Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Relationship between attention and social functioning in individuals with schizophrenia or schizoaffective… Ott, Fred Julius 2001

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

Item Metadata

Download

Media
831-ubc_2001-0486.pdf [ 7.19MB ]
Metadata
JSON: 831-1.0090139.json
JSON-LD: 831-1.0090139-ld.json
RDF/XML (Pretty): 831-1.0090139-rdf.xml
RDF/JSON: 831-1.0090139-rdf.json
Turtle: 831-1.0090139-turtle.txt
N-Triples: 831-1.0090139-rdf-ntriples.txt
Original Record: 831-1.0090139-source.json
Full Text
831-1.0090139-fulltext.txt
Citation
831-1.0090139.ris

Full Text

RELATIONSHIP B E T W E E N ATTENTION A N D SOCIAL FUNCTIONING IN INDIVIDUALS WITH SCHIZOPHRENIA OR SCHIZOAFFECTIVE DISORDER by FRED JULIUS OTT B. A. , University of British Columbia, 1985 B.Sc. (OT), University of British Columbia, 1988 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE F A C U L T Y OF G R A D U A T E STUDIES (School of Rehabilitation Sciences) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A August 2001 © Fred Julius Ott, 2001 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. D e ^ a r t » ^ o f ^yh^JjAJ^rry Afi/tsyTM The University of British Columbia Vancouver, Canada Date DE-6 (2788) Relationship between Attention and Social Functioning ii Abstract There is growing evidence suggesting a relationship between cognition and social functioning in individuals with schizophrenia. Researchers have theorized that cognitive deficits may impair the ability of an individual with schizophrenia or schizoaffective disorder to recognize interpersonal cues and to process other information necessary for effective social functioning. Impairments of social functioning can affect any of the individual's occupational performance areas: self-care, productivity, or leisure. Individuals with schizophrenia experience significant information-processing deficits in a wide range of cognitive processes including attention, memory, reasoning ability and language. Yet, there is relatively little consensus as to which cognitive processes may be related to social functioning. The purpose of the study was to investigate the relationship between attention and social functioning in a sample of 35 individuals with schizophrenia or schizoaffective disorder. This study employed a cross-sectional design. Attention was defined as the rate of information processing and was evaluated by the Paced Auditory Serial-Addition Task (PASAT). The PAS AT measured the rate of information processing by determining the time required per correct response. Social functioning was evaluated with the Social Dysfunction Index (SDI). The SDI is a 27-item semi-structured interview, developed and validated on individuals with schizophrenia and schizoaffective disorder. The SDI provided an overall social dysfunction score, a satisfaction score, and a score for each of the nine components of dysfunction assessed. A Pearson correlation analysis found a small but statistically insignificant relationship between the SDI overall percent scores and the PASAT mean time/response scores (r = 0.086, p = 0.624). A Welch's approximate t-test determined that the PASAT mean time/response scores of the present study's subjects were significantly lower than those of Gronwall's normal control Relationship between Attention and Social Functioning subjects (t (34) = 3.59, p < 0.05). Therefore, the PASAT may be an effective tool for assessing attention impairments in individuals with schizophrenia. The findings of this study have indicated that the manner in which attention and social functioning are defined and measured needs to be considered when examining these complex concepts. The study supports the need for further investigation in this area. Relationships between specific cognitive deficits and impairments in activities of daily living must be made before remediation of cognitive deficits is attempted. Relationship between Attention and Social Functioning T A B L E OF CONTENTS Abstract ii Table of Contents iv List of Appendices vii List of Tables viii List of Figures ix Acknowledgements x CHAPTER I Introduction 1 1.1 Statement of the problem 3 1.2 Purpose of the study 5 CHAPTER II Literature Review 8 2.1 Relationships between cognition and social functioning in individuals with schizophrenia 9 2.2 Models and measures of attention 23 2.3 Measures to assess social functioning 32 2.4 Conclusion 35 Relationship between Attention and Social Functioning CHAPTER UI Methods 36 3.1 Subjects 36 3.2 Variables 44 3.3 Test administration 49 3.4 Design & procedure 53 3.5 Data analysis 54 CHAPTER TV Results 56 4.1 Pearson correlation between SDI overall percent scores and PAS A T mean time/response scores 57 4.2 Coefficient of determination 57 4.3 Calculation of power 57 4.4 Confidence interval around the correlation coefficient 58 4.5 Calculation of sample size 59 4.6 Pearson correlation between the occupational functioning subcomponent of the SDI and PASAT mean time/response scores 60 Relationship between Attention and Social Functioning vi CHAPTER V Discussion 62 5.1 Factors associated with PAS A T and SDI that may have contributed to findings 66 5.2 Strengths and Limitations of the study 75 5.3 Suggestion for future research and implication of findings 77 5.4 Conclusion 80 References 82 Relationship between Attention and Social Functioning vii LIST OF APPENDICES Appendix A Summary of Studies that Examined the Relationships between Cognition and Social Functioning in Individuals with Schizophrenia 101 Appendix B Limitations of Studies Examining the Relationships between Cognition and Social Functioning in Individuals with Schizophrenia I l l Appendix C SPSS printout of Pearson correlation analyses 114 Appendix D SPSS printout of Independent t-test analyses 117 Appendix E Sample items from Social Dysfunction Index 123 Appendix F Reliability Analysis - Intraclass Correlation Coefficient for SDI Raters 124 Appendix G Reliability Analysis - Cohen's Kappa Output for SDI Raters 127 Appendix H SPSS printout of Pearson correlation analysis of the relationship between the SDI overall percent and PAS A T mean time/response scores: 141 Appendix I SPSS printout of Pearson correlation analysis of the relationship between the total percent occupational functioning subcomponent scores of the SDI and PAS AT mean time/response scores 142 Appendix J Strength of relationship between attention and social functioning in previous research 143 Appendix K Summary of Green et al.'s (2000) Meta-analysis of key neurocognitive Constructs 147 Appendix L Review of frequently cited measures of attention 148 Relationship between Attention and Social Functioning viii LIST OF TABLES Table 1 Sohlberg & Mateer's Model of Attention with Proposed Measures to Assess Levels of Attention 29 Table 2 Descriptive Statistics of Subjects' Demographic Information 39 Table 3 Pearson Correlation Analyses between Subjects' Demographic Information and SDI overall percent scores and PASAT mean time/response Scores 40 Table 4 Frequency Count of Subjects' Demographic Information 41 Table 5 Independent t -test Analysis between Subjects' Demographic Information and SDI overall percent scores and P A S A T mean time/response scores 44 Table 6 Social Dysfunction Index Component Scoring 48 Table 7 Summary of ICC and Cohens Kappa scores for the SDI raters 52 Table 8 Descriptive Statistics for the Present Study Subjects and Gronwall's (1977) Control Group 68 Table 9 Studies involved in the development of the SDI 73 Relationship between Attention and Social Functioning ix