, (i.e., the child was able to refrain from button pressing for the appropriate 4 second intervals) — 5 = 44.79, sd= 11.74 and £ = 34.33, sd= 10.06, respectively. Statistical significance, (i.e., p£.05), between mean Total E.R. score differences for the two groups was not indicated through one-way analysis of variance procedures, (i.e., F =3.54, MSe = 0.04, p>.05). According to Gordon's (1,56) norms (Gordon, 1986a), the mean control Total E.R. score (5 = 0.69, sd=0.18) for the normally achieving control group falls within NORMAL range whereas that of the "high risk" group (5 = 0.59, sd = 0.24) falls within BORDERLINE range. RESULTS / 71 Table 1_ Repeated Measures Analysis of Variance (ANOVAR) Summary Table for GDS Delay Task Scores Item Source MSe F(l,56) Block Correct Responses Between Groups 28.08 14.11* Across Blocks 5.22 0.35 Group X Block 5.22 0.28 Block Responses Between Groups 160.63 0.18 Across Blocks 32.33 0.68 Group X Block 32.33 0.51 Block ER Between Groups 0.17 4.46 Across Blocks 0.02 0.74 Group X Block 0.02 0.77 *pS0.0167, with Bonferroni adjustment Additional data observations of interest on the Delay Task are students' performance over time. In determining significant mean score differences in performance over time between the two sample groups, repeated measures analysis of variance (ANOVAR) procedures were conducted on Block Correct, Block Responses and Block E.R. scores, (see Table 7) Mean Block scores for both Block Correct and Block Responses did not appear to vary much over the four Blocks, suggesting that the level of impulsive responding for each of the normally achieving and "high risk" groups did not change much over time, (see Table 3 and Figure 3) Mean Block scores were significantly different between the groups for Correct Responses while mean score differences for Block Responses were not. Mean Block E.R. scores, also, appeared relatively consistent over time, with no statistically significant mean score differences occuring between groups nor across RESULTS / 72 0.5-10. ao4 io I f ! — * — ABNORMAL to *0 JO 40 T NftUUL fo *> to So HO 30 Jo I it 3 M fioHMAJ-* « £ o . 6 5 - CONTROLS Figure _3_ Mean GDS Delay Task Performance Over Time for the  "High Risk" and Control Students RESULTS / 73 blocks. Additionally, no statistically significant interactive effect between group and block performances was achieved on any of the Block scores, (i.e.* Block Correct, Block Responses, or Block E.R.). b. Part 1: GDS Vigilance Task While the GDS Vigilance Task primarily makes demands on one's ability to focus and maintain attention over time, the subject must also be able to suppress any tendencies to respond impulsively. Analysis of variance (ANOVA) procedures reveal significant mean score differences between the two sample groups on several of the variables of this task, (see Table 8) According to Gordon (1986a), Total Correct Responses and Total Commission Errors scores are most indicative of impulse control and sustained attentional abilities. As anticipated, the normally achieving control group produced significantly more Total Correct Responses on average (x = 26.64, sd=2.64) than did the "high risk" group (x = 21.56, sd=5.62), with F =18.93, MSe= 19.71, (1,56) p<0.05. According to the GDS norms, these scores placed each group, respectively, into the NORMAL and BORDERLINE ranges. Thus, under conditions of relatively high arousal, the control students, on average, were able to press the button immediately following the appearance of a number "1" on the display screen significantly better than the "high risk" students. The GDS Percentile Table for five-year-olds (Gordon, 1986a) places the mean control group at the 16th percentile. Significant differences are also observable on mean Total Commission Errors scores between the groups. Total Commission Errors scores, with F =6.16, MSe = 317.23, and p<0.05). Once again, the mean control group (1,56) RESULTS / 74 Table _8 Repeated Measures Analysis of Variance (ANOVAR) Summary Table for GDS Vigilance Task Scores Item Source MSe F(l,56) Block Correct Responses Between Groups 6.57 18.94* Across Blocks 1.61 214.45* Group X Block 1.61 1.99 Block Omission Errors Between Groups 6.57 18.93* Across Blocks 1.61 4.73* Group X Block 1.61 1.99 Block Commission Errors Between Groups 106.30 5.29 Across Blocks 7.14 13.94* Group X Block 7.14 1.32 * p S 0.0167, with Bonferroni adjustment score lies within NORMAL range for five-year-olds (5 = 4.25, sd=5.87) whereas that of the "high risk" group lies well within the BORDERLINE range (5 = 15.87, sd=24.09), as per GDS norms (Gordon, 1986a). This difference in mean scores between the groups is statistically significant, with F =6.16, (1,56) MSe = 317.23, p<0.05. The high standard deviations obtained reflect widely variable individual performances within each group, (see Table 4) The GDS Percentile Tables for five-year-olds provided in the GDS manual suggest that the normally achieving kindergarten students, on average, performed better than approximately 66% of Gordon's normative population, whereas the average performance of teacher-nominated "high risk" students ranked close to the 90th percentile for total errors of commission. RESULTS / 75 c. Part 2: Sustained Attention Over Time Several observations of students' performance over time on the Vigilance Task can be made from the data analyses conducted, (see Tables 4 and 8) Repeated measures analysis of variance (ANOVAR) procedures were conducted on Block Correct Responses, Block Omission Errors and Block Commission Errors scores in examining the effect of time/repeated trials on students' performance and the possibility of an interactive effect existing between group and block performances. The Bonferroni adjustment of the p value was employed, (i.e., 0.05/3), as a means of reducing the possibility of making a Type I error. ANOVAR procedures indicate that the mean score differences between the sample groups on Block Correct Responses are significantly higher for the normally achieving students than for the "high risk" students. Examining individual Block scores, the mean number of correct responses made by each group is highest on the Block 1 trial and noticeably decreases with each 2-minute time block, (see Figure 4) This deterioration in mean performance scores for this variable is evident more so between Blocks 1 and 2 than between Blocks 2 and 3. ANOVAR procedures reveal that this deterioration in Block performance over time is statistically significant. Mean performance scores on Block Correct Responses suggest, therefore, that students from both sample groups were better able to appropriately push the button when the "hot" target stimulus, (i.e., number "1"), appeared on the screen within the first time block administration, (i.e., Block 1), more often than in each of the successive administrations, (i.e., Blocks 2 and 3), eliciting fewer correct responses with each successive time block. No significant interactive effect between group and block performances was RESULTS / 76 lo-l °4 t o * at io EM rtS 1 3o 20 to A » ' X « t M M . 2© io 2 0 BLOCKS ( Z - « m w * t . -*TC»VI) across t ( M L Figure _4_ Mean GDS Vigilance Task Performance Over Time of "High Risk" and Control Students RESULTS / .77 obtained for Block Correct Responses scores. Significant mean score differences in Block Omission Errors are evident (see Table 8), especially on Block 2. Examination of these scores between the. two groups reveals interesting findings. Standard deviations, in general, tend to be higher than the mean values over time for both groups, reflective of a fairly broad range of individual omission error scores within each group. Mean Block Omission Error scores reveal sizeable differences between the groups. While "high risk" students' mean scores reflect approximately three times the number of errors of omission on Blocks 1 and 3 than do those of the control group, the most sizeable discrepancy is seen on Block 2 on which it surpasses a four-fold difference. Moreover, no change in mean omission error scores is seen in the normally achieving control group from Blocks 1 to 2; however, a jump is observed from Blocks 2 to 3. The "high risk" group, conversely, displayed a considerable increase in mean score values between Blocks 1 and 2, and only a minimal increase between Blocks 2 and 3. Performance across Blocks was found to be statistically significant. No significant interactive effects between group and block, however, were obtained for this variable. Similar patterns in Block Commission Errors performances were exhibited by both groups with control mean values expectedly lower than those of the "high risk" group. These mean score differences between the groups, however, did not achieve statistical significance (using ANOVAR procedures). The number of commission errors made from block to block, however, was found to be statistically significant independent of sample group. Students within each group tended to make greater errors of commission on Block 1, with a sharp decrease evident on Block 2 and a minimal further decrease observed on Block 3. RESULTS / 78 Between groups, "high risk" students' made close to three times as many-commission errors on Block 1 than did controls, and in excess of quadruple the number of errors on Blocks 2 and 3. Once again, no statistically significant interactive effects between group and block were found. In summary, few significant differences in mean scores between groups were apparent on the GDS Delay Task variables. A significant difference between normally achieving and "high risk" teacher-nominated students was observed on the Total Correct Responses mean scores with the controls expectedly yielding more correct responses, suggestive of a better ability to refrain from impulsive responding than the "high risk" group. The Total Efficiency Ratio scores, purportedly the best quantitative GDS indicator of the degree of impulsive responding, produced insignificant mean score differences between the sample groups. Mean Correct Responses scores, elicited in each of the four time blocks, yielded significant differences between groups but not across Blocks, with no significant interactive effects between group and block performances. Insignificant mean performance differences on both Block Responses and Block E.R. scores, however, were obtained between groups and across blocks, with no significant interactive effects evident between group and block performances. Performance on the GDS Vigilance Task yielded statistically significant mean score differences between the sample groups for Total Correct Responses and Total Commission Errors scores. Significant changes in performance over time were evident in Block scores between groups and across blocks. No significant interactive effects, however, were evident between group and block performances. RESULTS / 79 7. Hypothesis 7 There will be no significant (p<.05) relationship (correlation) between the GDS measures of sustained attention (Vigilance Task — Total Correct Responses and Total Commission Errors scores) and impulsivity (Delay Task — Total Efficiency Ratio score) for the normally achieving and "high risk" groups. Examination of Pearson product-moment correlation coefficients suggests the existence of very weak to moderate relationships between sustained attention and impulsivity amongst the sample. The largest correlations obtained appear for both groups between the Total E.R. scores of the Delay Task and the Total Correct Responses scores of the Vigilance Task, (see Table 9) Table _9 Intercorrelations of Sustained Attention and Impulsivity Among Groups Vigilance Task Delay Task Group Total Correct Total Commission Variable Responses Errors Total E.R. 1. GDS Vigilance Task i) Total Correct Controls - -0.084 0.316 Responses High Risk - -0.071 0.519* ii) Total Commission Controls — — 0.245 Errors High Risk - - -0.059 *p<£0.05 While the coefficient for the control group is insignificant and relatively low (r=0.316), the "high risk" group's coefficient (r=0.510) denotes the presence of a statistically significant and relatively high correlation between the above two RESULTS / 80 mentioned variables. The difference between the strength of these two correlations was not significant using Fisher's z transformation of r (Glass & Hopkins, 1984). These data would suggest the presence of a significant and relatively strong relationship between levels of impulsive responding on the Delay Task and the ability to maintain attention on the Vigilance Task for the "high risk" group only. Correlation coefficients obtained for Total E.R. (Delay Task) and Total Commission Errors (Vigilance Task) scores suggest that there was really no statistically significant relationship between kindergarten students' levels of impulsive performance on the Delay Task and their ability to maintain attention and refrain from impulsive responding on the Vigilance Task, so as to niinimize the number of errors of commission made. Exploratory inspection of the nature of intercorrelations between the Delay and Vigilance Task performances reveals some interesting findings, (see Table E-l) Correlation coefficients for the normally achieving control group tended to be predominantly quite low and insignificant. The only noteworthy significant, but moderate, correlation coefficients, were obtained between the Delay Task E.R. Block Variability scores and (1) Block 1 Commission Errors, and (2) Total Commission Errors scores of the Vigilance Task. The "high risk" group, however, exhibited greater significant coefficients of higher value. The highest significant Pearson r coefficients for the "high risk" group were predominantly achieved in correlations of Block variables between the two GDS tasks. This was most apparent with coefficients obtained for Blocks 1, 2 and 3 E.R. scores of the Delay Task and Blocks 2 and 3 (1) Correct Responses, and (2) Omission Errors scores of the Vigilance Task. Amongst these significant correlations, the RESULTS / 81 coefficients were noticeably higher (range: ±0.536 to ±0.667) when Block E.R. scores were correlated with Block 2 Correct Responses and Omission Errors scores than when Block E.R. scores were correlated with Block 3 Correct Responses and Omission Errors scores (range: ±0.502 to ±0.573). The differences between the strengths of these correlations were not found to be significant using Fisher's z transformation of r (Glass & Hopkins, 1984). Yet further examination of the intercorrelational data yielded higher significant correlations on Blocks 1 and 3 E.R. scores when correlated to the above mentioned Block 2 Vigilance Task scores than on Blocks 2 and 4 E.R. scores. Additionally, Blocks 1, 2 and 3 E.R. scores exhibit a similar correlational pattern when correlated with Total Correct Responses scores of the Vigilance Task, (i.e., range: 0.563 to 0.651, pSO.OOl). Coefficients for Total E.R. were relatively high (range: 0.519 to ±0.572) and significant when correlated to only (1) Total Correct Responses, (2) Block 2 Correct Responses, and (3) Block 2 Omission Errors scores. In summary, intercorrelational examination of the nature of the relationship between kindergarten students' performances on the GDS Delay and Vigilance Tasks indicates the presence of very weak to relatively strong relationships for the sample groups. Of all variables intercorrelated for each group, the "high risk" group's Total Correct Responses performance (Vigilance Task), exclusively, suggests the presence of a significant and relatively strong relationship with Total Efficiency Ratio performance (Delay Task). Correlational data on students' impulsivity and sustained attentional skills over time, (i.e., Block variables), revealed few significant relationships for normally achieving control subjects. "High risk" students, on the other hand, yielded a greater number of significant and stronger correlations in impulsivity RESULTS / 82 and vigilance performances over time, particularly evident in correlations of Delay Task Block E.R. scores with Vigilance Task Blocks 2 and 3 Correct Responses and Omission Errors scores. 8. Hypothesis 8 There will be no significant (p<.05) difference between boys and girls on their inattentiveness as displayed by their test scores on the GDS Vigilance Task. General inspection of all of the data analyses reveals some differences in performances/ratings amongst the sexes, (see Tables E-2, E-3, and E-4) Teacher ratings on the School Learning Profile indicate that normally achieving control boys were assigned generally lower qualitative ratings than normally achieving control girls. Mean score differences on rating values between the control group sexes were not statistically significant on any of the SLP variables, as demonstrated through ANOVA procedures. Standard deviations were generally moderate to high, with respect to the obtained mean scores, being noticeably higher for the control boys than the girls. This would suggest that there was a greater consistency in teacher ratings of control girls compared with the control boys. "High risk" students' mean ratings scores indicate that boys and girls received predominantly similar qualitative rating values on the SLP with the exception of item #13E (defiant and uncooperative). Teachers appeared to identify "high risk" boys as exhibiting noticeably more defiant and uncooperative behaviour than did the girls, however, these mean score differences were not found to be statistically significant. Although boys received twice the numerical rating value RESULTS / 83 as did the girls, (i.e., x = 1.105, sd= 1.449 and x = 0.545, sd = 0.522, respectively), the behavioural rating equivalents are not substantially different, (i.e., "Never" to "Rare"). Compared to the normally achieving control group sexes' ratings, boys also received higher numerical ratings (close to three times higher), yet still not achieving statistical significance, than did girls, (i.e., x = 0.400, sd=0.737 and x = 0.154, sd = 0.375, respectively), however, their behavioural rating equivalents approximated a "Never" value. Standard deviations for the "high risk" boys and girls tended to approximate each other. a. GDS Delay Task No significant mean score differences between the sexes were obtained for either the normally achieving or "high risk" groups even though mean scores between the sexes tended to approximate one another quite consistently. Within each group, neither sex scored consistently higher or lower on any of the test variables. Standard deviations for both boys and girls amongst the two sample groups tended to be generally low with respect to variable means. Noteworthy, however, is that standard deviations for both normally achieving and "high risk" girls are generally higher than for the boys in each sample group. This is particularly noticeable on the following Delay Task variables: Total Correct Responses, Total Responses for both groups, and additionally on Block 2 Correct and Total Responses, Blocks 3 and 4 Total Responses scores for the normally achieving group only. Compared to the mean scores of normally achieving control boys and girls on Total Correct Responses, substantially less disparity between the sexes in mean scores was seen for control students than for "high risk" students. This RESULTS / 84 difference, however, did not achieve statistical significance, using ANOVA procedures. Interestingly enough, however, the Total Responses mean scores approximated each other among all the students, regardless of sex or group, with the differences in mean scores being insignificant. Analysis of variance procedures do not provide evidence of a significant two-way interactive effect of sex and group performances. Less disparity between boys and girls on Block Correct and Block Responses mean performances was evident for the control group. Although the mean performance scores on these two variables for the "high risk" group suggest that this group tended to make somewhat greater numbers of responses within each block than the controls, the "high risk" students produced fewer correct responses than the controls. Within the "high risk" group, boys generated noticeably more correct responses than girls over all four blocks. Repeated measures analysis of variance (ANOVAR) was conducted on Block Correct Responses, Block Responses, and Block E.R. scores in examining the effect of repeating trials on students' performances. The Bonferroni adjustment of the p value was used, (i.e., 0.05/3 = 0.0167). No significant mean score differences for any of the above mentioned scores were obtained for the sexes and the groups. Additionally, no significant interactive effects were evident between (1) sex and group; (2) sex and blocks; and (3) sex, group and blocks. The calculated Total Efficiency Ratios, purportedly the best single indicator of impulsive responding (Gordon, 1986a), yielded mean score differences between the groups that were not significant, (i.e., p>0.05). According to the GDS norms for five-year-olds, mean scores on this variable for both control boys and girls are within NORMAL limits for impulsive responding. "High risk" boys and girls RESULTS / 85 generated mean Total E.R. scores at the upper end of the BORDERLINE range. 6. GDS Vigilance Task Several general observations of the data analyses for this task are apparent. Significant mean score differences between the sexes on all variables are more prominent for the "high risk" group than for the control group, with the girls tending to achieve better qualitative scores than boys within both groups. Standard deviations tend to approximate one another between the sexes more so for the normally achieving control subjects than for the "high risk" subjects. "High risk" students' lower mean scores reflect poorer overall performance, for both boys and girls, on this vigilance task than do those of the controls. While insignificant mean score differences between control boys and girls are apparent on the variable Total Correct Responses, significant differences arise between the "high risk" sexes with mean scores favouring the girls' performance, (i.e., boys' 5 = 20.474, sd = 5.853; girls' 5 = 23.454, sd=4.865). Analysis of variance procedures indicate mean score differences on Total Correct Responses between the normally achieving and "high risk" groups, however, as being highly statistically significant. In examining boys' and girls' ability to sustain attention over time, (i.e., Block scores), repeated measures analysis of variance procedures were conducted in determining significance of mean score differences. The Bonferroni adjustment of the p value was employed, (i.e., 0.05/3 = 0.0167). As discussed earlier, mean score differences on Block Correct Responses were found to be significant between normally achieving and "high risk" groups. RESULTS / 86 Mean Block Correct score differences between boys and girls in this study's sample were not found to be statistically significant, ^ = 0.89, MSe = 6.41, p> 0.0167. Additionally, no significant interactive effect between group and sex was evident over all three blocks on this variable. Mean performances for boys and girls between the blocks did achieve significant differences in the number of correct responses elicited from the first block to the last one, F, =199.07, (2,108) MSe=1.66, p<0.0167. Once again, no evidence of significant interactive effects between (1) block and sex; (2) group and block; nor (3) sex, group and block were achieved. Examination of the performance patterns over blocks reveals that the normally achieving control group mean scores are consistently marginally higher for the boys than for the girls. Although the "high risk" group yielded mean scores significantly lower over time on this variable, mean score differences were also insignificant for this sample group between boys and girls, with scores somewhat favouring the girls' performance. Examining the quality of performance over time, it is evident that a similar degenerative pattern in responding correctly to the criteria of the task exists between both sexes within each group. The greatest drop in Correct Responses scores occurs between Blocks 1 and 2 with a substantially smaller decrease apparent between Blocks 2 and 3. ANOVAR procedures on Block Omission Errors scores indicate that the "high risk" students produced significantly greater mean differences in errors than did the normally achieving controls over the three blocks. Insignificant differences between boys' and girls' performances, however, were observed, F =0.89, (1,54) MSe=6.41, p>0.0167. In addition, statistical significance was not achieved for mean score differences from one block to the next for the sample, RESULTS / 87 F =4.17, MSe=1.66, p>0.0167. No significant interactive effects between (2,108) (1) sex and group; (2) sex and block: (3) group and block; nor (4) sex, group and block were achieved. Examination of performance patterns over trials indicates that control boys tended to produce slightly fewer omission errors, (i.e., not pressing the button immediately after the number "1" appeared on the digital display screen). For the "high risk" group, mean block score differences on this variable between the sexes were more apparent, with boys consistently eliciting greater omission errors than girls. Over time, both boys and girls within each group elicited mean scores suggestive of a generally steady increase in the production of omission errors from Block to Block. Noteworthy, also, standard deviations tended to be fairly high on the obtained mean scores for this variable, regardless of sex, being more obvious within the control group. The most significant differences in vigilance performance, both among the groups and among the sexes, were evident on Block Commission Errors and Total Commission Errors scores. Standard deviations were very high on all of these variables, (i.e., greater than their mean score values), suggestive of a high degree of variance in individual performances within each group. As the mean scores for the sexes were the greatest amongst both sample groups on the initial task trial, (i.e., Block 1), all students had difficulty controlling impulsive responding, which generated errors of commission upon initial exposure to this task. ANOVAR results suggest that significant mean score differences in Block Commission Errors between boys and girls arose only between blocks, with F =12.46, MSe = 7.12, p<0.0167. Significant differences between boys and . (2,108) girls participating ' in this study, regardless of group, were not achieved, with RESULTS / 88 F (154) = 1 ' 6 7 ' M S e = 1 0 4 ' 6 2 > P > 0 - 0 1 6 7 - Additionally, significant differences between the groups were also not achieved, with F =4.38, MSe= 104.62, p>0.0167. 6 V (1,54) Interactive effects between (1) sex and group; (2) sex and block; (3) group and block; and (4) sex, group and block did not achieve statistical significance. Little difference in mean Block Commission Errors scores for the normally achieving control group sexes is observed on Block 1, with control boys scoring slightly higher, but not statistically significant, than the girls. Blocks 2 and 3 mean Commission Errors scores reveal a dissimilar performance pattern occurring over time between control boys and girls. Boys' mean scores decreased dramatically on Block 2, (i.e., from 2.600 on Block 1 to 0.600 on Block 2), however, their performance noticeably improved on Block 3 (x = 1.467, sd=2.560), yielding only half the number of errors compared with the initial trial. The control girls yielded mean Commission Error scores that steadily diminished over time, with the greatest decrease exhibited on Block 3 on which the mean score value dropped by a factor of nine. Thus, it would appear from these results that control girls' ability to sustain attention to this task and refrain from impulsive responding gradually improved with time, being especially apparent on the third trial. While control boys demonstrated an overall reduction in the production of commission errors over time, their mean scores reflect somewhat of a deterioration in sustaining attention and refraining from impulsive responding on the final trial, (i.e., Block 3), on which the mean score more than doubled. Conversely, the "high risk" group displayed sizeable mean Block Commission Errors score differences between the sexes on Block 1 with the boys' scores being more than double that of the girls, (i.e., x = 8.895, sd= 10.322 and x = 4.273, sd = 4.519, respectively). Block 2 mean scores reflect a substantial RESULTS / 89 reduction in the production of commission errors for both sexes, with boys' scores being three and a half times greater than those of the girls. Only a marginal further reduction is seen for the "high risk" boys in Block 3. The girls' Block 3 mean scores, on the other hand, increased, yet still suggestive of a better ability to sustain attention and control impulsive responding on this task, (i.e., approximately two-fold compared with boys). Comparisons of mean Commission Errors score differences of the sexes between the sample groups indicates that while "high risk" boys' performance diminishes steadily over time, that of the control boys' appears to be more erratic with an initial improvement observed (on Block 2) followed by somewhat of a deterioration exhibited on the final trial, (i.e., Block 3). Moreover, mean scores on this variable reveal that "high risk" boys produced approximately (1) 3.5 times more commission errors than both the control boys and girls on Block 1 and twice as many as "high risk" girls; (2) 16 times more than control boys, 4 times more than each of control and "high risk" girls on Block 2; and (3) 3.5 times more than control boys, 34 times more than control girls, and twice as many as "high risk" girls. The girls' mean scores, conversely, reveal less of a performance discrepancy between the two sample groups. The control girls' mean scores indicated a steady improvement in performance over time, (i.e., mean scores decreased between time blocks), while a more erratic performance pattern, similar to that of the control boys, was observed for the "high risk" girls. Mean Commission Errors scores between the two sample groups indicate that "high risk" girls elicited approximately (1) twice as many errors of commission on Block 1 as did the control girls; (2) slightly more on Block 2; and (3) 17 times RESULTS / 90 the number of errors on Block 3. "High risk" girls' ability to sustain attention and refrain from impulsive responding over time, therefore, deteriorated dramatically on the last trial. Examination of mean Total Commission Errors scores for the Vigilance Task does not appear to reflect the extreme changes in performances between the sexes over time. Control boys obtained a mean score slightly higher (5 = 4.667, sd=6.366) than that of the control girls (1 = 3.769, sd=5.464), and, compared with the "high risk" subjects, approximately 4.5 times less that of the boys (5 = 80.158, sd=28.851) and just less than half that of the girls (5 = 8.454, sd=9.470). Using analysis of variance procedures, these mean score differences between only the groups were found to achieve statistical significance, with F =5.58, MSe = 345.95, p<0.05. Significant performance differences between (1,54) F the sexes were not obtained. Over the full duration of the Vigilance Task, "high risk" boys were observed to produce substantially greater errors of commission than any of the other subject groupings, (i.e., control boys, control girls and "high risk" girls). Additionally, no significant two-way interactive effect between sex and group was obtained on Total Commission Errors scores. 9. Hypothesis 9 There will be no significant (p<.05) relationship (correlation) between test scores of impulsivity (GDS) and visual memory (Bead Memory subtest — Stanford-Binet Intelligence Scale: Fourth Edition) for normally achieving and "high risk" kindergarten students. RESULTS / 91 10. Hypothesis 10 There will be no significant (p<.05) relationship (correlation) between test scores of sustained attention (GDS) and visual memory (Bead Memory subtest — Stanford-Binet Intelligence Scale: Fourth Edition) for normally achieving and "high risk" kindergarten students. For both sample groups, very low Pearson correlation coefficients were predominantly obtained when Bead Memory performance was correlated with the GDS Delay Task and Vigilance Task performances, (see Table E-l) Moreover, the majority of coefficients amongst both the normally achieving and "high risk" subjects yielded negative values, implicit of an apparent tendency for inverse relationships to exist between visual memory and impulsivity/vigilance performances on the assessment measures used. Within the normally achieving control group, a single statistically significant correlation having a moderate value, (i.e., r=0.423, p=0.01), was obtained between Block 1 E.R. (Efficiency Ratio) scores of the Delay Task and Bead Memory. Although a positive correlation between these two variables was obtained for the "high risk" group as well, the coefficient was very low (r=0.168) and statistically insignificant. While two other Delay Task variables (Block 1 Responses and Total Efficiency Ratio) yielded correlations with Bead Memory that approached significance, the coefficient values were low, (i.e., r=-0.295 and 0.272, respectively), and statistically insignificant. Examining the correlations between the Delay Task variables and Bead Memory for the "high risk" group reveals only one statistically significant coefficient (r=-0.442) of moderate value for Block 4 Correct Responses and Bead RESULTS / 92 Memory. No statistically significant correlations were obtained between any of the Vigilance Task variables and Bead Memory. In summary, the results do not appear to suggest generally significant relationships between visual memory and impulsivity/vigilance performances for either of the normally achieving or "high risk" groups.' C. S U M M A R Y The main effects obtained in this study are summarized below. Supported Findings 1. Data supported the presence of significant relationships between teacher ratings of school readiness (overall ability to learn school materials) for the normally achieving group, with 'attention span and distractibility, as related to following classroom directions' and 'short attention span, as related to failing to finish things started' exhibiting the strongest relationships to school readiness. 