LIST OF FIGURES Figure 1 Model of Social Functioning 92 Figure 2 Hypothesized role of attention deficits in the development of social dysfunctions among person with a genetic susceptibility to schizophrenia 93 Figure 3 Scatter Plots of the relationships between subjects' age and SDI overall percent scores and PAS AT mean time/response scores 94 Figure 4 Scatter plots of the relationships between subjects' level of education and SDI overall percent scores and PASAT mean time/response scores 95 Figure 5 Scatter plots of the relationships between the number of months subjects were diagnosed and the SDI overall percent scores and PASAT mean time/response scores 96 Figure 6 Boxplots of the relationships between the subjects' gender and SDI overall percent scores and PASAT mean time/response scores 97 Figure 7 Boxplots of the relationships between the subjects' diagnosis and SDI overall percent scores and PASAT mean time/response scores 98 Figure 8 Boxplots of the relationships between the subjects' medication dosage levels and SDI overall percent scores and PASAT mean time/response scores 99 Figure 9 Scatter Plot displaying relationship between SDI overall percent scores and PASAT mean time/response scores 100 Relationship between Attention and Social Functioning x A C K N O W L E D G E M E N T S There are many people to whom I would like to express my appreciation for their contribution to this thesis: Dr. Lyn Jongbloed, my research advisor, for her expert guidance and unwavering support during the development of this project; Dr. Susan Harris, for her supportive feedback and encouragement; Dr. Peter Liddle, for his expertise in schizophrenia research and provision of thoughtful and relevant comments; Dr. Jonathan Money, for his invaluable advice about statistical analysis; Dr. David Irwin, Annie Kuan, LauraLynn Reinhardt, Rosanne Wozny, Miriam Cohen & Otto Lim, for their untiring and highly valued assistance; The clients of the U B C Schizophrenia Rehabilitation Day Program for their cooperation during data collection; The BC Schizophrenia Foundation, for their generous financial assistance; My mother & family, for their constant words of support; and last but not at all least, Beth Ott, my loving wife, for her understanding and invaluable support. Relationship between Attention and Social Functioning 1 CHAPTER 1 Introduction Schizophrenia is a serious psychiatric disorder that affects an estimated one person in a hundred (Bellack, Morrison, Wixten, & Mueser, 1990; Davison & Neale, 1982; Morrison, 1990; Perm, Mueser, Spaulding, Hope, & Reed, 1995). Deterioration in social functioning is a central feature of schizophrenia (Allen, 1995; Bellack et al., 1990; Dworkin, 1992; Dworkin, et al., 1991; Kelly, Urey & Patterson, 1980; Kielhofner, 1985; Liberman, DeRisi, & Mueser, 1989; Massel, Corrigan, Liberman, & Milan, 1991; Pan & Fisher, 1994; Perm, et al., 1995), and is one of the defining diagnostic criteria specified in the DSM- iV (American Psychiatric Association, 1994). Social isolation or withdrawal and marked impairment in major life role functioning are recognized as prodromal and residual symptoms of the illness (Bellack et al., 1990; Liddle, 1987; Perm et al., 1995). Deterioration in social functioning has been thought to result from the positive and negative symptoms of the disorder (Ancil, Holliday, & Higenbottam, 1994; Green, 1993). Research has demonstrated that individuals with schizophrenia who exhibit severe negative symptoms, such as avolition, anhedonia and alogia, are more socially impaired than individuals with less marked negative symptoms (Andreasen, 1982; Bellack et al., 1990; Crow, 1985; Dickerson, Boronow, Ringel, & Parente, 1996; Morrison, 1990). As an alternative to the positive-negative dichotomy of symptom classification, Liddle (1987) classified symptoms into three syndromes: (a) psychomotor poverty (poverty of speech, lack of spontaneous movement, and various aspects of blunting of affect); (b) disorganization (inappropriate affect, poverty of content of speech, and disturbances of the form of thought); and (c) reality distortion (particular types of delusions and hallucinations). Liddle (1987) found that individuals with schizophrenia who exhibited psychomotor poverty syndrome also tended to have poor pre-morbid social competence and impaired social relationships. Relationship between Attention and Social Functioning 2 Ample evidence now exists to suggest that impairments in social functioning are not simply secondary consequences of negative and/or positive symptoms (Ancil et al., 1994; Bellack et al., 1990; Green, 1993; Harvey, Sukhodolsky, Parrella, White, & Davidson, 1997; Jackson, Minas, Burgess, Joshua, Charisiou, & Campbell, 1989; Liberman & Green; 1992; Perm et al., 1995; Tryssenaar & Goldberg, 1994). Bellack et al. (1990) argued that negative symptoms and social impairments could be mediated by some third variable. Recent studies suggest that the cognitive deficits occurring in schizophrenia are possible underlying factors of impaired social function and skill acquisition (Lysaker, Bell, Zito, & Bioty, 1995; Perm et al., 1995). Even with the implementation of medication regimes, cognitive deficits continued to persist long after the positive and negative symptoms subsided (Ancil et al., 1994; Green, 1993; Perm et al., 1995; Rund, 1998). Crow (1985) theorized that the cognitive deficits associated with negative symptoms interfere with skill acquisition. Individuals with schizophrenia experience significant information-processing deficits in a wide range of cognitive processes including memory, attention, reasoning ability and language (Allen, 1995; Ancil et al., 1994; Bellack, 1992; Braff, Saccuzzo & Geyer, 1991; Kietzman, 1991; Nuechterlein & Asarnow, 1989; Perry & Braff, 1994; Tryssenaar & Goldberg, 1994). Many authors have suggested that the initial stage of information processing, such as attention, underlies the information-processing deficits (Bowen, Wallace, Glynn, Nuechterlein, Lutzker, & Kuehnel, 1994; Corrigan, Green, & Toomey, 1994; Duchek, 1991; Perm et al., 1995; Perry & Braff, 1994; Van Zomeren & Brouwer, 1994). McKee, Hull, and Smith (1997) found that deficits in attention, memory, and executive functioning were associated with impairments in social skill acquisition and social functioning. Although several studies have suggested that attention impairments are associated with social functioning impairments (Freedman, Rock, Roberts, Cornblatt, & Erlenmeyer-Kimling, 1998; Kern, Green, Satz, 1992; McKee et al., 1997; Relationship between Attention and Social Functioning 3 Perm et al., 1995) other studies have not supported this finding (Addington & Addington, 1999; Dickerson et al., 1996). Green (1996) reviewed 17 studies that examined the relationships between cognitive deficits and functional outcomes in schizophrenia. The review included studies that evaluated neuropsychological measures as predictors and correlates of the following functional outcomes: (a) community functioning (social and occupational); (b) social problem solving; and (c) social skill acquisition. Despite wide variation in the selection of neuropsychological measures and limited statistical power, Green (1996) found that vigilance was positively associated with social-problem solving and skill acquisition. Green (1996) argued that vigilance was an "attentional process"; he defined vigilance as "the ability to discriminate signal (target) from noise (nontarget) stimuli across an entire vigilance period" (p. 326). He also noted that none of the community functioning