2. Predominantly significant differences between the sample groups were achieved on teacher ratings of attention and school readiness, with the normal achievers receiving higher qualitative ratings on displayed behaviour than the "high risk" students. 3. Significant differences between normally achieving and "high risk" groups were evident in their abilities to sustain attention and control impulsive responding on the GDS Vigilance Task, with better performance favoring the normal achievers. RESULTS / 93 4. Data supported the presence of significant differences between the groups in their performance on the GDS Vigilance Task over time. While the "high risk" group made significantly fewer correct responses and greater omission errors from Block to Block than the controls, significant differences in Block errors of commission were not obtained between the groups. A significant deterioration in performance was observed over time on the number of correct responses and errors of omission made by both groups. A significant improvement over time, however, was observed in the decreased number of commission errors made by both groups. 5. Significant relationships between sustained attention and impulsive performances on the GDS tasks were found primarily for the "high risk" group. 6. Significant differences in vigilant performances between boys and girls were more prominent among the "high risk" students than among the normal achievers, with girls tending to achieve better qualitative scores than boys within both groups. Unsupported Findings 1. No support was found for the existence of significant relationships (correlations) between test scores of impulsivity (GDS) and teacher ratings (SLP) for the normally achieving and "high risk" students. Although significant relationships between teacher ratings of attentional skills and students' vigilant performances on the GDS were not evident amongst either the normally achieving nor "high risk" groups for the Total Correct RESULTS / 94 Responses scores, the number of commission errors were found to significantly correlate with*, only a few of the SLP items, and only for the "high risk" group. 2. Significant differences in levels of impulsive responding displayed on the GDS Delay Task were not found between the normally achieving and "high risk" groups. 3. No support was found for the presence of significant relationships between visual memory performance and sustained attention and impulse control abilities for either of the sample groups. V. DISCUSSION A N D CONCLUSIONS . Early identification of educationally at-risk preschool/kindergarten children can serve as a preventive measure leading to intervention so that the likelihood of later academic failure can be minimized. Although school readiness tests have long been utilized in early screening of learning problems, no standardized screening methods/psychometric instruments have as yet emerged as adequate measures of school readiness skills. Incresed rates of attention deficits, hyperactive behaviour and impulsivity have frequently accompanied school difficulties (Richards, Samuels, Turnure and Ysseldyke, 1990). Amongst children diagnosed as ADHD, between 25 and 60 percent are likely to experience learning problems (Barkley, 1981). Precursors of attentional problems have been shown to be identifiable in a large number of the preschool/kindergarten population (Palfrey et al., 1985). The importance of establishing reliable early detection programs cannot be underscored as these programs would identify those children in need of receiving appropriate remedial services. Additionally, early identification procedures can assist educators in better comprehending a child's learning style so that individualized educational programs and/or customized curriculum materials could be developed in order to accommodate a child's poor self-monitoring, impulsive and/or distractible tendencies. Prior to presenting a discussion of the results, it is important to realize that how attentive a child appears may not be necessarily related to his/her actual ability to attend. Gordon, DiNiro, and Mettelman (1989) contend that: . . . just because a child may appear inattentive does not necessarily 95 DISCUSSION AND CONCLUSIONS / 96 indicate that the child actually suffers from deficits in attention. Some children are physically active and inattentive because of anxiety, learning disorders, or conduct problems, yet still perform successfully on measures of attention and, under appropriate conditions, on academic material. Conversely, compliant, seemingly attentive youngsters can exhibit clear deficits on tasks that require attention, (p. 142) Thus, assessment approaches to attention deficits should not rely solely upon a single measure, but instead should encompass a wide variety of information obtained through multiple assessment sources. A. SUMMARY The purposes of this exploratory study were to investigate the similarities and differences between two assessment measures — the Kindergarten School Learning Profile teacher ratings and the Gordon Diagnostic System — in identifying children who would likely be at-risk for experiencing school failure as a result of attentional/impulse control deficits displayed in kindergarten. As attentional skills are believed to influence memory, visual memory was also investigated in relation to attention and impulse control. Twenty-eight teacher-nominated "high risk" kindergarten students were identified as functioning within the lowest 10% for overall school readiness through completion of a devised teacher rating scale — the Kindergarten School Learning Profile (SLP) (Carter & Conry, 1988). Following this, 30 control students were obtained through computerized systematic random selection, with SLPs also completed on each one of them. A greater disparity between the sexes selected for participation in this study appeared within the "high risk" group than for the controls, favouring a greater number of boys than girls within both groups, however, more noticeably within the former group. DISCUSSION AND CONCLUSIONS / 97 Measures of impulse control and sustained attentional skills for all students were obtained through the preschool setting of the Delay and Vigilance Tasks, respectively, of the Gordon Diagnostic System (GDS). Individual performance consistency over time was examined through several trial, or Block, presentations of each task. In general, performance in both impulse control and attentional skills deteriorated with time. Visual memory abilities were examined through use of the Bead Memory subtest of the Stanford-Binet Intelligence Scale: Fourth Edition. Means, standard deviations, Pearson product-moment correlations and one-way analyses of variance were conducted in the analysis of both the subjective teacher ratings and the objective standardized test results. Repeated measures analysis of variance procedures were used on the GDS Block measures. Based on the proposed research questions presented in chapter 1, a summarization and discussion of the obtained results of this study are forthcoming. B. DISCUSSION OF RESULTS 1. Teacher Ratings of School Readiness and Attention a. Summary A brief outline of obtained significant differences on mean teacher ratings between the groups on this measure is summarized below. Highly Significant (p< 0.0038) • overall learning ability/school readiness DISCUSSION AND CONCLUSIONS / 98 • attention span/distractibility — following instructions in class • fidgets • difficulty staying seated • easily distracted • difficulty listening • short attention span — fails to finish things started • SLP total behaviour score (item 13) Insignificant (p>0.0038) • temper tantrums (explosive and unpredictable) • blurts out answers to questions before they have been completed • difficulty waiting turns in games or group activities • defiant and uncooperative • difficulty playing quietly b. Discussion Poorer teacher-rated behaviour, as related to attention and impulse control abilities, exhibited both within and outside of the classroom situation, significantly differentiated the normally achieving control and "high risk" groups. "High risk" students were rated as displaying problematic behaviours on considerably more variables of the School Learning Profile than were the controls. While it is difficult to ascertain the extent to which attentional/behavioural criteria represented by the subitems of SLP#13 truly evaluated attention and impulse control abilities in "normal" behaviour, some of these subitems would appear to relate to one ability more so than to the other. In this study, the greatest mean score differences, achieving the highest degrees of significance, between the two sample DISCUSSION AND CONCLUSIONS / 99 groups were apparent on those subitems appearing to pertain to attentional abilities, (i.e., difficulty listening, attention span/distractibility as related to following classroom instructions, and short attention span as related to failing to complete things started) more so than on those subitems relying on intact impulse control abilities, (i.e., difficulty waiting turn, difficulty playing quietly, fidgeting, difficulty staying seated, and blurting out answers to questions before they have been completed). Of the SLP#13 subitems, teachers assigned the highest ratings for both the normally achieving control and "high risk" groups to 13D (easily distracted), approximating qualitative rating values of "Rare" and "Frequent", respectively. It would appear, therefore, that distractibility, as understood by teachers, is a behaviour most commonly exhibited both within and outside of the classroom situation by all students participating in the study. It could be argued that teachers' expectations of appropriate classroom behaviour (Rist, 1970) could be a factor of these obtained results and that participating teachers were less tolerant (Rubin & Balow, 1978) of distractible behaviours, assigning higher ratings on this variable indicative, in this case, of more frequently displayed behaviour. According to Cooper and Farran (1988), kindergarten teachers can tolerate disobedience and poor peer interactions more than behaviours associated with being off-task and inattentive to routines, instructions and completion of assignments. Teacher-rating results obtained in the present study generally support Cooper and Farran's findings. The lowest qualitative behavioural ratings ('Never') for both sample groups arose on subitems 13E (defiant and uncooperative) and 13F (temper tantrums), on which a significant difference between the groups in mean rating scores was' not achieved on either of these subitems. Teachers DISCUSSION AND CONCLUSIONS / 100 additionally considered "high risk" students to exhibit significantly greater frequencies of fidgeting behaviour, difficulties listening and staying seated, and a shorter attention span (related to failure in completing things started) compared to control students. In Day and Peters' investigation (1989) of cognitive and behavioural variables that differentiate normally achieving and underachieving Grade 3 and 4 students, teacher ratings of inattentive-passive classroom behaviour were powerful correlates of academic achievement. This finding was also evident in the current study. The highest and most significant correlation between teacher-rated overall learning ability (SLP#1) and behaviours involving intact attention and impulse control abilities (remaining SLP items used in this study) was apparent on SLP#2 (attention span and distractibility — following classroom instructions) for the control group. (Note: Correlation coefficients could not be computed for the "high risk" group on these variables due to the lack of variance in SLP#1 ratings, as defined by the study.) Of the behavioural ratings of item #13, subitem 131 (short attention span — failing to finish things started) achieved the highest correlation, being most significant, with SLP #1. Although possessing a negative value, the coefficient represented the existence of a relatively strong relationship between teacher ratings of overall learning ability and short attention span, (i.e., the higher the mean control group rating for overall learning ability [approximating the upper 30%, but not the highest 10% of the population on the normal curve], the less frequently they were considered to display difficultites in completing things started due to a short attention span). Correlations of this magnitude, (i.e., r&0.50), have been known to demonstrate good predictive efficiency with many screening' devices, correctly identifying approximately 70 to DISCUSSION AND CONCLUSIONS / 101 80 percent of kindergarten children who later perform at or near the bottom of their class in the elementary grades (Mercer, Algozzine & Trifiletti, 1979; Simner, 1982 [as cited in Simner, 1983]). While the intent of this study was not to examine later school failure, it is of interest to note that the results obtained do indicate the presence of significant and strong enough correlations of overall learning ability with one specific warning sign considered to be one of the best overall in-class indicators of future academic success (Simner, 1983), namely, in-class attention span/distractibility. Mean teacher ratings of control boys and girls on the SLP items for both the normally achieving and "high risk" groups revealed insignificant differences between the sexes. Between the sample groups, "high risk" boys and girls were found to be rated as exhibiting more problematic behaviours associated with attention and impulse control abilities than control boys and girls, not to a statistically significant degree. Intercorrelations of SLP#2 and SLP#13 subitems revealed the presence of many highly significant and strongly correlated relationships for both groups. Teacher ratings of attention span and distractibility, as related to displayed difficulty in following classroom directions, correlated with their ratings of exhibited problematic behaviours, associated with attention and impulse control abilities, somewhat stronger, in general, for the normally achieving students than for the "high risk" group. Of both groups, the strongest evident relationship in teacher ratings appeared between 'attention span/distractibility' and 'difficulty listening'. This is not too surprising as the ability to follow directions/instructions given by a teacher in the classroom would be highly dependent on a student's DISCUSSION AND CONCLUSIONS / 102 intact listening skills, which require adequate sustained attentional skills with minimal attention paid to ongoing distractions. Considering, on the other hand, that the unknown discriminative power of the SLP items limits interpretations of such obtained relationships, teachers may well have been identifying the behavioural aspects of 'attention span/distractibility, as related to following classroom directions' as being the same as 'difficulty listening'. If this, indeed, was the case, then it would not be surprising to find such a high correlation in the ratings of these two variables. The correlational results also indicated strong and highly significant relationships existing for both control and "high risk" students between SLP#2 ratings and teacher ratings on students being easily distracted, and having a short attention span related to failure to complete things started. While there generally appeared to be more consistent qualitative ratings for the control group between SLP#2 and the SLP#13 subitems (attributed to generally high Pearson r values on most subitems), teacher ratings for the "high risk" group showed less consistency, (i.e., lower r values). For both groups, teachers demonstrated least consistent ratings on SLP#2 and temper tantrums and difficulty playing quietly, which would lend support to Cooper and Farran's findings (1988) that kindergarten teachers have more tolerance for disobedience and poor peer relationships than they do for inattentiveness, off-task behaviours and incompletion of assignments. Interestingly, defiance and uncooperative behaviour ratings for the control students showed a relatively strong relationship with ratings of attention span/distractibility, however, a somewhat weak relationship between ratings of these same two variables was evident for the "high risk" group. These differences may be attributed to teachers finding "high risk" students DISCUSSION AND CONCLUSIONS / 103 demonstrated greater variability, (i.e., less consistency as a group), in such behaviour than control students, (i.e., defiant and uncooperative behaviour and difficulties with attention span/distractibility being seldom displayed in this group as a whole). In summary, teacher ratings on all but two of the SLP variables, namely, 'temper tantrums' and 'blurts out answers to questions before they have been completed', significantly differentiated the "high risk" and control students. The most commonly exhibited problematic behaviour identified by teachers for both sample groups was that students became easily distracted. These results support Cooper and Farran's contention (1988) that kindergarten teachers demonstrate greater tolerance for disobedient behaviour and poor peer interactions than they do for off-task behaviours, inattentiveness, and incompletion of assignments. The two SLP variables most highly correlated with overall learning ability (SLP#1) for the control group were, in order, SLP#2 (attention span/distractibility — following directions in class) and SLP#13I (short attention span — fails to complete things started), both yielding Pearson r values considered to be most effective predictors of later academic success/failure (Mercer et al., 1979). Teachers' ratings of impulse control behaviours on the SLP#13 subitems correlated most strongly and significantly with their ratings of attention/distractibility (SLP#2). This was somewhat more evident for the normally achieving control group than for the "high risk" group, suggestive of greater rating consistency for the former group. Between the sexes within each sample group, insignificant (p> 0.0038) mean rating score differences were obtained on all of the SLP items amongst the control boys and girls, and for 'defiant and uncooperative' behaviour amongst the "high risk" boys and girls. Between the groups, "high risk" boys and girls were DISCUSSION AND CONCLUSIONS / 104 rated qualitatively lower than control boys and girls, respectively, suggesting that gender has little influence between the sample groups on the incidence of exhibited problematic behaviours associated with attenion and impulse control abilities. 