studies evaluated vigilance. Green (1996) hypothesized that the same measures that predicted social skill acquisition, such as vigilance, would also predict social functioning in the community. Statement of the Problem Although a number of studies have demonstrated relationships between cognitive processes and aspects of social functioning, the mechanisms by which cognitive impairments limit social functioning are not clearly understood. Researchers have speculated that attention deficits could impair an individual's ability to recognize interpersonal cues and process other information necessary for effective social interaction (Corrigan et al., 1994; Morrison, 1990). Wallace, Nelson, Liberman, Aitchison, Lukoff, Elder, & Ferris (1980) proposed a sequential model of social functioning (see Figure 1). According to this model, impaired "receiving" skills will affect performance on subsequent "processing" and "sending" skills (Wallace et al., 1980). Addington & Addington (1999) noted that Wallace et al.'s (1980) model of social functioning supported the theory that deficits in early information processing could influence later cognitive Relationship between Attention and Social Functioning 4 processing. For example, attention deficits could impair the recognition or "receiving" of social cues. Processing of limited social cues could jeopardize the "sending" of appropriate social behaviors. Wallace et al.'s model of social functioning is consistent with Cornblatt and Keilp's (1994) hypothesis (see Figure 2) that attention deficits compromise the social functioning among individuals with schizophrenia. Attention deficits can impair an individual's ability to receive and process relevant social information. The individual may then either avoid social contact, resulting in decreased stress and symptom reduction, or may attempt social contact resulting in increased stress and symptom exacerbation (Cornblatt & Keilp, 1994). Thus, Wallace et al.'s (1980) sequential model of social functioning and Cornblatt and Keilp's (1994) hypothesized role of attention deficits in the development of social dysfunctions among individuals at risk of schizophrenia seem to support Green's (1996) hypothesis that deficits in attention would impair an individual's social functioning within the community he or she resides. Although Cornblatt and Keilp (1994) proposed a model of how attention deficits could impair social functioning in individuals at risk for schizophrenia, their research did not actually assess the attention deficit behaviors theoretically associated with social functioning impairments. Cognition has been identified as a major determinant of occupational functioning (Lysaker et al., 1995) and guides all aspects of performance. Therefore, treating cognitive dysfunction is well within the occupational therapist's (OT's) professional domain (DePoy & Burke, 1992; Duchek, 1991). According to the Canadian Model of Occupational Performance, (CMOP; Canadian Association of Occupational Therapy, 1997) a dynamic relationship exists between a person, his/her occupations and roles, and the environment in which he/she lives. The CMOP presents the person "as an integrated whole who incorporates spirituality, social and cultural experiences, and observable occupational performance components" (pg. 41-42). The three occupational performance components of a person are: (a) affective, (b) cognitive, and (c) Relationship between Attention and Social Functioning 5 physical. The cognitive performance component of the person is defined as "the domain that comprises all mental functions both cognitive and intellectual, and includes, among other things, perception, concentration, memory, comprehension, judgment and reasoning" (pg. 44). Although treating cognitive dysfunction is well within the OT's professional domain, OTs in mental health typically teach behavioral strategies that bypass deficits in functional areas, rather than attempt to remediate deficiencies in cognitive abilities (Allen, 1995; Averbuch & Katz, 1992; Tryssenaar & Goldberg, 1994). The hesitancy of OTs to remediate cognitive deficiencies occurring in schizophrenia may be related to the lack of agreement about which "core" cognitive impairments underlie the diverse information-processing deficits (Bellack, 1992; Liberman & Green, 1992). Without a clear understanding of which specific cognitive impairments are related to functioning in activities of daily living, efforts to remediate cognitive impairments would be unfocused (Green, 1996). Purpose of the Study Attention deficits commonly observed in individuals with schizophrenia include: (a) inability to selectively attend to relevant stimulation and filter out irrelevant stimulation, (b) inability to sustain attention over a period of time, and (c) difficulty shifting attention between stimuli (Braff et al., 1991; Cullum et al., 1993; Kietzman, 1991). Any of these attention deficits may impair learning and daily functioning of the individual with schizophrenia. A better understanding of the attention deficits individuals with schizophrenia experience and how these deficits interfere with functional behavior, can offer direction for more comprehensive and effective interventions (Ducheck, 1991; Green, 1993; Green, 1996; Penn et al., 1995; Spaulding, 1992). The present study was designed to document the relationship between attention and social functioning in individuals with schizophrenia or schizoaffective disorder. The instrument used to assess attention in this study was the Paced Auditory Serial Addition Task (PASAT; Relationship between Attention and Social Functioning 6 Gronwall, 1977). The P A S A T has been identified as an appropriate measure to diagnosis attention impairments and assess treatment efficacy of attention remediation protocols (Gronwall, 1977; Sohlberg & Mateer, 1989). For the purpose of this study, the construct of attention was defined as the rate of information processing as measured by the PASAT. The instrument used to assess social functioning in this study was the Social Dysfunction Index (SDI). The SDI is a semi-structured interview designed to measure social dysfunction in individuals with schizophrenia and schizoaffective disorder (Munroe-Blum, Collins, McCleary, & Nuttall, 1996). Social functioning was defined as the individual's ability to maintain desired roles and relationships within his or her community as measured by the SDI. Examples of social functioning areas assessed by the SDI include ability to present self in public and maintain relationships other than family. This study differed from other studies that examined the relationships between cognition and social functioning in individuals with schizophrenia or schizoaffective disorder in that: (a) attention was the only aspect of cognition assessed, and (b) attention was defined as the rate of information processing and measured with the PASAT. Results of this study may add to the theoretical understanding of how attention and social functioning in individuals with schizophrenia are related and offer direction for more comprehensive and effective interventions. Green (1996) argued that relationships between specific cognitive deficits and impairments in activities of daily living must be made before remediation of cognitive deficits is attempted. Future research could investigate the efficacy of attention remediation protocols designed to improve the rate of information processing on social functioning using the PASAT and SDI as measurement tools. The following hypotheses were developed based on the review of the literature: H 0 : There will be no relationship between attention and social functioning in individuals with schizophrenia or schizoaffective disorder. Relationship between Attention and Social Functioning H i : There will be a significant positive relationship between attention and social functioning in individuals with schizophrenia or schizoaffective disorder. Relationship between Attention and Social Functioning 8 CHAPTER 2 Literature Review Impairments in cognition and social functioning have been well documented in individuals with schizophrenia (Green & Nuechterlein, 1999). There is also a growing body of literature suggesting that cognition and social functioning are correlated. However, there is no consensus as to which aspect of cognition is most closely related to social functioning (Dickerson et al., 1996). Addington and Addington (1999) argued that generalization of cognitive remediation efforts to improve social functioning would be enhanced if the cognitive processes responsible for the mediation of social functioning were targeted for cognitive remediation. Because little is known about precisely which cognitive deficits compromise social functioning in individuals with schizophrenia, it is not clear which cognitive process or processes should be targeted for rehabilitation (Green, 1996). The purposes of this chapter are to: (a) review the literature that has examined the relationships between cognition and social functioning in individuals with schizophrenia or schizoaffective disorder; (b) review the construct of attention and how it has been defined and measured in the literature; and (c) discuss some of the tools used to evaluate social functioning in studies that examined the relationship between cognition and social functioning in individuals with schizophrenia. Several studies have suggested that attention is correlated with social functioning. However, other studies have suggested that other aspects of cognition are correlated with social functioning. Since there is no consensus as to which aspect of cognition is most closely related to social functioning in individuals with schizophrenia, the review of the literature will include studies that have examined the relationships between various aspects of cognition, including attention, and social functioning. Summaries of the measures used to assess cognition and social functioning in the literature are displayed in Appendix A. Relationship between Attention and Social Functioning 9 The literature selected for this review included citations listed in the following University of British Columbia Library OVID databases: (a) CINAHL, from 1982 to 1999; and (b) MEDLINE, from 1960 to 1999. Key words used in the search included: schizophrenia, neurocognition, cognitive functioning, cognition, information processing, social dysfunction, social skills, and social functioning. Relationships between Cognition and Social Functioning in Individuals with Schizophrenia Cognitive dysfunction has been recognized as a pathognomonic characteristic of schizophrenia. Brenner, Hodel, Roder, and Corrigan (1992) have proposed several models that could explain the multiplicity of cognitive dysfunctions associated with schizophrenia. One model argued that incorrect attention and encoding of incoming information prevent accurate processing of information at higher levels of cognitive functioning, such as decision-making and response selection. A contrasting model proposed that deficits at higher levels of cognitive functioning diminish the ability of an individual with schizophrenia to encode and consolidate simpler information. Stress and vulnerability models suggest that sub-clinical cognitive vulnerabilities impede the acquisition of coping skills during premorbid years. Subtle information processing deficits may present even while the disease is in remission, consequently preventing accurate recognition of social cues and retrieval of appropriate responses (Brenner et al., 1992). To ameliorate cognitive and social dysfunctions associated with schizophrenia, Brenner et al. (1992) recommended the use of Integrated Psychological Therapy (EPT). IPT was comprised of five hierarchical intervention subprograms. "Early interventions" targeted basic cognitive skills. "Middle interventions" shaped cognitive skills into verbal and social responses. "Later interventions" trained individuals to solve more complex and challenging interpersonal problems. Remediation of elementary cognitive processes theoretically allowed higher order cognitive processes to integrate incoming information. Better integration of incoming Relationship between Attention and Social Functioning 10 information would improve an individual's ability to acquire social and coping skills. By acquiring more social and coping skills an individual is less vulnerable to social stressors and possible relapse of the illness (Brenner et al., 1992). Although the evaluative studies of IPT were cited as part of the 8 t h World Congress of Psychiatry proceedings in 1989, and specifics of the studies were unavailable, Brenner et al. (1992) argued that the IPT had favorable effects on elementary cognitive processes such as attention, abstraction, and concept formation. Evidence was mixed regarding LPT's effect on more complex cognitive skills that coordinate and integrate information. Many therapists have improved the social functioning of clients with schizophrenia by conducting social skills training programs (Massel et al., 1991). Social skills training (SST) often involves the use of instruction, modeling, rehearsal, feedback, and reinforcement to facilitate the acquisition of "microsocial skills", such as appropriate voice volume, eye contact and hand gestures, as well as "molar interpersonal skills", such as making a positive statement, requesting additional information, and resolving interpersonal problems (Massel et al., 1991). Despite positive findings, the characteristics of some client populations appear to impede SST efforts. Information processing deficits associated with schizophrenia persist during all phases of the illness (Massel et al., 1991). Distractibility and diminished information processing may hamper individuals' abilities to attend to SST instructional materials, to acquire social behaviors by observing models, and to comprehend the trainer's feedback (Massel et al., 1991). Massel et al. (1991) used a multiple baseline design across behaviors for 3 subjects to compare the acquisition and generalization of skills after a traditional SST program with the effects of an attention-focusing procedure (AFP). The AFP differed from the SST, in that the AFT used discrete trials and faded prompts to help individuals focus their attention during the training skills. Massel et al. (1991) found that AFP greatly facilitated the acquisition of Relationship between Attention and Social Functioning 11 conversational skills in thought-disordered individuals with schizophrenia for whom traditional SST had little effect. Both Brenner et al. (1992) and Massel et al. (1991) argued that a relationship between attention and social functioning in individuals with schizophrenia existed, however they neglected to mention how attention was assessed in their studies. In the present study, attention was assessed using the Paced Auditory Serial Addition Task (PASAT). Some studies found an association between attention impairments and social functioning by employing methods of assessing social functioning requiring direct observation of behavior. Kern et al. (1992) used a longitudinal study with a