2. The Gordon Diagnostic System a. Delay Task Summary A brief outline of obtained significant differences on mean Delay Task scores between the groups is summarized below. Significant • Total Correct Responses (p^O.05) • Block Correct Responses (pS0.0167) Insignificant (p>0.05) • Total Efficiency Ratio • Total Responses • E.R. Block Variability (p> 0.0167) • Block Responses • Across block score differences on Block Correct, Block Responses and Block E.R. • Interactive effects between group and block on Block Correct, Block Responses and Block E.R. DISCUSSION AND CONCLUSIONS / 105 b. Delay Task Discussion The GDS Delay Task is purportedly a measure of one's ability to suppress or delay impulsive behavioural responding (Gordon, 1986a). Reinforcement is obtained when responding only after a sufficient minimum time interval has elapsed, unknown to the examinee. Three summary scores useful in performance interpretation are number of correct responses, number of responses, and the ratio between these two scores, (i.e., Total Efficiency Ratio). The obtained results on this Task in the present study revealed few significant performance differences between the "high risk" and normally achieving groups. As the Total Efficiency Ratio (E.R.) is the best single indicator of the level of exhibited impulsivity, obtained scores on this variable provide an important overview of degree of impulsivity displayed by each sample group. According to Gordon's statistically derived Threshold Tables (1986a), the control and "high risk" groups yielded performances within the NORMAL (37th percentile) and BORDERLINE (21st percentile) ranges, respectively. Statistical significance was not achieved on these mean score differences between the groups suggesting that "high risk" students, on average, were no more impulsive than were the controls. This finding contradicts Gordon's earlier results (1979) in which highly significant differences in Total E.R, scores were obtained for hyperactive and nonhyperactive groups of 6- to 8-year old boys. Differences in group selection criteria for these two studies, (i.e., presence or absence of hyperactive behaviour [Gordon's study] vs. rated as placing below or above the lower 10th percentile for overall learning ability [present study]), may be a factor in these contradictory findings. Additionally, the results of the current study may well be attributed to DISCUSSION AND CONCLUSIONS / 106 the mean Total E.R. scores for both groups closely approximating the cut-off point between the NORMAL and BORDERLINE ranges. Although Gordon (1986a) identifies an obtained BORDERLINE score on either of the GDS Tasks as a sufficient diagnostic criterion for ADD/Hyperactivity (adapted from Barkley, 1981), the obtained mean Total E.R, score for the "high risk" group too closely approximated the cut-off point between NORMAL and BORDERLINE ranges to unequivocally conclude that this group, on average, was truly a representative sample of ADD/Hyperactivity students. Examination of individual performances would be a more reliable and valid means of determining the diagnostic composition of the "high risk" group. Gordon (1986c) states that " . . . most nbn-ADD children can achieve an E.R. score of at least 0.72 [42nd percentile] . . . " (p. 15), with the lower limit for NORMAL range being 0.62, and " . . . children with an impulsive style often produce scores of 0.42 or lower" (p. 15), with 0.42 being the upper limit for the ABNORMAL performance range. Control and "high risk" groups' mean Total E.R. scores (5 = 0.69 and 5 = 0.59, respectively) suggest that while neither group's mean performance was indicative of an unequivocally displayed impulsive response style on this task, tendencies toward impulsive responding were suggested for both of these groups, and expectedly being somewhat more apparent for the "high risk" group. While the Total Responses score represents a child's overall response rate on the Task, it is primarily significant in ascertaining the validity of the Total E.R. score. Mean score differences between the groups approximated one another and were not found to be statistically significant. These results reflect an acceptable rate in attempting to score rewards without demonstrating excessive DISCUSSION AND CONCLUSIONS / 107 impulsive and inaccurate responding for both groups. Even though significant mean score differences on Total Responses were not obtained between the groups, the "high risk" children made fewer statistically significant mean correct responses over all four Blocks than the control children. Examining the progression of performance over time is important in detecting any changes in response patterns which would reflect a tendency toward deterioration in the ability to control impulsive responding. While statistically significant differences between normally achieving and "high risk" students arose on their abilities to produce correct responses from block to block, (i.e., Block Correct scores), mean scores for each group appeared fairly consistent over time. Thus, although scores generally increased somewhat between blocks, they did not do so to a significant degree. Although "high risk" students yielded fewer correct responses and showed greater performance consistency over time compared to the controls, both groups demonstrated an approximately equal response rate within each 90-second trial and also minimal deterioration in their ability to control impulsive responding over time. As in Gordon's study (1979) with hyperactive and nonhyperactive boys, the most noteable improvement in Correct Responses scores produced in the current study, although small compared to Gordon's results, arose on the second Block trial for both sample groups. "High risk" students may well have required the first block trial to adapt to the task while control students were better able at locking into a relatively efficient response pattern already on the first trial. While the control group yielded fairly consistent mean Block E.R. scores over time, the "high risk" group demonstrated the greatest deterioration in performance between Blocks 1 and 2 after which further but mild deterioration DISCUSSION AND CONCLUSIONS / 108 occurred, supported by Block E.R. scores for this group decreasing the most dramatically between these two trials. As statistical significance between the sample groups on mean Block E.R. scores was not achieved, prolonged time on task did not statistically differentiate the control and "high risk" groups. The "high risk" students may have fatigued easier, lost interest more in the task, become less attentive or more distracted/impulsive with time, especially in the latter two trials of the Task, than the control group. As no significant interactive effects between group and block were obtained, it would appear that block performance was not influenced by overall learning ability, (i.e., criterion differentiating normal achievers from "high risk" achievers in this study). The E.R. Block Variability scores, indicative of the overall degree of consistent responding over all four Blocks, did not reveal statistically significant mean score differences between the normally achieving control and "high risk" groups, even though scores within the NORMAL and BORDERLINE ranges were achieved for these two groups, respectively. Thus, insignificant differences in the degree of variability in performances over time between the groups, (i.e., very low standard deviations), suggest both groups displayed approximately equal degrees of consistency in delaying or inhibiting impulsive responding, similar to same-aged peers. Interestingly, both groups' mean scores on this variable closely approximated the cut-off point between NORMAL and BORDERLINE ranges. This may well suggest that the students participating in this study, although demonstrating normal variability in performance, also showed a tendency toward some inconsistency in delaying responding over time and, therefore, some general performance deterioration. Low and NORMAL Slope scores, depicting an Even Slope in performance DISCUSSION AND CONCLUSIONS / 109 from Block to Block for both groups, support the E.R. Block Variability results and reflect generally consistent performance, (i.e., no clear trends toward improvement nor deterioration), over time. The "high risk" group's Slope score closely approximated the cut-off point between the NORMAL and BORDERLINE ranges, as it did with the mean E.R. Block Variability score. Noteworthy, however, is that the large standard deviation values obtained for both groups suggest great variability in individual performances over time. Thus, once again, these results lack clarity and interpretations need to be made with caution. Coupled with other Block score data obtained on the Delay Task, however, it is reasonable to postulate that the "high risk" group demonstrated tendencies toward performance deterioration over time. While the control boys and girls obtained better scores on approximately half of the Delay Task variables compared with "high risk" boys and girls, no significant differences in impulse control abilities were indicated for boys and girls within each group nor between the groups, as per obtained Total E.R. scores. Additionally, no significant differences between the sexes within each group were evident on any of the task variables. Thus, gender differences do not appear to significantly influence the capacity to inhibit responding, (i.e., impulse control), neither immediately nor over time amongst kindergarten students in this research sample. Intercorrelational results between teacher ratings on the School Learning Profile and students' impulse control performances on the GDS Delay Task, (i.e., mean Total E.R. scores), yielded no significant correlations for either group. Thus, there appears to be no significant relationship between demonstrated impulse control abilities in the testing situation and teacher ratings of attentional/impulsive DISCUSSION AND CONCLUSIONS / 110 behaviours displayed within and outside of the classroom, on the SLP variables. One possible explanation of these findings is that perhaps teachers were not assessing primarily impulsivity on this SLP variable or that the Total E.R. score does not reflect a subject's demonstrated level of impulsivity as it would be displayed within a school setting, as it purports to. As all the Pearson r values for the intercorrelations between these two measures are very low, it is possible that each measure is evaluating something different. If, in fact, any of the SLP variables evaluate primarily impulsivity and performance on the Delay Task truly reflects levels of impulsivity as they would be displayed at school, then higher correlations with the Delay Task scores would have been anticipated, especially since both control and "high risk" groups' scores suggest tendencies towards impulsive responding amongst the members of each group. These findings contradict those of Gordon's investigation (1979) in which teacher ratings of hyperactivity, using the Conners' Behavior Rating Scale (Conners, 1969) were significantly correlated with all of the DRL (6-second schedule) variables; coefficients with the whole-session efficiency score were, for example, r = 0.53, p-SO.01. Interestingly enough, Delay Task performance reveals practically no relationship to teachers' ratings of control students on 'overall learning ability', the very variable designated, by definition of the study in sample selection, to identify and differentiate normally achieving control students from "high risk" students. In summary, the prediction that kindergarten students identified as being at "high risk" for later experiencing academic failure would encounter greater difficulties in refraining from impulsive responding on the GDS Delay Task than DISCUSSION AND CONCLUSIONS / 111 controls was not confirmed by the results of this study. In fact, average performances of students in both control and "high risk" groups were indicative of general tendencies towards impulsive responding, as most scores for both groups closely approximate the cut-off points between NORMAL and BORDERLINE ranges of performance, as per Gordon's statistically determined Threshold Tables (1986a). Mean Total Correct Responses scores alone yielded significant differences between the two sample groups, contrary to Gordon's earlier findings (1979) in which a highly significant main group (hyperactive vs. nonhyperactive) effect was obtained for Total Correct Responses, Total Responses and especially for Total Efficiency Ratio scores. Considering that the criterion for group membership in Gordon's study was quite different from that of this study, comparative conclusions cannot be reliable nor valid. As the Total E.R. scores generally correlated very weakly with teacher ratings on the majority of the SLP variables, the validity that each of these measures, or particular variables on these measures, assessed impulse control abilities is somewhat questionable. While the "high risk" students demonstrated less response accuracy on each trial compared to the controls, their response rate paralleled that of the controls and was not found to be significantly different. Some performance deterioration was noted for the "high risk" group. Additionally, the "high risk" group was less able to refrain from impulsive responding over time, as some performance deterioration was observed on the last two blocks. • Gender differences in impulse control abilities within each sample group revealed insignificant results, (i.e., boys performed neither worse nor better than girls). DISCUSSION AND CONCLUSIONS / 112 c. Vigilance Task Summary A brief outline of obtained significant differences on mean Vigilance Task performance between the groups and across blocks is summarized below. Significant (p^O.0167) • Total Correct Responses • Blocks Correct Responses • Total Commission Errors • Block Commission Errors (across blocks only) Insignificant (p>0.0167) • Interactive effects between group and block on Block Correct Responses, Block Omission Errors and Block Commission Errors d. Vigilance Task Discussion The Vigilance Task examines a child's ability to focus and maintain attention over time in the absence of reinforcement. As no immediate feedback on performance is provided, this task requires that a child must be self-motivated to succeed. Impulse suppression is an additional and an important task demand, as success is also dependent on the child's ability to refrain from impulsive/random responding to alerting but non-target stimuli, (i.e., "non-hot" numbers which immediately follow the target number "1"). Two noteworthy summary scores provide valuable interpretive information: number of Correct Responses and number of Commission Errors. Statistically significant performance differences, in general, between the control and "high risk" groups were achieved on this Task. The "high risk" DISCUSSION AND CONCLUSIONS / 113 group responded to significantly fewer target stimuli than did the control group, with respective displayed performances at the 16th and 47th percentiles. Under conditions of relatively high arousal, children rated by their teachers as being within the lowest 10% for overall learning ability revealed tendencies toward experiencing deficits in sustaining attention over time. Their demonstrated weakness in achieving and maintaining alert, vigilant responding is reflected in their obtained Total Correct Responses scores which places their mean performance at the upper end of the BORDERLINE range, according to Gordon's Threshold Tables, compared with the NORMAL mean performance of the normally achieving control group. Lapses in sustained attention over time, although evident in both sample groups' Block Correct Responses performances, further differentiated the control and "high risk" groups. This differentiation was particularly evident on Block 1 and 2 scores. While performance deteriorated most dramatically between the first and second trials for both groups, greater deterioration in appropriately responding to the target stimulus was observed amongst the "high risk" students. This response pattern continued into the third trial, that is, Block 3, with less deterioration indicated by the scores for both groups. Thus, these results suggest that both normally achieving and underachieving kindergarten students in this study's sample experienced some deficits, in varying degrees, in focusing attention and maintaining vigilance to the task appropriately over a 6-minute (3 trials) period, with the "high risk" group demonstrating significantly greater difficulties. Additionally, while standard deviations were not particularly high, in general, on any of the 'Correct Responses' variables, the "high risk" group inevitably yielded approximately double the values compared to the controls, DISCUSSION AND CONCLUSIONS / 114 suggestive of twice as much variability in individual performance within the "high risk" group. Like the Correct Responses scores, Errors of Omission scores provide an index of lapses in sustained attention, reflecting times at which the child lacked sufficient attention to the Task, so that no response to the target number "1" was made when it flashed onto the screen. Correct Responses scores, therefore, mirror those of the Omission Errors since the former reflects the number of target stimulus presentations to which the student responded. Quite expectedly, therefore, "high risk" students made significantly more Errors of Omission during each Block with greater variability in individual performances than did the controls with both groups demonstrating increased sustained attentional deficits over time. Interestingly enough, however, while the "high risk" students' performance showed the greatest increase in Omission Errors between Blocks 1 and 2, mirroring their Block Correct Responses scores, the control groups' Block Omission Errors scores did not reflect their Block Correct Responses scores. Instead, stability in the production of omission errors was maintained from Blocks 1 to 2, however, the control group experienced considerable difficulties sustaining attention on the third trial, as indicated by a sharp increase in errors of omission, unlike the plateauing deterioration pattern of the "high risk" group. Thus, neither Omission Errors scores nor a general pattern of performance accurately mirrored the Correct Responses scores achieved by the control group. "High risk" students tended to respond significantly more often to incorrect stimuli than the controls, reflecting a significantly greater degree of exhibited impulsive responding amongst "high risk" students. Commission Errors are unrelated to either Correct Responses or Omission Errors performances (Gordon, DISCUSSION AND CONCLUSIONS / 115 1986a) but instead represent responding to stimuli " . . . in an ill-conceived or poorly controlled fashion" (Gordon, 1986a, p. 26) instead of reflecting lapses in sustained attention. Once again, BORDERLINE and NORMAL range data were obtained for "high risk" and control groups, respectively. The Total Commission Errors variable, however, depicted "high risk" students as having unequivocal impulse control deficits on tasks demanding vigilance and alertness as their scores fell well into the middle of the BORDERLINE range, contrary to scores obtained on other GDS Delay and Vigilance Task variables, on which scores more closely approximated the cut-off points between the NORMAL and BORDERLINE ranges making interpretation less definitive. Interpretive consistency between two GDS variables purporting to measure the degree of impulsive responding appear to be somewhat weak in these data. This inconsistency may suggest that impulsive behaviour is more apparent under conditions demanding vigilance and alertness than when such conditions are absent. Intercorrelational data between Vigilance Task Total Commission Errors scores and teacher ratings on the SLP variables reveal the presence of more frequent significant positive coefficients amongst only the "high risk" students, reflecting the presence of relatively weak to relatively strong relationships. The normally achieving control group, on the other hand, produced insignificant and very low coefficients between the Delay Task Total E.R, scores and the SLP variables. The largest and most significant relationship with Total Commission Errors scores for the "high risk" group arose on the SLP variable 'defiant and uncooperative', implying that such behaviour is not governed by deficits in impulse control but additionally such behaviour becomes more apparent in situations which demand sustained attention and alertness. Intercorrelational DISCUSSION AND CONCLUSIONS / 116 results between this same SLP variable and Vigilance Task Total Correct Responses scores (a purported measure of sustained attention) revealed insignificant and very low correlations for both groups, suggesting that sustained attention and impulse control alone do not appear to influence teacher ratings of defiant and uncooperative behaviour. Only under conditions requiring vigilance and alertness did impulse control deficits become apparent through behaviours such as defiancy and lack of cooperation. 'Difficulty waiting turns in games or group activities' is an additional such behaviour that appeared to be governed to a similar degree, (i.e., relatively high coefficient value) by impulse control deficits under the above-stated conditions. Both control and "high risk" groups exhibited the same performance pattern over time. While errors of commission decreased from Block to Block, the most dramatic such decrease arose on Block 2. Thus, all students demonstrated some ability to refrain from impulsive responding after the first trial, responding less to incorrect stimuli, especially noticeable on the second trial after which a general plateauing effect occurred. Noteworthy also, are the large standard deviations obtained for mean Block performances for both groups, suggestive of great variability in individual performances within each group. Additionally, there appeared to be no significant interactive effects between group and block performances. Intercorrelational data between the Delay and Vigilance Tasks yielded few significant and predominantly low correlation coefficients for the controls. "High risk" students revealed some highly significant and moderate-to-strong relationships between Delay Task Block E.R. and Vigilance Task Blocks 2 and 3 Correct Responses and Omission Errors scores. Block 2 Correct Responses and Omission DISCUSSION AND CONCLUSIONS / 117 Errors also correlated highly and significantly with Blocks 1 and 2 Responses scores (Delay Task) and the Total E.R. score. As the Delay Task Total E.R. score purports to measure a child's level of impulsive responding, it would be expected that a highly significant and strongly correlated relationship would have existed with the total number of commission errors being made on the Vigilance Task as this variable is also considered to reflect the level of impulsive responding. Since this was not obtained, it would suggest that these two variables may represent different aspects of impulse control abilities and that one's true ability to control impulsive responding is challenged more under conditions demanding both sustained attention and alertness. Gender differences in performance are particularly apparent within the "high risk" group. "High risk" boys were found to have significantly greater difficulties sustaining attention responding to the target stimulus and also greater difficulties in refraining from responding to incorrect stimuli than did the "high risk" girls. Over time, "high risk" boys exhibited generally weaker performances than girls, particularly evident on the mean number of commission errors made from Block to Block. Both sexes displayed the same pattern of mean responses over time as revealed by the entire "high risk" group scores, (i.e., greatest number of commission errors made during the first trial followed by noticeable decreases in the subsequent two trials, particularly during Block 2). "High risk" boys generally made twice as many errors of commission in each Block than did the girls. In summary, the control students demonstrated a significantly better ability in sustaining attention and controlling impulsive responding over three two-minute trials than did the "high risk" students. The control group generated significantly DISCUSSION AND CONCLUSIONS / 118 more correct responses to the target stimulus "1", fewer errors of omission, and greater errors of commission. Intercorrelational data indicated that Total E.R. scores on the Delay Task and Total Commission Errors scores on the Vigilance Task were not correlated for either sample group suggesting that these two variables are not measuring the same aspects of impulsivity. Several significant correlations for the "high risk" group between Total Commission Errors scores and teacher ratings were obtained which were not apparent for this same group between Total E.R. performances and teacher ratings. Thus, it appears that under conditions requiring sustained attention and alertness, impulse control abilities strongly correlated with teacher ratings of exhibited attentional/impulsive behaviours, achieving statistically significant levels. Gender differences in performance were noticeable moreso within the "high risk" group than within the normally achieving group. Boys exhibited generally weaker performances than girls, having greater difficulties both with sustaining attention and with refraining from impulsive responding. These differences, however, did not achieve statistical significance. No significant interactive effects between gender, block and group were apparent. 3. Bead Memory - S B F E Highly significant mean score differences in short term visual memory abilities were apparent between the control and "high risk" groups. Short Term Memory Area Standard Age Scores (SAS) indicated control students to be functioning at the 45th percentile (SAS=98) and within the AVERAGE range, while the "high risk" students were found to be functioning at only the 8th percentile (SAS = 78) and within the SLOW LEARNER range, according to the DISCUSSION AND CONCLUSIONS / 119 SBFE normative data (Thorndike, Hagan & Sattler, 1986). Success on this task requires intact attentional and concentration skills. Such disparity between the groups in mean task performance could imply significantly greater difficulties with any form of perception and discrimination, spatial relations and/or alertness to details amongst the underachievers in this study. As these "high risk" students were found to display a significantly greater impulsive response pattern on the GDS Vigilance Task, (i.e., under conditions requiring sustained attention and alertness), they may also have greater motor-coordination difficulties in manipulating objects, thus, leading to lower performance scores. Intercorrelational data suggest the lack of any significant relationships between performances on the Bead Memory subtest and the GDS Vigilance Task for both groups. Very low coefficients were obtained on the Total Correct Responses scores (an indicator of sustained attention and alertness) and the Bead Memory subtest. Thus, short term visual memory abilities, while dependent on intact attentional and concentration skills, do not appear to be significantly related to vigilant performance, as measured by the GDS. Additionally, a significant but moderately weak negative coefficient (reflecting a relatively weak inverse relationship) was obtained for the "high risk" group only when correlating Bead Memory performance and Total Commission Errors on the GDS Vigilance Task. As this latter variable purports to assess levels of impulsivity, examination of correlational data between Bead Memory and the Total E.R, score of the Delay Task (another purported measure of impulse control) revealed insignificant and very weak relationships for both the control and "high risk" groups, once again differentiating the conditions under which impulsivity is best displayed. DISCUSSION AND CONCLUSIONS / 120 C. CONCLUSIONS The results of this exploratory study permit the following conclusions to be drawn about the relationships of sustained attention, impulse control and short term visual memory abilities within normally achieving and underachieving kindergarten children. 1. Hypothesis 1 There will be no significant (p<.05) relationship (correlation) between test scores of impulsivity (GDS) and teacher ratings on the 13 variables of the SLP for normally achieving and "high risk" kindergarten students. 2. Hypothesis 2 There will be no significant (p<.05) relationship (correlation) between test scores of vigilance (GDS) and teacher ratings on the 13 variables of the SLP for normally achieving and "high risk" kindergarten students. 1. As it is extremely difficult to unequivocally determine the extent to which each of attentional and impulse control abilities influence the DSM LII-R (1987) behavioural criteria for ADHD/Undifferentiated ADD incorporated in item 13 of the SLP, conclusive statements about the presence of relationships between vigilance and impulse control GDS performances and teacher ratings of the same lack reliability and validity. Justifiable conclusions can be drawn, however, from teacher ratings alone. Kindergarten teachers rated underachievers as exhibiting significantly greater behavioural DISCUSSION AND CONCLUSIONS / 121 problems characteristic of ADHD/Undifferentiated ADD both within and outside of the classroom situation than normally achieving control students. 2. While control students elicited significantly more frequent responses after waiting the appropriate time interval on the GDS Delay Task, no significant differences in the percentage of correct button presses was evident between the groups. Thus, Total E.R. scores revealed insignificant differences in exhibited levels of impulsivity for both normally achieving and underachieving students, however, a tendency for both sample groups toward impulsive responding was evident, as both groups' scores approximated the cut-off point between statistically derived NORMAL/BORDERLINE performance ranges. Gordon's purported best single indicator of the level of demonstrated impulsivity, (i.e., Total E.R. score), therefore, failed to differentiate normally achieving from underachieving kindergarten students in this study's research sample. 3. "High risk" students demonstrated significantly greater difficulties refraining from impulsive responding than controls on a task demanding both sustained attention and alertness. It is these underlying task conditions that are believed to influence the exhibition of impulsivity particularly within the teacher-nominated underachieving group. Additionally, the lack of immediate reinforcement of performance on this Task, contrary to the Delay Task, may also have been a factor in significantly greater impulse control problems having been displayed by "high risk" students. Thus, Total Commission Errors performance on the GDS Vigilance Task differentiated normally achieving from underachieving kindergarten students significantly better than did the Total E.R. scores of the Delay Task, Gordon's DISCUSSION AND CONCLUSIONS / 122 purported best single indicator of impulsivity. 4. The GDS Vigilance Task identified control students as displaying a significantly better capacity of successful vigilant responding than "high risk" students. Relatively strong correlations to teacher ratings of attention/impulse control behaviour problems achieved significance for only the "high risk" group. Greater frequencies of failing to complete things started (short attention span), experiencing temper tantrums and having difficulties waiting turns in games/group activities bore the greatest relationship to focusing and maintaining attention over time. 3. Hypothesis 3 There will be no significant (p<.05) relationship (correlation) between teacher ratings of attention and school readiness (overall ability to learn school materials) for each of the normally achieving and "high risk" groups. 4. Hypothesis 4 There will be no significant difference (p<.05) between normally achieving and "high risk" groups on teacher ratings of attention and school readiness (overall ability to learn school materials). 1. Predominantly significant and qualitatively positive relationships between teacher ratings of attention/impulse control and school readiness were obtained on most SLP variables for the control group only. Strong confidence can be given to the existence of highly significant relationships DISCUSSION AND CONCLUSIONS / 123 between overall learning ability and (1) attention span/distractibility as related to following classroom directions; and (2) short attention span as related to completing tasks started. Behaviours considered by teachers not to influence or be related to a child's overall learning ability were 'temper tantrums' and 'blurting out answers to questions before they have been completed'. 