pretest-posttest design to examine the efficacy of neuropsychological measures to predict skill competence, skill acquisition, and task-related behaviors in 16 individuals with schizophrenia or schizoaffective disorder. According to Green (1996), skill competence and skill acquisition are integral aspects of social functioning. Difficulty acquiring or relearning psychosocial skills could compromise an individual's efforts to enhance his or her social functioning. The study employed a neuropsychological battery that sampled a wide range of cognitive and psychomotor abilities (Kern et al. 1992). Kern et al. (1992) argued that increased verbal learning, sustained attention, and decreased susceptibility to distraction were associated with better social functioning in individuals with schizophrenia or schizoaffective disorder. McKee et al. (1997) examined cognitive and symptom correlates of social functioning in 19 individuals diagnosed with chronic schizophrenia or schizoaffective disorder. They hypothesized that attention, memory, and executive function would be better predictors of social functioning than positive and negative symptoms. The study employed a neuropsychological battery that sampled a wide range of cognitive and psychomotor abilities (McKee et al., 1997). The Schedule for the Assessment of Positive Symptoms (SAPS) and the Schedule for the Assessment of Negative Symptoms (SANS) were used to assess the psychotic symptoms upon Relationship between Attention and Social Functioning 12 enrollment into the social skills training program. "Attendance" and "level of participation" were measures of social functioning. Using stepwise linear regressions, McKee et al. (1997) found that selective attention was a predictor of "level of participation", while negative symptoms and sustained attention were predictors of "attendance". Of the various neuropsychological measures used, Kern et al. (1992) and McKee et al. (1997) found that attention was most closely associated with social functioning. This evidence lends support to the hypothesis that the capacity to sustain attention free from distraction may allow individuals to receive and accurately process social information, thereby allowing them to interact more effectively. In a cross-sectional study, Dickerson et al. (1996) investigated the relationship between specific deficits in cognitive and social functioning in 88 stable outpatients with schizophrenia or schizoaffective disorder. Subjects were administered a battery of neuropsychological tests that sampled a variety of cognitive abilities. To assess social functioning, subjects were administered the Social Functioning Scale (SFS). Dickerson et al. (1996) found that the best neuropsychological predictors of social function were the subjects' performance on the aphasia screening, visual motor and spatial organization ability tests. Dickerson et al. (1996) employed the WAIS-R Digit Span and WAIS-R Arithmetic measures to assess attention and found no relationship between attention and social functioning. Kern et al. (1992) also employed the WAIS-R Digit Span measure to assess attention, and found a relationship between attention and social functioning. Inconsistent findings between studies that used similar measures suggest that further research is required to determine whether attention and social functioning are associated. Some researchers have employed longitudinal studies to examine the relationship of attention impairments in childhood with social dysfunctions in adulthood (Cornblatt, Lenzenweger, Dworkin, & Erlenmeyer-Kimling, 1992; Freedman'et al., 1998). Cornblatt et al. Relationship between Attention and Social Functioning 13 (1992) compared 39 unaffected adult offspring of parents with schizophrenia (high risk for schizophrenia individuals) with 86 normal control individuals. Subjects were recruited when the children were between 7 and 12 years of age and judged to be free of overt psychopathology. Abnormalities in childhood attention were assessed by means of the Attention Deviance Index (ADI). Social dysfunctions in adulthood were measured by the Personality Disorder Examination (PDE), a semi-structured interview assessing DSM-IJI-R Axis II personality disorders. The PDE was administered approximately 14 years after the attention data had been collected (Cornblatt & Keilp, 1994). Based on their findings, Cornblatt and Keilp (1994) proposed a model to describe the role of attention deficits in the development of social dysfunctions among individuals with a genetic susceptibility to schizophrenia (see Figure 2). According to Cornblatt and Keilp's (1994) model, impaired attention is likely a chronic disturbance throughout the development of an individual at risk for schizophrenia. The attention impairment leads to an inability to efficiently process information from the environment, especially subtle and highly complex interpersonal cues and communications. Deficiencies in processing social and interpersonal information make interactions with others increasingly difficult and stressful. Over time, unsuccessful attempts to initiate or maintain relationships can exacerbate symptoms and act as environmental triggers for the clinical expression of the illness. Alternatively, susceptible individuals may reduce stress and control their overt symptomatology by actively avoiding intense interpersonal contacts and situations requiring active relationships with other people. Therefore, Cornblatt and Keilp (1994) hypothesized that the active avoidance of others acts as a compensating mechanism that provides a measure of symptom control for some individuals predisposed to schizophrenia. Freedman et al. (1998) also employed a longitudinal study to examine the relationship of attention impairment in childhood with measures of physical anhedonia and social dysfunction in adulthood. Freedman et al. (1998) found that attention impairment in childhood was predictive Relationship between Attention and Social Functioning 14 of greater suspicious solitude and lack of empathy in adulthood. However, Freedman et al. (1998) argued that the relationship between attention impairment and social insecurity in adulthood was indirect and was mediated through physical anhedonia. While not all measures of social function were directly linked to early attention impairment, Freedman et al. (1998) concluded that attention impairment and physical anhedonia may play a specific role in the link between risk for schizophrenia and later social dysfunction. Although Cornblatt et al. (1992) and Freedman et al. (1998) used different measures to identify attention deficits in childhood, both groups of researchers found a relationship between attention deficits in childhood and social dysfunction in adulthood for individuals at risk of schizophrenia. In another longitudinal study, Perm, Van der Does, Spaulding, Garbin, Linszen, and Dingemans (1993) examined the relationships between information processing and social functioning in 24 individuals with schizophrenia and 24 normal individuals as controls. Perm et al. (1993) also found that sustained attention and cognitive flexibility were related to social functioning. In 1995, Perm et al. re-examined the relationship between information processing and social functioning among 38 individuals with chronic schizophrenia or schizoaffective disorder. However unlike Perm et al. (1993), Perm et al. (1995) assessed social functioning in the laboratory with an unstructured role-play test. Perm et al. (1995) found that higher global social functioning scores were associated with