2. Teachers rated "high risk" students as exhibiting attentional/impulsive behaviour problems significantly more frequently than control students on all variables except 'temper tantrums' and 'blurting out answers to questions before they have been completed', the two variables that revealed very weak correlations with ratings of 'overall ability to learn school materials'. Teachers most readily differentiated control from "high risk" students based on behaviours such as being easily distracted, having difficulty listening, fidgeting and exhibiting a short attention span (related to failing to complete tasks started). 5. Hypothesis 5 There will be no significant (p<.05) difference between the normally achieving and "high risk" groups on the GDS measures of impulsivity (Delay Task — Total ER score) and sustained attention (Vigilance Task — Total Correct Responses and Total Commission Errors scores). DISCUSSION AND CONCLUSIONS / 124 6. Hypothesis 6 There will be no significant (p<.05) difference between normally achieving and "high risk" groups on their Vigilance Task performance over time (i.e., Blocks 1, 2, and 3). 1. Significant differences in teacher ratings of attentional/impulse control difficulties and objective performances of sustained attention and impulsivity on the GDS are present for only few of the SLP variables and only among the "high risk" group. 2. Inconsistent intercorrelational results between GDS impulse control performance and teacher ratings for the "high risk" group suggest that teacher ratings tend to correlate more with impulse control performance generated under conditions demanding both vigilance and alertness. 3. Under conditions demanding vigilance and alertness, "high risk" students demonstrated levels of impulsivity which were greater than that of their peers, or normally achieving control students, yielding a normal level of impulsive responding. 7. Hypothesis 7 There will be no significant (p<.05) relationship (correlation) between the GDS measures of sustained attention (Vigilance Task — Total Correct Responses and Total Commission Errors scores) and impulsivity (Delay Task — Total Efficiency Ratio score) between the control and "high risk" groups. DISCUSSION AND CONCLUSIONS / 125 1. A significant relatively strong relationship between levels of impulsive responding (Delay Task) and the ability to focus and maintain attention over time (Vigilance Task) was apparent for the "high risk" group only. No noteworthy relationship for either sample group was found to be present between the two GDS measures of impulsivity, namely Delay Task Total E.R. scores and Vigilance Task Total Commission Errors scores, suggesting that these two measures may well be evaluating different aspects of impulse control as the nature of these two GDS impulse control tasks is quite different. 2. No interactive effects across blocks nor between group and block were seen on the Delay Task. Significant interactive effects were not obtained for group and block on the Vigilance Task as well. 8. Hypothesis 8 There will be no significant (p<.05) difference between boys and girls on test scores of the GDS Vigilance Task (inattentiveness). 1. While gender appeared not to influence vigilance abilities for control students, teacher-nominated underachieving "high risk" boys experienced significantly greater difficulties sustaining attention and alertness throughout the trials of the Vigilance Task than did the "high risk" girls. 2. Under conditions demanding vigilant and alert responding, insignificant gender differences in a displayed normal level of impulsivity were evident DISCUSSION AND CONCLUSIONS / 126 for the control students. Boys at-risk for later academic problems experienced significantly greater difficulties refraining from impulsive responding, making at least twice as many errors of commission on each trial than did at-risk girls. 3. Between the groups, the normally achieving sex performed, on average, better than the corresponding sex within the underachieving group, with boys experiencing greater difficulties both refraining from impulsive responding and sustaining attention than did the girls. 4. Gender did not appear to influence teacher ratings of attentional behaviour problems nor GDS Delay Task performances (refraining from impulsive responding), in general, neither for the normal achievers nor for the under achievers. 5. The GDS better differentiated gender performances of vigilance and impulse control abilities, particularly for the "high risk" students, than did teacher ratings of attentional/impulsive behaviour problems, following the diagnostic criteria of ADHD/Undifferentiated ADD set forth by the DSM III-R. 9. Hypothesis 9 There will be no significant (p<.05) relationship (correlation) between test scores of impulsivity (GDS) and visual memory (Bead Memory subtest — Stanford-Binet Intelligence Scale: Fourth Edition) for normally achieving and "high risk" kindergarten students. DISCUSSION AND CONCLUSIONS / 127 10. Hypothesis 10 There will be no significant (p<.05) relationship (correlation) between test scores of sustained attention (GDS) and visual memory (Bead Memory subtest — Stanford-Binet Intelligence Scale: Fourth Edition) for normally achieving and "high risk" kindergarten students. 1. Although short term visual memory performance significantly differentiated the normally achieving from the academically at-risk students, no indications of significant relationships to sustained attentional skills nor to impulse control abilities were apparent. Gender, as well, did not appear to influence short term memory performances. D. LIMITATIONS Several limitations of this study are evident. • As the School Learning Profile (SLP) was not experimentally validated prior to its' use in this study, the reliability and validity of its items in truly measuring school readiness, attentional skills and impulse control abilities are unknown. Confidence in deriving any definite conclusions from the obtained data therefore, is limited. Although these limitations do exist, the purpose of this study was not to validate the SLP. Well documented support (Gresham et al., 1987; Sattler, 1988) of teachers being good judges of students' behaviours was relied upon in the decision to use a teacher rating scale as one of the research instrumrnts in this study. The fact that the Conners' Teacher Rating Scale (1969) and the DSM IH-R (1987) diagnostic criteria DISCUSSION AND CONCLUSIONS / 128 for ADHD/Undifferentiated ADD, from which the SLP items are derived, are valid and reliable means of assessing attention and impulse control, lends further justification to the use of the SLP as a means of generating teachers' opinions of kindergarten students' behaviours. The extent to which the SLP items, especially the #13 subitems, differentiate the evaluation of attentional skills from impulse control abilities is unknown. For example, the instrument does not reveal the extent to which teacher ratings on items #13B ('difficulty staying seated') and #13C ('difficulty waiting turn') arise from intact/problematic attentional abilities compared with impulse control abilities, or the combination of the two. The unknown loading, therefore, of attentional and impulse control factors in each of the items, influencing teacher ratings, limits the validity of any conclusions made from the data obtained. Related to the preceding limitation, the unknown discriminatory power between items in measuring the same factor(s), (i.e., attention and/or impulse control), limits the conclusions that can be drawn from intercorrelational data analyses of the results examining the presence, strength and significance of relationships existing between teacher ratings of displayed behaviours. For example, can a teacher truly make a distinction between what s/he is being asked to evaluate on SLP#2 ('attention span and distractibility, as related to following classroom instructions') and SLP#13G ('difficulty listening')? Are these two items discriminatory enough in assessing the same factor? Answers to these questions are unknown as the internal consistency of this instrument was not investigated. Thus, the formulation of any definite conclusions is limited. DISCUSSION AND CONCLUSIONS / 129 • A larger sample size than that used in this study (28 normally achieving and 30 "high risk" students) would improve the reliability and validity of the results obtained. E. IMPLICATIONS AND FUTURE RESEARCH This study attempted to investigate the relationship of teacher ratings of kindergarten students' school readiness with respect to attention and impulse control factors, and students' sustained attentional abilities and levels of impulsivity as measured by the Gordon Diagnostic System. As the GDS purports to be able to accurately discriminate hyperactive from nonhyperactive children (Gordon, 1984), the GDS was used in this study to examine its usefulness in differentiating between children exhibiting early symptoms of ADHD/Undifferentiated ADD that interfere with/impair their overall ability to learn school materials. As the reliability and validity of teacher rating scales in the assessment of behaviour problems have been heavily criticized in the past, this exploratory study also sought to ascertain if the GDS may be a more suitable assessment measure of sustained attention and impulse control abilities, providing objective data, than the more subjective information obtained through teacher rating scales. Performance on the GDS Delay Task revealed practically no relationship to teachers' ratings of normally achieving and "high risk" students of 'overall learning ability' (designated to differentiate control from "high risk" students in this study) and attention and impulsivity. Contrary to this, GDS Vigilance Task performance yielded significant relationships to teachers' ratings primarily for the "high risk" group. These results suggest that sustained attention and impulse DISCUSSION AND CONCLUSIONS / 130 control performance measured by the Vigilance Task of the GDS may provide more useful information than the Delay Task in assessing academically at-risk children who also exhibit problematic behaviours symptomatic of ADHD/Undifferentiated ADD. Very low and insignificant correlations between teachers' ratings of 'overall learning ability' and GDS performance were obtained for the normally achieving control students, suggestive of the GDS being an inadequate psychometric instrument to be used in differentiating normally achieving from underachieving kindergarten students. As the GDS was designed to differentiate hyperactive from nonhyperactive children, the aforementioned correlational finding is not surprising. Unequivocally conclusive results were not obtained in this study that would suggest the GDS measure be used to replace teacher ratings of problematic attentional/impulsive behaviours, or vice versa, in the assessment process. Instead, it appears that the GDS would be a useful diagnostic tool to use in conjunction with other standard measures to obtain more in-depth information on vigilance and impulse control functioning. For example, observational performance data provided by the GDS can be most beneficial in determining various cognitive strategies (or lack of) employed during the tasks, motivational levels, emotional factors, (i.e., passive versus aggressive), invoked by the tasks indicative of frustration tolerance, etc. There would be a risk, however, if such observational data were used as the sole means of collecting information on these additionally important factors influencing performance. A teacher-identified hyperactive child may be so completely involved with the task that classroom-exhibited hyperactive behaviour may well not be displayed due to the novelty, high interest, or challenge of the task conditions set forth by the GDS microprocessor device. It is DISCUSSION AND CONCLUSIONS / 131 important to remember that the testing situation reflects only some of the features of the classroom, (e.g., sitting at a table and chair). It cannot, however, reflect the socially and environmentally stimulating elements of the classroom which may well be necessities for impulsive or inattentive behaviours to be exhibited. On the other hand, GDS performance may reveal that a child displaying no hyperactive behaviour in the classroom situation would exhibit tendencies toward attentional and impulse control difficulties, undetected by the classroom teacher. Thus, this study supports use of the GDS in the assessment of attention and impulse control abilities in conjunction with, and not as a replacement for, teacher rating scales. Although intact attentional and impulse control skills are important prerequisites in facilitating the learning process (Murphy-Berman, Rosell & Wright, 1986), the likelihood that learning will be hampered when difficulties are encountered in either of these two areas is probable, however, it is not valid to assume that learning cannot occur adequately. Since learning tends to be very individualistic, giving recognition to some basic commonalities, some children experiencing difficulties in sustaining their attention to a task or controlling their impulses may still perform academically within average to high average/superior ranges of functioning through possible use of various effective learning/coping strategies. To suggest, therefore, that attentional or impulse control deficits (as experienced by ADHD/Undifferentiated ADD diagnosed children). conclusively leads to academic underachievement would be inaccurate. As important as it is to identify those children experiencing attentional/impulse control deficits who are underachieving in their school learning DISCUSSION AND CONCLUSIONS / 132 situation, it is conceivably equally important to be aware of those children experiencing these same deficits but who appear to be achieving normally or even exceptionally well. Rather than focusing on the underachievers, it would be more important to question why some ADHD/Undifferentiated ADD children are underachievers and others are not, (i.e., what are the factors that seem to differentiate these two groups?). For example, are these differences attributable merely to different intellectual abilities or to varying usages of coping skills, learning strategies, environmental cues, etc.? Considering that the construct of ADHD/ADD is a hypothetical one whose diagnostic criteria are subject to periodic change, it may not be too surprising to find that some such children are underachievers while others are not. Directions in future research need to address these issues/questions. The GDS could prove to be a useful tool in identifying some of these factors that may well influence learning and place certain children at-risk for later academic failure. Vigilance and impulse control may be found to play a much greater role in learning than has been previously thought. Further exploration into the development of attentional and impulse control skills would be most beneficial, specifically focusing on evaluations of how early identification of attention/impulse control problems in the classroom could be used to aid in educational planning and decision-making. Providing an at-risk student with appropriate learning strategies, (e.g., modification of cognitive styles, organizational and reflective strategies, etc.), early in their school years may greatly help the child circumvent certain learning problems, increasing the likelihood of him/her experiencing more academic successes than failures. Longitudinal studies comparing GDS performance, teacher ratings of DISCUSSION AND CONCLUSIONS / 133 attention and impulse control with achievement measures later in the primary, intermediate and secondary grades could serve useful predictive purposes in the early identification of academically at-risk students manifesting early symptoms of ADHD/Undifferentiated ADD. Interesting, too, would be to ascertain both the short-term and long-term effects of implementing remedial interventions at various grade levels on student achievement. Would an early identified at-risk kindergarten student who had been taught to use appropriate learning strategies to minimize inattentiveness and impulsivity achieve better in Grade 4, for example, than an at-risk peer, also identified in kindergarten, who had not received any educational intervention? How might the former student's achievement in Grade 4 compare with that of a fellow Grade 4 student identified later,(e.g., in Grade 2),and who had also received remediation but which had commenced only during the Second Grade? Additionally, investigating the impact of short-term remediation of attentional deficits and impulsivity, (e.g., 6 to 10 week programs), through pre- and post-tesing, on achievement could have very important implications for the remedial role of special educators such as Learning Assistance Teachers. In conclusion, attention deficits in the school years represent a significant area of developmental-behavioural and academic dysfunction. The high prevalence of this problem, together with its major impact on the quality of life, makes it imperative to study its possible etiologies, clinical manifestations, and treatments. Early identification of risk factors and initial manifestations is critical in the early remediation of children exhibiting attentional deficits, giving these children better opportunities at experiencing academic success and social/emotional development early in their lives. This study suggests that teachers' opinions of a DISCUSSION AND CONCLUSIONS / 134 children's behaviours and academic performances continue to play a critical role in detecting attentional/impulse control difficulties early in a child's academic life. Support for replacing subjective teacher ratings with the Gordon Diagnostic System, a more objective means of obtaining similar information, was not achieved in this study. Instead, it is recommended that the Gordon Diagnostic System be used in conjunction with teacher ratings in early identification procedures when attentional deficits are observed in a child's academic and/or nonacademic performances/displayed behaviours. REFERENCES American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed. Revised). Washington, D.C: pp. 50-53, 95. Barkley, R.A. (1977). A review of stimulant drug research with hyperactive children. Journal of Child Psychology and Psychiatry, 18, 327-345. Barkley, R.A. (1981). Hyperactive Children. New York: The Guildford Press. Barkley, R.A., Fischer, M., Newby, R.F., & Breen, M.J. (1988). Development of a multimethod clinical protocol for assessing stimulant drug response in children with ADD. Journal of Clinical Child Psychology, 2 7(1), 14-24. Becker, L.D., & Snider, M.A. (1979). Teachers' ratings and predicting special class placement. Journal of Learning Disabilities, 12, 96-99. Benezra, E., & Douglas, V.I. (1988). Short-term serial recall in ADDH, normal, and reading-disabled boys. Journal of Abnormal Child Psychology, 16(5) 511-525. Borcherding, B., Thompson, K., Kruesi, M., Bartko, J., Rapport, J.L., & Weingartner, H. (1988). Automatic and effortful processing in attention deficit/hyperactivity disorder. Journal of Abnormal Child Psychology, 16(3), 333-345. Brown, R.T., & Wynne, M.E. (1982). Correlates of teacher ratings, sustained attention, and impulsivity in hyperactive and normal boys. Journal of Clinical Child Psychiatry, 11 (3), 262-267. Bryan, T. (1974). An observational analysis of classroom behaviors of children with learning disabilities. Journal of Learning Disabilities, 7, 35-43. Cairns, E. & Cammock, T. (1978). Development of a more reliable version of the MFFT. Developmental Psychology, 14, 555-560. Campbell, SB., Endman, M.W., & Bernfeld, G. (1977). A three-year follow-up of hyperactive preschoolers into elementary school. Journal of Child Psychology and Psychiatry, 18, 239-249. Cantwell, D.P. (1987). How are DSM-IH and DSM-III(R) used to make the diagnosis of attention deficit disorder? Journal of Children in Contemporary Society, 19(1/2), 5-17. 135 / 136 Carlson, C.L., Laney, B.B., & Neeper, R. (1986). Direct assessment of the cognitive correlates of attention deficit disorders with and without hyperactivity. Journal of Psychopathology and Behavioral Assesment, 8(1), 69-86. Carter, J., & Conry, J. (1988). Kindergarten School Learning Profile. Unpublished manuscript, The University of British Columbia, British Columbia. Chamberlin, R.W. (1977). Can we identify a group of children at age two who are at risk for the development of behavioral or emotional problems in kindergarten or first grade? Pediatrics, 59(Suppl.), 971-981. Charles, L., Schain, R.J., Zelniker, T., & Guthrie, D. (1979). Effects of methylphenidate on hyperactive children's ability to sustain attention. Pediatrics, 64, 412-418. Cohen, N., Douglas, V., & Morganstern, G. (1972). The effect of methylphenidate on attentive behavior and autonomic activity in hyperactive children. Psychopharmacologia, 22, 282-294. Conners, C.K. (1969). A teacher rating scale for use in drug studies with children. American Journal of Mental Deficiency, 126, 884-888. Cooper, D.H., & Farran, D.C. (1988). Behavioral risk factors in kindergarten. Early Childhood Research Quarterly, 3, 1-19. Day, A.M.L., & Peters, R.D.V. (1989). Assessment of attentional difficulties in understanding children. Journal of Educational Research, 82(6), 356-361. Douglas, V.I. (1972). Stop, look and listen: The problem of sustained attention and impulse control in hyperactive and normal children. Canadian Journal of Behavioural Science, 4, 259-282. Douglas, V.I. (1983). Attentional and cognitive problems. In M. Rutter (Ed.), Developmental neuropsychiatry (pp. 280-329). New York: Guildford Press. Douglas, V.I. & Peters, K.G. (1979). Toward a clearer definition of attentional deficit of hyperactive children. In G.A. Hale & M. Lewis (Eds.) Attention and Cognitive Development. New York: Plenum Press. Dykstra, R. (1967). The use of reading readiness tests for prediction and diagnosis: A critique. Perspectives in Reading, 8, 35-50. Eaves, L.C., Kendall, D.C, Crichton, J.U. (1972). The early detection of minimal brain dysfunction. Journal of Learning Disabilities, 5(8), 454-462. Edelbrock, C , & Achenbach, T.M. (1984). The teacher version of the Child Behavior Profile: I. Boys aged 6-11. Journal of Consulting and Clinical Psychology, 52, 207-217. / 137 Enns, J.T., & Akhtar, N. (1989). A developmental study of filtering in visual attention. Child Development, 60, 1188-1199. Farran, D.C, & McKinney, J.D. (1986). Risk in intellectual and psychosocial development. New York: Academic Press. Feshbach, S., Adelman, H., & Fuller, W.W. (1974). Early identification of children with high risk of reading failure. Journal of Learning Disabilities, 7, 639-644. Fischer, M., Rolf, J.E., Hasazi, J.E., & Cummings, L. (1984). Follow-up of a preschool epidemiological sample: Cross-age continuities and predictions of later adjustment with internalizing and externalizing dimensions of behavior. Child Development, 55, 137-150. Fisk, A.D., & Schneider, W. (1981). Control and automatic processing during tasks requiring sustained attention: A new approach to vigilance. Human Factors, 23, 737-750. Fletcher, J.M., & Satz, P. (1984). Test-based versus teacher-based predictions of academic achievement: A three-year longitudinal follow-up. Journal of Pediatric Psychology, 9, 193-203. Garfinkel, B.D. (1986). Recent developments in attention deficit disorders. Psychiatric Annals, 16(1), 11-15. Gerber, M.M., & Semmel, M.I. (1984). Teacher as imperfect test: Reconceptualizing the referral process. Educational Psychologist, 19, 137-148. Glass, G.V. & Hopkins, K.D. (1984). Statistical methods in education and psychology. 2nd Edition. Englewood Cliffs, N.J.: Prentice Hall. Gordon, M. (1979). The assessment of impulsive and mediating behaviours in hyperactive and nonhyperactive children. Journal of Abnormal Child Psychology, 7(3), 317-326. Gordon, M. (1986a). Technical manual and interpretive guide to the Gordon Diagnostic System. Syracuse: Gordon Systems, Inc. Gordon, M. (1986b). How is a computerized test used in the diagnosis of attention deficit disorder? Journal of Children in Contemporary Society, 19(1/2), 53-64. Gordon, M. (1986c). Microprocessor-based assessment of attention deficit disorders. Psychopharmacology Bulletin, 22, 288-290. Gordon, M., DiNiro, D., & Mettelman, B.B. (1988). The effect upon outcome of nuances in selection criteria for ADHD/Hyperactivity. Psychological Reports, 62, 539-544. / 138 Gordon, M., DiNiro, D., Mettelman, B.B., & Tallmadge, J. (1989). Observations of test behavior, quantitative scores, and teacher ratings. Journal of Psychoeducational Assessment, 7 141-147. Gordon, M., & McClure, F.D. (1983). The objective assessment of attention deficit disorders. Paper presented at the 91st Annual Convention of the American Psychological Association, Anaheim, CA. Gordon, M., & McClure, F.D. (1984). Assessment of attention deficit disorders using the Gordon Diagnostic System. Paper presented at the 92nd Annual Convention of the American Psychological Association, Toronto, Ont. Gordon, M. & Mettelman, B.B. (1988). The assessment of attention: I. Standardization and reliability of a behavior-based measure. Journal of Clinical Psychology, 44, 682-690. Gresham, F.M., Reschly, D.J., & Carey, M.P. (1987). Teachers as "tests": Classification accuracy and concurrent validation in the identification of learning disabled children. School Psychology Review, 16(4), 543-553. Hardman, S., & Oldridge, O.A. (1985). Predictive validity of the WISC-R on the Woodcock-Johnson Achievement Battery, the B.C. Quick Individual Educational Test, and teacher rankings of student achievement. Canadian Journal of School Psychology, 1, 31-38. Hoge, R.D., & Luce, S. (1979). Predicting academic achievement from classroom behavior. Review of Educational Research, 49, 479-496. Hynd, G.W., Nieves, N., Connor, R.T., Stone, P., Town, P., Becker, M.G., Lahey, B.B., & Lorys, A.R. (1989). Attention deficit disorders with and without hyperactivity: Reaction time and speed of cognitive processing. Journal of Learning Disabilities, 22(9), 573-580. Kavale, K.A. (1988). The long-term consequences of learning disabilities. In M.C. Wang, H.J. Wallberg, M.C. Reynolds (Eds.), The handbook of special education: Research and practice (pp. 303-344). Keogh, B.K., & Daley, S.E. (1983). Early identification: One component of comprehensive services for at-risk children. Topics in Early Childhood Special Education, 3, 7-15. Keogh, B.K., Wilcoxen, A.G., & Bernheimer, L. (1983). Prevention services for risk children: Evidence for policy and practice. In D.C. Farran & J.D. McKinney (Eds.), The concept of risk in intellectual and psychosocial development. New York: Academic Press. Keogh, B.K., & Margolis, J.S. (1976). A component analysis of attentional problems of educationally-handicapped boys. Journal of Abnormal Child Psychology, 4(4), 349-359. / 139 Keogh, B.K., Tehir, C , & Windeguth-Behn. A. (1974). Teachers' perceptions of educationally high risk children. Journal of Learning Disabilities, 7 367-374. Klee, S.E., & Garfinkel, B.D. (1983). The computerized continuous performance task: A new measure of inattention. Journal of Abnormal Child Psychology, 27(4), 489-496. Kohn, M., & Rossman, B. (1972). A social competence scale symptom checklist for the preschool child. Developmental Psychiatry, 6, 430-444. Koppitz, E.M. (1964). The Bender Gestalt Test for young children. New York: Grune and Stratton. Krupski, A. (1981). An interactionist approach to the study of attention problems in children with learning handicaps. Exceptional Education Quarterly, 2, 2-11. Kupietz, S.S., & Richardson, E. (1978). Children's vigilance performance and inattentiveness in the classroom. Journal of Child Psychology and Psychiatry, 19, 145-153. Leigh, J.E. (1983). Early labeling of children: Concerns and alternatives. Topics in Early Childhood Special Education Services, 3, 1-6. Leigh, J.E., & Riley, N. (1982). Learning disabilities in the early years: Characteristics, assessment, and intervention. Topics in Learning & Learning Disabilities, 2, 1-15. Lindsay, G.A., & Wedell, K. (1982). The early identification of educationally "at-risk" children revisited. Journal of Learning Disabilities, 15, 212-217. Mayes, S. (1987). Assessment of preschool hyperactivity: Combining rating scale and objective observation measures. Topics in Early Childhood Special Education, 6(4), 49-61. McClure, F.D., & Gordon, M. (1984). The performance of disturbed hyperactive and nonhyperactive children on an objective measure of hyperactivity. Journal of Abnormal Child Psychology, 22(4), 561-572. McKinney, J.D. (1975). Teacher perceptions of classroom behavior of reflective and impulsive children. Psychology in the Schools, 12, 348-352. McKinney, J.D. (1989). Longitudinal research on the behavioral characteristics of children with learning disabilities. Journal of Learning Disabilities, 22(3), 141-150. / 140 McKinney, J.D., Mason, J., Perkerson, K., & Clifford, M. (1975). Relationship between classroom behavior and achievement. Journal of Educational Psychology, 67, 198-203. Mercer, CD., Algozzine, B., & Trifiletti, J. (1979). Early identification - An analysis of the research. Journal of Learning Disabilities, 15, 212-217. Murphy-Berman, V., Rosell, J., & Wright, G. (1986). Measuring children's attention span: A microcomputer assessment technique. Journal of Educational Research, 80(1), 24-28. Palfrey, J.S., Levine, M.D., Walker, D.K., & Sullivan, M. (1985). The emergence of attention deficits in early childhood: A prospective study. Journal of Deveolopmental and Behavioral Pediatrics, 6(6), 339-348. Parasuraman, R. (1983). Vigilance, arousal and the brain. In A. Gale and J.A. Edwards (Eds.), Physiological correlates of human behavior: Attention and performance (pp. 1-32). London: Academic Press. Richman, J.E. (1987). Use of a sustained visual attention task to determine children at risk for learning problems. Journal of the American Optometric Association, 57(1), 20-26. Richards, G.P., Samuels, S.J., Turnure, J.E., & Ysseldyke, J.E. (1990). Sustained and selective attention in children with learning disabilities. Learning Disabilities, 23(2), 129-136. Rist, R. (1970). Student social class and teacher expectations: The self-fulfilling prophecy in ghetto education. Harvard Educational Review, 40, 411-451. Rolf, J., & Garmezy, N. (1974). The school performance of children vulnerable to behavior pathology. In D. Ricks, A. Thomas, & M. Rolf (Eds.), Life history research in psychopathology (Volume 3). Minneapolis: University of Minnesota Press. Ross, D.M., & Ross, S.A. (1982). Hyperactivity: Current issues, research and theory. New York: Wiley & Sons. Rosvold, H.E., Mirsky, A.F., Saranson, I., Bransome, E., & Beck, L.A. (1956). A continuous performance of brain damage. Journal of Clinical Psychology, 20, 343-352. Rubin, R. & Balow, B. (1978). Prevalence of teacher identified behavior problems: A longitudinal study. Exceptional Children, 54, 102-111. Salvia, J., & Ysseldyke, J.E. (1985). Assessment in special and remedial education. Third Edition. Boston: Houghton Mifflin Company. Sattler, J.M. (1988). Assessment of Children. Third Edition San Diego: Jerome Sattler, Publisher. / 141 Schachar, R., Logan, G., Wachsmuth, R., & Chajczyk, D. (1988). Attaining and maintaining preparation: a comparison of attention in hyperactive, normal, and disturbed control children. Journal of Abnormal Child Psychology, 19(4), 361-378. Siegel, L.S., & Ryan, E.B. (1989). The development of working memory in normally achieving and subtypes of learning disabled children. Child Development, 60, 973-980. Simner, M.L. (1983). The warning signs of school failure: An updated profde of the at-risk kindergarten child. Topics in Early Childhood Special Education, 3, 17-27. Simner, M.L. (1987). Predictive validity of the Teacher's School Readiness Inventory. Canadian Journal of School Psychology, 3, 21-32. Skinner, B.F. (1938). The Behavior of Organisms. New York: Appleton-Century-Crofts. Spivack, G. & Swift, M. (1977). "High-risk" classroom behaviors in kindergarten and first grade. American Journal of Community Psychology, 5, 385-397. Stevenson, H., Parker, T., Wilkinson, A., & Fish, E. (1976). Predictive value of teachers' ratings of young children. Journal of Educational Psychology, 68, 507-517. Swanson, H.L. (1980). Auditory and visual vigilance in normal and learning disabled readers. Learning Disability Quarterly, 3, 71-78. Swanson, H.L., Cochran, K.F., & Ewers, CA. (1990). Can learning disabilities be determined from working memory performance? Journal of Learning Disabilities, 23(1), 59-67. Sykes, D.H., Douglas, V.I., & Morgenstern, G. (1973). Sustained attention in hyperactive children. Journal of Child Psychology and Psychiatry, 14, 213-220. Sykes, D.H., Douglas, V.I., Weiss, G., & Minde, K.K. (1971). Attention in hyperactive children and the effect of methylphenidate. (Ritalin). Journal of Child Psychology and Psychiatry, 12, 129-139. Szumowski, E.K., Ewing, L.J., & Campbell, S.B. (1987). What happens to "hyperactive" preschoolers? Journal of Children in Contemporary Society, 19(112), 75-88. Thorndike, R.L., Hagen, E.P., & Sattler, J.M. (1986). The Stanford-Binet Intelligence Scale: Fourth Edition. Chicago, LL: Riverside Publishing Company. / 142 Torgesen, J.K. (1981). The relationship between memory and attention in learning disabilities. Exceptional Education Quarterly, Nov., 51-59. van der Meere, J., & Sergeant, J. (1988a). Controlled processing and vigilance in hyperactivity: Time will tell. Journal of Abnormal Child Psychology, 16(6), 641-655. van der Meere, J., & Sergeant, J. (1988b). Focused attention and pervasively hyperactive children. Journal of Abnormal Child Psychology, 16(6), 627-639. Vogel, S.A. (1990). Gender differences in intelligence, language, visual-motor abilities, and academic achievement in students with learning disabilities: A review of the literature. Journal of Learning Disabilities, 23(1), 44-52. Werry, J.S., & Aman, M. (1975). Methylphenidate and haloperidol in children: Effects on attention, memory, and activity. Archives of General Psychiatry, 32, 790-795. Whalen, C.K. (1983). Hyperactivity, learning problems, and attention deficit disorders. In T.H. Ollendick & M. Hersen (Eds.), Handbook of Child Psychopathology (pp. 151-199). New York: Plenum Press. APPENDIX A LETTER TO TEACHERS INVITING THEIR PARTICIPATION IN THE STUDY 143 A P P E N D I X B K I N D E R G A R T E N S C H O O L L E A R N I N G P R O F I L E ( T E A C H E R R A T I N G S C A L E ) 145 / 146 SCHOOL LEARNING PROFILE 1. Student's. Name Teacher's Name C o m p a r e d to other chi ldren vou have observed, please ra te this chi ld 's overa l l ability to learn school ma te r i a l . USE THE FIVE POINT SCALE BELOW; Circle one number between 1 and 6. lowest 10% lower 30% middle 40% upper 30% highest 10% but not but not lowest 10% highest 10% Attention span and distractibility: In class, does this child have difficulty following directions? Rating: Poor Attention Very good attention 1 2 S 4 5 3. Spoken language skills: Is this child able to articulate and speak clearly? Rating: Poor Very good articulation 1 2 3 4 5 4. Verbal sequences: Can this child verbally describe a sequence of events? Rating: Poor Very good skills 1 2 3 4 5 / 147 6. A l p h a b e t rec i tat ion: C a n this child recite the a lphabet? Ra t i ng : Poor V e r y good 1 2 3 4 5 6. Le t te r identif icat ion ski l ls : C a n this child correct ly n a m e uppercase letters shown in r a n d o m order? Ra t ing : N o n e A l l correct 1 2 3 4 5 7. N u m b e r ident i f icat ion: C a n this chi ld correct ly name n u m b e r s between one and twenty shown in r a n d o m order? Ra t i ng : N o n e A l l correct 1 2 3 4 5 8. P r i n t i ng : C a n this chi ld pr int his/ her n a m e correct ly w i thout reversa ls , deletions, addit ions, or misa l ignments ? Ra t i ng : Poor Perfect p r in t ing 1 2 3 4 5 9. F ine-motor sk i l l s : C a n this chi ld use sc issors to cut p a p e r correct ly? Ra t i ng : Poor V e r y 1 2 3 4 10. G r o s s motor ski l ls : Ra t e this chi ld 's abi l i ty for m o v e m e n t in phys i ca l educat ion and sports abi l i t ies. R a t i n g : Poo r V e r y 1 2 3 4 11. Socia l par t i c ipat ion : Ra te th is chi ld 's p l a y behav ior w i th the other ch i ldren. Ra t i ng : A g g r e s s i v e N o t aggress ive 1 2 s 4 r 12. Co lour d i sc r imina t ion : C a n this chi ld correct ly identify b y n a m e p r i m a r y colours shown in r a n d o m order? Ra t i ng : N o n e A l l correct 1 2 3 4 5 / 148 Please rate the behaviour of this child, with reference to your observations of him/her in the classroom, on the playground, or in other situations you have seen. To what degree does this child exhibit each behaviour below? Circle one number on the scale for each item. A. Fidgets B. Difficulty staying seated C. Difficulty waiting turn in games or group activity D. Easily distracted E. Defiant & uncooperative F. Has temper tantrums (explosive & unpredictable) G. Has difficulty listening H . Has difficulty playing quietly 0 I. Fails to finish things started (short attention span) 0 J . Blurts out answers to questions before they have been completed Never Rare Occasional Frequent Constant 0 1 2 3 2 2 2 2 2 3 3 3 3 3 3 3 APPENDIX C PARENT PERMISSION LETTER 149 / 150 The University of British Columbia The Education Clinic Faculty of Education 2125 Main Mall Vancouver, B.C. V6T 1Z5 April 28, 1988 Dear ParentVGuardian: RE: Study of Standardized Tests and Kindergarten Teacher Ratings of Readiness. This is to request your permission to allow your child, , to participate in a research project which is planned for May, 1988, in School District No. 71 (Courtenay). This project has been approved by the school district. The purpose of the project is to compare several newly developed standardized ability tests and kindergarten teacher ratings of school readiness. In this way we may determine how effective the tests are in diagnosing learning difficulties early in a child's school career, so that appropriate intervention may occur. It is hoped that follow-up assesment will be possible in one to two years" timeT" AT this time, however, permission for your child's participation is requested for the kindergarten screening. The tests we will be using involve some thinking tasks, problem solving, vocabulary and drawing skills. Based on past experience, it has been found that children enjoy working with the test materials. Approximately 15% of kindergarten students in School District No. 71 (Courtenay) will be selected for testing which requires approximately two hours in total. The results of the project will be used for research purposes. Your child's scores on the tests, therefore, will not become part of your child's school record. In the event that the tests do indicate some specific difficulties, we would contact you to ask for permission to consult with your child's school and discuss the possibilities of additional testing or recommendations for special education programs. We would be pleased to answer any questions you may have regarding the project. Please contact us at either of the telephone numbers or addresses below. It is important to note that your child's participation in this project is completely voluntary. If you decide that your child should not participate in the project, or wish to withdraw at any time, this decision will not affect your child's progress or status in school in any way. Please see the Parent Permission Form on the following page. ...12 Parent Permission Form Page 2 I do or do not (circle one) grant permission for my child to participate in this project, and I acknowlege receipt of a copy of this letter. I understand that my child will be tested by a qualified examiner in the child's school, and that my child's teacher may be asked to complete some brief rating forms about him/her. I also understand that my child's individual results will be kept strictly confidential. I am this child's parent or legal guardian and I am completing this form on the child's behalf. Name: (please print) Signature: Relationship to child: Address: • Telephone: If you consent for your child to participate, please complete the following confidential background information. Child's full name: Sex: _ _ _ Child's Age: Birthdate: Year Month Day . Parent's education (please check one in each column): Yean of Education Completed Mother or Father or Female Guardian Male Guardian Up to Grade 8 Grade 9 to 11 High school diploma or equivalent 1 • 3 years of college or technical school Four years, of university or more Does the child live with this person? (Check if yes) APPENDIX D THE GORDON DIAGNOSTIC SYSTEM (MODIFIED RECORDING SHEETS) 153 / 154 THE VJIORDON DELAY TASKS Date Tested: Date ot Birth: Age: R ESULTS YlAft MONTH OAT VIA* MONTH OAV T I M * MONTHS O A 1 C O R R E C T Sti. SUMMARY DATA Pot. TRACKING DATA Tntal C.nrtmrA t 0- 0.99 Total Responses 9 1 -1.99 Block l Correct •» 2- 2.99 Responses 4 3- 3.99 Block 2 Correct «. 4- 4.99 Responses B 5- S.99 R l i v k 1 CMrmrl 6- 6.99 Responses A 7- 6.99 Block 4 Correct 0 9-17.99 Resoontes in > 18 00 SUMMARY STATISTICS Score Normal BordVAbnl Total ER ER Block Variability Slope Scope Tola) Responses Total Correct STRATEG I ES/COMMENTS : Block: 1 2 3 4 . Total Score: R E S P O N S E S 1.0 | •H •H •H •H Block Sco'e EFF IC IENCY RATIO vo - s • Jt 1 2 3 4 Total PeakER YES / NO Valley ER YES / NO / 155 VIGILANCE TASKS TASK PARAMETERS  J Standard U Parallel l_J Preschool l_J Preschool 1 / 9 3 / 5 1" Interval 180" Blocks i Mode 0 Mode 2" Interval 120" Blocks _ | Other Parameters Block 1 Presentation Interval Block Length RESULTS Sel. SUMMARY DATA Po» Tolsl Correct 1 . Block 1 Correct. Omission. Commission. Block2 Correct. Omission . Commission . Blocks Correct. Omission. Commission . 2 3 * S 6 7 8 . 8 . 1 0 , TRACKING DATA iex XXB Errors XXI ot _ _ XI X Commission xsx XXX Block 1 Block 2 _ i t n C y Block 3 ».l M O Total C O R R E C T Block: 1 2 3 Total Score: Peak Score: YES / NO Valley Score: YES / NO OMISSIONS io • Block Score: Total SUMMARY STATISTICS Score Normal Bord/AbnL Total Commissions Commissions Block Variability Total Correct S T R A T E G I E S / C O M M E N T S : COMMISS IONS a rs a * — . _ M P a a »s «_.-« r_ a .r*CL .4 a at ra - - - -a -.—>«*_, ta a • _-f t _ IS •s a a - * ca SB m ES • . sa _ B _ Score: 1 2 3 Total Peak Score: YES / NO Valley Score YES / NO APPENDIX E INTERCORRELATIONS FOR CONTROL AND HIGH RISK STUDENTS ON GORDON DIAGNOSTIC SYSTEM TASKS; MEANS AND STANDARD DEVIATIONS OF GENDER PERFORMANCE DIFFERENCES ON THE SLP, GDS AND BEAD MEMORY 156 Iabje £1 Intercorrelations oi the GDS Delay and Vigilance Tasks and jkad Memory fSBFR) GDS Delay Tasks Total Total Block 1 Block 1 Block 2 Block 2 Block 3 Block 3 Block 4 Block 4 GDS Vigilance Tasks Group Correct Responses Correct Responses Correct Responses Correct Responses Correct Responses Total Correct Responses Control .331* -.084 .076 -.123 .241 -.219 .382* -.096 .393* .171 High Risk .401* -.518 .108 -.625** .294 -.532** .566** -.370* .316* .001 Block 1 Correct Control .273 -.150 .163 .032 .155 -.281 .308 -.251 .264 -.007 High Risk .343* -.367* .066 -.454* .227 -.358* .488* -.211 .311* -.068 Omission Control -.273 .150 -.163 -.032 -.155 .281 -.308 .251 -.264 .007 High Risk -.343* .367* -.066 .454* -.227 .358* -.488* .211 -.311* .068 Commission Control .204 -.109 .153 -.196 .128 -.125 .252 .035 .129 -.009 High Risk .169 .190 .127 .214 .009 .125 .103 .171 .281 .076 Block 2 Correct Control .154 -.085 -.033 -.096 .059 -.233 .227 -.140 .266 .205 High Risk .255 -.521* .046 -.582** .074 -.546** .341* -.453* .338* .055 Omission Control -.154 -.085 .033 .096 -.059 .233 -.227 .140 -.266 -.205 High Risk -.255 .521* -.046 .582** -.074 .546** -.341* .453* -.338* -.055 Commission Control .199 -.195 .193 -.205 .210 -.121 . .156 -.151 .076 -.187 High Risk :199 .108 .138 .161 .118 .048 .107 .066 .267 .067 Block 3 Correct Control .298 .019 .055 -.172 .292 -.016 .306 .114 .325* .151 High Risk .416* -.431* .162 -.552* .445* -.452* .602** -.281 .146 .022 Omission Control -.298 -.019 -.055 .172 -.292 .016 -.306 -.114 -.325* -.151 High Risk -.416* .431* -.162 .552* -.445* .452* -.602** .281 -.146 -.022 Commission Control .037 -.021 .018 -.058 .007 -.011 .008 -.031 .090 .038 High Risk .185 .154 .082 .093 .156 .056 .068 .164 .278 .183 Total Commission Errors Control .187 -.127 .150 -.191 .138 -.111 .191 -.075 .125 -.048 High Risk .196 .157 .130 .164 .092 .079 .098 .133 .292 .117 Bead Memory Control .076 -.200 .141 -.295 .057 -.149 .106 -.082 -.068 -.155 High Risk -.149 -.011 .083 -.217 -.064 .003 -.035 .027 -.442* .162 Table El continued GDS Delay Tasks Total ER Block Slope Block 1 Block 2 Block 3 Block 4 Bead GDS Vigilance Tasks Group ER Variability Scope ER ER ER ER Memory Total Correct Responses Control .316 .016 -.167 .241 .428* .300 .031 -.167 High Risk .519* .176 -.134 .646** .563** .651** .389* .219 Block 1 Correct Control .361* .085 .150 .138 .402* .460* .201 -.088 High Risk .459* .120 -.017 .473* .388* .447* .387* .140 Omission Control -.361* -.085 -.150 .138 -.402* -.460 -.201 .088 High Risk -.459* -.120 .017 -.473* -.389* -.447* -.387* -.140 Commission Control .264 -.581** -.213 .320* .209 .135 .087 -.100 High Risk -.025 -.146 .106 -.171 -.166 -.120 -.030 -.335* Block 2 Correct Control .199 .149 -.080 .146 .281 .269 -.053 -.212 High Risk .572** .174 -.198 .667** .536** .637** .333* .185 Omission Control -.199 -.149 .080 -.146 -.281 -.269 .053 .212 High Risk -.572** -.174 .198 -.667** -.536** -.637** -.333* -.185 Commission Control .349* -.493* -.110 .367* .270 .195 .271 .103 High Risk -.033 -.212 .073 -.145 -.062 -.023 -.003 -.254 Block 3 Correct Control .171 -.155 -.350* .231 .281 .012 -.038 -.072 High Risk .281 .154 -.134 .502* .508* .573** .259 .232 Omission Control -.171 .155 .350 -.231 -.281 -.012 .038 .072 High Risk -.281 -.154 .134 -.502* -.508* -.573** -.259 -.232 Commission Control -.023 -.337* -.023 -.045 -.037 -.075 .013 -.215 High Risk -.117 -.143 -.037 -.095 -.044 -.117 .082 -.270 Total Commission Errors Control .245 -.588** -.160 .275 .186 .109 .135 -.098 High Risk -.060 -.168 .053 -.128 -.103 -.083 -.045 -.307* Bead Memory Control .272 -.037 -.225 .423* .176 .190 .085 High Risk -.082 .058 -.184 .168 -.062 -.039 -.071 •p<0.05 "pSO.001 lahls £2 M__S (-Standard Deviations) iy Group and _£2 and Analysis _f Variance ior _ _ _Q _ _ _ _ learning Profile (SLE) CONTROL "HIGH RISK" SLP Total1 Boys1 Girls' Total4 Boys5 Girls' MSe F(1.52) 1. Overall Learning Ability 3.929 (0.766) 3.733 (0.799) 4.154 (0.689) 1.000 (0.000) 1.000 (0.000) 1.000 (0.000) 0.64 2.28 2. Attention Span & Distractibility 4.107 (1.031) 3.733 (1.163) 4.538 (0.660) 1.968 (0.850) 2.000 (0.882) 1.909 (0.831) 2.66 3.20 13A. Fidgets 1_14 (0.832) 1.333 (0.976) 1.077 (0.640) 2.345 (1.203) 2.316 (1.204) 2.400 (1.265) 0.22 0.20 B. Difficulty Staying Seated 0.929 (0.979) 1.000 (1.195) 0.846 (0.689) 2.100 (1.269) 2.158 (1.259) 2.000 (1.342) 0.26 0.19 C. Difficulty Waiting Turn 0.893 (1.066) 1.000 (1.254) 0.769 (0.832) 1.633 (1.402) 1.684 (1.455) 1.545 (1.368) 0.21 0.13 D. Easily Distracted 1.286 (1.013) 1.600 (1.183) 0.923 (0.640) 3.000 (0.983) 3.053 (0.848) 2.909 (1.221) 2.34 2.42 E. Defiant & Uncooperative 0.286 (0.600) 0.400 (0.737) 0.154 (0.375) 0.900 (1.213) 1.105 (1.449) 0.545 (0.522) 2.26 2.45 F. Temper Tantrums 0.143 (0.448) 0.200 (0.561) 0.077 (0.277) 0.267 (0.691) 0.316 (0.820) 0.182 (0.404) 0.23 0.65 G. Difficulty Listening 0.893 (1.066) 1.133 (1.302) 0.615 (0.650) 2.700 (1.119) 2.632 (0.955) 2.818 (1.401) 0.38 0.32 H. Difficulty Playing Quietly 0.821 (0.819) 0.800 (0.941) 0.846 (0.689) 1.500 (1.253) 1.684 (1.335) 1.182 (1.079) 0.73 0.63 I. Short Attention Span 0.679 (0.905) 1.133 (0.990) 0.154 (0.375) 2.333 (1.155) 2.263 (1.194) 2.454 (1.128) 2.16 2.17 J. Blurts Out Answers 0.786 (0.995) 0.733 (1.100) 0.846 (0.899) 1.267 (1.285) 1.105 (1.197) 1.545 (1.440) 1.07 0.79 Total Behaviour Score 7.929 (7.086) 9.333 (8.707) 6.308 (4.385) 17.967 (8.487) 18.316 (8.750) 17.364 (8.394) 55.10 0.88 *p^  0.0038, with Bonferroni adjustment _=28 2n=15 'n=13 <n = 30 5n=19 Tahle F3 Means (Standard Deviations) bv Group and Sex on the GDS Delay Task CONTROL "HIGH RISK" GDS Delay Task Total1 Boys1 Girls' Total' Boys5 Girls' Total Correct Responses 44.786 (11.133) 44.867 ( 6.621) 44.692 (15.091) 34.333 (10.060) 36.316 ( 8.076) 31.909 (12.478) Total Responses 68.607 (21.226) 67.533 (13.674) 69.864 (28.151) 71.400 (28.657) 72.684 (24.768) 69.182 (35.620) Block 1 Correct 10.893 ( 3.685) 10.933 ( 3.081) 10.864 ( 4.413) 8.667 ( 3.231) 9.474 ( 2.875) 7.273 ( 3.467) Responses 16.929 ( 6.588) 17.333 ( 6.275) 16.461 ( 7.160) 16.767 ( 8.881) 17.790 ( 8.734) 15.000 ( 9.274) Block 2 Correct 11.321 ( 3.560) 10.933 ( 1.668) 11.770 ( 4.986) 8.833 ( 2.878) 9.368 ( 2.338) 7.909 ( 3.562) Responses 17.179 ( 5.831) 16.733 ( 4.080) 17.692 ( 7.521) 19.433 (11.358) 19.158 ( 8.408) 19.909 (15.700) Block 3 Correct 11.107 ( 3.270) 11.067 ( 2.344) 11.154 ( 4.200) 8.300 ( 3.153) 8.526 ( 2.756) 7.909 ( 3.859) Responses 17.393 ( 6.839) 16.600 ( 3.960) 18.308 ( 9.232) 17.800 ( 8.829) 17.210 ( 8.059) 18.818 (10.362) Block 4 Correct 11.464 ( 3.305) 11.933 ( 2.434) 10.923 ( 4.132) 8.533 ( 3.340) 8.947 ( 3.100) 7.818 ( 3.763) Responses 17.107 ( 5.593) 16.867 ( 3.870) 17.385 ( 7.263) 17.400 ( 8.177) 18.526 ( 7.344) 15.454 ( 9.501) Total ER 0.693 ( 0.182) 0.681 ( 0.120) 0.707 ( 0.240) 0.588 ( 0.236) 0.596 ( 0.211) 0.743 ( 0.285) ER Block Variability 0.121 ( 0.059) 0.135 ( 0.062) 0.104 ( 0.052) 0.145 ( 0.088) 0.137 ( 0.084) 0.158 ( 0.098) Slope Scope 0.001 ( 0.164) 0.021 ( 0.183) -0.021 ( 0.142) -0.038 ( 0.243) -0.020 ( 0.245) -0.070 ( 0.248) Block 1 ER 0.708 ( 0.227) 0.689 ( 0.234) 0.730 ( 0.226) 0.645 ( 0.278) 0.643 ( 0.259) 0.649 ( 0.321) Block 2 ER 0.702 ( 0.206) 0.682 ( 0.146) 0.725 ( 0.264) 0.585 ( 0.269) 0.570 ( 0.235) 0.610 ( 0.331) Block 3 ER 0.708 ( 0.227) 0.696 ( 0.176) 0.718 ( 0.282) 0.570 ( 0.270) 0.580 ( 0.264) 0.553 ( 0.294) Block 4 ER 0.717 ( 0.211) 0.733 ( 0.159) 0.698 ( 0.264) 0.578 ( 0.245) 0.555 ( 0.250) 0.617 ( 0.242) '11=28 Jn=15 5n=13 4n=30 sn=19 «n=n o Table E4 Means fStandaid Deviations) by Group and Sex on the GDS Vigilance Task CONTROL "HIGH RISK" GDS Vigilance Task Total1 Boys' Girls' Total4 Boys5 Girls' Total Correct Responses 26.643 ( 2.642) 27.000 ( 2.591) 26.231 ( 2.743) 21.567 ( 5.618) 20.474 ( 5.853) 23.454 ( 4.865) Block 1 Correct 11.357 ( 0.989) 11.467 ( 0.743) 11.231 ( 1-235) 10.167 ( 2.365) 9.842 ( 2.713) 10.727 ( 1.555) Omission 0.643 ( 0.989) 0.533 ( 0.743) 0.769 ( 1-235) 1.833 ( 2.365) 2.158 ( 2.713) 1.273 ( 1.555) Commission 2.464 ( 3.305) 2.600 ( 3.291) 2.308 ( 3.449) 7.200 ( 8.849) 8.895 (10.322) 4.273 ( 4.519) Block 2 Correct 8.357 ( 1.193) 8.400 ( 1.056) 8.308 ( 1.377) 6.233 ( 2.144) 5.947 ( 2.248) 6.727 ( 1.954) Omission 0.643 ( 1-193) 0.600 ( 1.056) 0.692 ( 1.377) 2.768 ( 2.144) 3.053 ( 2.248) 2.273 ( 1.954) Commission 0.929 ( 1.762) 0.600 ( 1.121) 1.308 ( 2.287) 4.267 ( 8.674) 5.789 (10.523) 1.636 ( 2.618) Block 3 Correct 6.929 ( 1.412) 7.133 ( 1.457) 6.692 ( 1-377) 5.167 ( 2.151) 4.684 ( 2.001) 6.000 ( 2.236) Omission 1.071 ( 1.412) 0.867 ( 1.457) 1.308 ( 1.377) 2.833 ( 2.151) 3.316 ( 2.001) 2.000 ( 2.236) Commission 0.857 ( 1.976) 1.467 ( 2.560) 0.154 ( 0.375) 4.200 ( 7.911) 5.158 ( 9.610) 2.545 ( 3.236) Total Commission Errors 4.250 ( 5.873) 4.667 ( 6.366) 3.769 ( 5.464) 15.867 (24.093) 20.158 (28.851) 8.454 ( 9.470) Bead Memory 49.429 ( 7.346) 47.467 ( 7.891) 51.692 ( 4.197) 39.167 ( 5.113) 38.368 ( 5.068) 40.545 ( 5.126) _=28 _=15 '_=13 4_=30 5n=l9 « _ = ] Ll 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

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"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0098724/manifest

Comment

Related Items