better performance on attention and reaction time tasks. Better nonverbal skills were associated with fewer errors on the concept manipulation task. Greater paralinguistic skill was associated with faster reaction times and with fewer errors on the concept manipulation task. Penn et al. (1995) suggested that non-verbal skills in schizophrenia such as poor eye contact, fidgeting, and restlessness may reflect internal states such as anxiety and distraction. Factors indicative of over arousal could compromise cognitive functioning. Relationship between Attention and Social Functioning 15 Thus, the relationship between nonverbal skills and errors on the concept manipulation task may be mediated by the capacity to focus and sustain attention. Poorer paralinguistic skills in the role-play were associated with slower reaction time scores. This relationship suggested that impaired paralinguistic skills, such as frequent pauses, speech dysfluency, and slow speech rate, were reflective of a longer latency to responses during a social encounter (Perm et al., 1995). Perm et al. (1995) concluded that the role of attention and reaction time in social functioning indicated that cognitive remediation should be directed at more of the early aspects of information processing such as attention. This finding supported Perm et al.'s (1993) earlier conclusion that attention is correlated with social functioning in individuals with schizophrenia. A key aspect of social functioning is the ability to perceive emotion in others. There is a growing body of literature suggesting that individuals with schizophrenia differ substantially from normal controls in processing emotional communications (Corrigan et al., 1994; Cramer, Bowen, & O'Neil, 1992; Ito, Shiragata, Kanno, Mori, Hoshino, & Niwa, 1998; Mandal, Pandey, & Prasad, 1998). Perception of emotion in schizophrenia is associated with basic neurocognition (Green & Nuechterlein, 1999). Deficits in the ability to perceive social cues or facial affect may contribute to impaired social adjustment and social skill learning. These impairments can lead to interpersonal tensions and abnormal behavior (Corrigan et al., 1994; Ito et al., 1998; Mandal et al., 1998). Researchers have hypothesized that social cue perception occurs in the early stages of information processing. Social cue perception has been conceptualized as the individual's ability to attend to and accurately encode the salient stimuli that describe interpersonal situations (Corrigan et al., 1994; Ito et al., 1998). Thus, some researchers have investigated the cognitive correlates to social cue perception in schizophrenia (Corrigan et al., 1994; Ito et al., 1998; Mandal et a l , 1998). Cramer et al. (1992) proposed three possible explanations for the inaccuracies individuals with schizophrenia make when attempting to judge the emotional states of others: (a) formal Relationship between Attention and Social Functioning 16 thought disorder, (b) selective avoidance of psychological factors that cause arousal and distress, and (c) perceptual or attention deficits that impair an individual's ability to organize meaningful material, including emotional factors, resulting in a failure to distinguish relevant from irrelevant material. To determine which of the three possible explanations accounted for the impairments in social functioning, Cramer et al. (1992) analyzed the verbal responses to questions regarding five videotaped social interactions of 34 in-patient subjects with the diagnosis of schizophrenia, and 34 controls who were matched for age and sex with the subjects. Cramer et al. (1992) argued that the pattern of responses were similar irrespective of the level of thought disorder, thus refuting the theory that individuals with schizophrenia displayed poor social judgment because of the effects of thought disorder alone. According to Cramer et al. (1992), the two findings in support of the theory that attention deficits impaired the individual's ability to organize meaningful material were: (a) individuals with schizophrenia, as a group, used fewer adjectives referring to the feeling state of the principal actor; and (b) individuals with schizophrenia were significantly less likely to comment on affect. Similar to Cramer et al. (1992), Corrigan et al. (1994) examined the relationships between cognitive functioning, social cue perception, and psychiatric symptoms in 26 individuals diagnosed with schizophrenia. Subjects were administered five neuropsychological tests that measured early input and more complex information-processing functions. To assess social functioning, subjects were administered the Social Cue Recognition Test (SCRT). Psychiatric symptoms were assessed by the Brief Psychiatric Scale (BPRS). Corrigan et al. (1994) found that social cue perception was positively correlated with early visual processing, recognition memory, and negative symptoms. A stepwise multiple regression determined that early visual processing, recognition memory, and negative symptoms independently accounted for a significant amount of variance in SCRT. Corrigan et al. (1994) argued that information-processing deficits and negative symptoms associated with Relationship between Attention and Social Functioning 17 schizophrenia might diminish individuals' sensitivity to social cues such that they mistakenly identify absent cues as being present. Although Corrigan et al. (1994) and Cramer et al. (1992) identified relationships between cognition and social cue perception, each group of researchers argued that different aspects of i cognition were correlated with social cue perception. Corrigan et al. (1994) suggested that early visual processing and recognition memory were correlated with social cue perception, yhile Cramer et al. (1992) suggested that attention was correlated with social cue perception., A limitation of Cramer et al.'s (1992) study was that no direct measures of attention or other cognitive functioning were made. Based on their findings, Cramer et al. (1992) deduced that attention was the core deficit in social cue perception. Addington, McCleary, and Munroe-Blum (1998) employed a cross-sectional study to examine the relationships between cognition and social function in 37 individuals with schizophrenia or schizoaffective disorder. Neuropsychological measures were selected iin order i to assess a range of cognitive functions (Addington et al., 1998). Unlike many other researchers, Addington et al. (1998) used direct and indirect methods of assessing social functioning. Social ! functioning was assessed indirectly with the Social Dysfunction Index (SDI) and the Social Adjustment Scale-II (SAS-II). Social functioning was assessed directly with The Assessment of Interpersonal Problem Solving Skills (AIPSS). i Addington and colleagues (1998) found that cognitive functioning did not generally I predict social functioning, as assessed by the SDI and the SAS-II. However, there was a significant positive correlation between low scores on processing skills and sending skills of the AIPSS and attention. The results of their study suggested that performance on an auditory attention task was correlated with social functioning of individuals with schizophrenia or schizoaffective disorder. This finding also lends support to the hypothesis that the capacity to Relationship between Attention and Social Functioning sustain attention free from distraction may allow individuals to receive and accurately process social information, thereby allowing them to interact more effectively. Addington and Addington (1999) also employed a cross-sectional research design to examine the relationship between cognitive and social functioning in a sample of 80 outpatients with a DSM-III-R diagnosis of schizophrenia. Although they used similar measures of jcognitive and social functioning, Addington and Addington (1999) obtained different results than Addington et al. (1998). Addington and Addington (1999) concluded that the 'receiving', 'processing', and 'sending' subscales of the AEPSS were correlated with verbal ability, yerbal memory, and cognitive flexibility. However, Addington et al. (1998) concluded that the i 'processing' and 'sending' subscales of the AIPSS were correlated with attention. Inconsistency in findings between similar types of research suggests that further research is necessary. I Some researchers studied the relationships between cognitive and social functioning i using 'global' cognitive measures. Although, many of these researchers found relationships | between cognitive and social functioning, the global cognitive measures lacked the capability of differentiating or identifying which of the elements of cognition were most impaired. Harvey et al. (1997) studied the relationship of cognitive functioning and various aspects of "social-adaptive functioning" in 208 elderly subjects diagnosed with schizophrenia. A l l subjects were examined with the Consortium to Establish a Registry for Alzheimer's Disease (CERAD) Cognitive Battery and the Social-Adaptive Functioning Evaluation (SAFE) scale. Both "instrumental and self care" and "social functions" subscales of the SAFE j accounted for significant variance increments in the cognitive composite score, while "impulse control" did not enter the regression analysis (Harvey et al., 1997). Harvey et al. (1997)! ! i suggested that instrumental and social skill deficits were more strongly correlated with cognitive impairments than the severity of uncontrolled behavior. Although Harvey et al. (1997) concluded that no single aspect of cognitive impairment was a better predictor of the leyel of Relationship between Attention and Social Functioning 19 social functioning, they admitted that the C E R A D was a "brief cognitive assessment that omitted some of the typical tests found in other cognitive test batteries. Because none of the CERAD tests were cited as measures of attention, potential impairments in attention may not have been assessed. Penny, Mueser, and North (1995) investigated the relationship between cognitive disability and social skills in 55 patients in an acute care adult inpatient psychiatric setting. Independent ratings of subjects' social skills were made during a semi-structured interview using the Social Interaction Test (SIT), followed by measurement of cognitive disability using the revised Allen Cognitive Level Test (ACT-90) (Penny et al., 1995). Penny et al. (1995) found a relationship between cognitive disability and social impairments. However, similar to the CERAD, the ACL-90 was a " global" measure of cognition, and appeared to lack the capability of differentiating which of the elements of cognition were most impaired. Green (1996) demonstrated that attention, memory, and executive functioning deficits were more strongly correlated than other aspects of cognitive functioning with poor responses to remediation of adaptive deficits in schizophrenia. Thus, global cognitive impairment was less important than specific cognitive impairments in the prediction of response to behavioral treatments. Identification of the specific cognitive deficit(s) related to impaired social functioning is a necessary prerequisite to attempts at cognitive remediation (Addington & Addington, 1999). Unlike Harvey et al. (1997) and Penny et al. (1995), the present study employed a measure designed to assess attention, therefore allowing some clarity as to which aspect of cognitive functioning will be measured. Lysaker et al. (1995) examined the relationship between cognitive deficits and social functioning in the work environment of 91 individuals with schizophrenia or schizoaffective disorder. Multiple regression analysis revealed, that after 10 weeks of work, the neuropsychological measures uniquely predicted improvement among individuals with impaired Relationship between Attention and Social Functioning 20 social functioning. Significant partial contributions were made by bizarre thought, impairments in concept formation, and concreteness of thought, with poorer performance on neuropsychological measures associated with the persistence of deficits in individuals' social functioning (Lysaker et al., 1995). Lysaker et al. (1995) proposed that "idiosyncratic processing of social information, combined with difficulties extracting and deducing the rules for appropriate social interaction from daily experience" (pg. 691), may be responsible for the social functioning impairments observed in individuals with schizophrenia in natural work settings. Because Lysaker et al. (1995) had not considered the influence of cognitive processes associated with early aspects of information processing, such as attention, their conclusions might not have been complete. For instance, both Lysaker et al. (1995) and Penn et al. (1995) used a card-sorting task to identify a correlation between concept formation and social functioning. However, because Penn et al. (1995) included a version of the Continuous Performance Test (CPT), they also found a correlation between attention tasks and social functioning. This finding allowed Penn et al. (1995) to developed a theoretical understanding of social functioning that incorporated early aspects of information processing. Summary The foregoing section examined the complex relationships between cognitive abilities and social functioning in individuals with schizophrenia. Growing evidence suggests a relationship between cognitive deficits and impaired social functioning in schizophrenia (Addington, et al., 1998; Ancil et al., 1994; Dickerson et al., 1996; Green, 1993; Green, 1996; Lysaker et al., 1995; Penn et al., 1995). Yet, relatively little is known about which cognitive deficits may impact social functioning (Addington & Addington, 1999; Penn et al., 1995). In fact, there is little consensus as to which aspect of cognition is related to social functioning, even among studies with similar methodologies. Relationship between Attention and Social Functioning 21 Many of the studies lacked guiding hypotheses to explain the relationships between certain cognitive measures and social functioning behaviors. The absence of guiding hypotheses meant that different researchers selected widely different measures, making it more difficult to draw general conclusions (Green, 1996). Many of the studies cited neuropsychological measures that assessed several aspects of cognition, such as: attention, memory, executive and intellectual functioning, ability to abstract, perceptual organization, fine motor speed, and dexterity (Addington & Addington, 1999; Addington et al., 1998; Cornblatt & Keilp, 1994; Corrigan et al., 1994; Dickerson et al., 1996; Freedman et al., 1998; Kern et al., 1992; Lysaker et al., 1995; McKee et al., 1997; Perm et al., 1993; Perm et al., 1995). The reviewed studies also assessed social functioning in a variety of different ways. Some of the studies (Addington et al., 1998; Harvey et al., 1997; Kern, et al., 1992; Massel et al., 1991; McKee et al., 1997; Penn et al., 1995; Penny et al., 1995) assessed social functioning by directly observing behaviors such as: poor eye contact, disruptive/non-disruptive behaviors, appropriate/non-appropriate responses, voice volume and pitch, tone of voice, speech rate and fluency, restlessness and attendance in group. Other studies (Addington & Addington, 1999; Cornblatt & Keilp, 1994; Corrigan et al., 1994; Dickerson et al., 1996; Freedman et al., 1998) used interviews or questionnaires to indirectly assess social functioning. Indirect methods assessed the following aspects of social functioning: sensitivity to interpersonal cues, social indifference, suspiciousness, social isolation, and ability to empathize. The lack of hypotheses meant that most of the analyses were conducted in a post hoc fashion. Post hoc analyses capitalized on chance, so some of the significant associations in this review were probably accidental (Green, 1996). Even though most studies found significant relationships between cognitive and social functioning measures, several of the correlation coefficients (Addington & Addington, 1999; Addington et al., 1998; Dickerson et a l , 1996; Lysaker et a l , 1995; McKee et al., 1997; Penn et al., 1993; Penn et al., 1995; Penny et al., 1995) Relationship between Attention and Social Functioning 22 would fall within the low to moderate range (0.26-0.49 and 0.50-0.69 respectively) (Domholdt, 1993). Green (1996) argued that many of the studies were "underpowered" and "overanalyzed" since they attempted to analyze the results of several different measures with a limited number of subjects. According to Domholdt (1993), significant correlation coefficients of a low magnitude may not describe clinically meaningful relationships. However, Domholdt (1993) also argued that low to moderate correlation coefficients may be considered strong when determining relationships between abstract constructs that are difficult to measure. It can be argued that terms such as attention and social functioning represent abstract constructs that are difficult to measure. Despite methodological limitations (see Appendix B for limitations of reviewed studies) and low to moderate correlation coefficients, some general conclusions are warranted. The literature reviewed indicated that although subjects performed poorly on some neuropsychological measures, they did not perform poorly on all measures. This finding suggests that social functioning impairments observed in individuals diagnosed with schizophrenia are likely related to specific cognitive deficits or a combination of several specific cognitive deficits, rather than to a global cognitive deficit. Of the various neuropsychological measures used, many of the studies found that attention measures were consistently correlated with social functioning measures (Addington et al. 1998; Cornblatt & Keilp, 1994; Freedman et al., 1998; Kern et al., 1992; McKee et al., 1997; Penn et al., 1993; Penn et al., 1995). The most frequently assessed aspects of attention were sustained and selective attention. This evidence supports the hypothesis that the capacity to sustain attention free from distraction may allow individuals with schizophrenia to accurately receive and process social information (Cornblatt & Keilp, 1994). However, because there is no consensus as to which aspect of cognition is related to social functioning, further research is necessary to effectively test this hypothesis. Relationship between Attention and Social Functioning 23 Models and Measures of Attention Turn-of-the-century researchers such as Wundt, Kraeplin, and Bleuler (as cited in Nuechterlein & Asarnow, 1989; Perry & Braff, 1994) were among the first to document cognitive deficits in schizophrenia. Kraeplin differentiated two types of attention impairments in schizophrenia. He claimed that disturbances in passive registration of information, or "auffassung" appeared only in the acute or terminal stages of the disorder, whereas disturbances in active, voluntary attention or "aufmerksamkeit'' were considered consistent features of schizophrenia (Nuechterlein & Asarnow, 1989). The study of attention declined in popularity with the advent of behaviorism at the beginning of World War I (Parasuraman & Davies, 1984). However, a lack of academic theory available to explain problems with servicemen maintaining vigilance during wartime operations, revived interest in the study of attention. During the late 1940's, the study of cognition adopted an "information processing" approach (Duchek, 1991; Parasuraman & Davies, 1984). The advent of the computer as an information processor provided a useful model for human cognition. According to Nuechterlein and Asarnow (1989): "Information processing models emphasize the structures and processes by which individuals register, encode, select, transform, store, and retrieve information" (p. 242). A crucial aspect of such models is the notion of limited processing capacity, that is, the amount of information that can be processed at any one time is limited (Kahneman & Treisman, 1984). Of the various structures and processes identified with information processing models, attention has been identified as the key factor in the information-processing model and cognitive functioning (Duchek, 1991; Kahneman & Treisman, 1984; Nuechterlein & Asarnow, 1989; Parasuraman & Davies, 1984; Tryssenaar & Goldberg, 1994). Attention can be conceptualized as processes that control the flow of information processing, focus mental effort, sustain this focus, and shift this focus back and forth when required (Duchek, 1991; Nuechterlein & Asarnow, 1989). Relationship between Attention and Social Functioning 24 The function of attention has been the topic of much debate by different theorists. Some theorists view attention as a mechanism for separating wanted from unwanted information. These theorists have attempted to identify the mechanism of attention and its location in the information processing sequence. Others view attention as a "processing resource" which can be variably allocated in relation to information processing demands. Still others focus on types of information processing and attempt to differentiate the kinds of processing that require varying degrees of attention involvement. As a result of much interest and research in the field of cognition, numerous models have been developed to conceptualize and describe attention (Broadbent, 1958; Deutsch & Deutsch, 1963; Ellis & Hunt, 1993; Norman, 1968; Treisman, 1964). An overview of the more significant models of attention identified in the literature may offer some understanding of the construct of attention. Early Selection Models. In the literature, selection models have also been referred to as filter, stage, structural, switch or set models. Selection models of attention are concerned with the series of information processing stages, in which the output of one stage is transferred to a subsequent stage for further processing (Green, 1993). Depending on the type of selection model, certain theoretical "filters" or "bottle necks" exist along the information processing stages that limit attention performance (Liberman & Green, 1992; Parasuraman & Davies, 1984). Broadbent (1958) hypothesized that attention operated early in the information sequence. According to this early selection model, attention is attracted and maintained by a sensory or physical attribute of the incoming information and selects the input message that will be further processed. Messages not selected remain in the sensory register and become completely blocked or filtered. An experimental design used to evaluate early selection models consisted of a combination of dichotic listening and shadowing tasks (Broadbent, 1958; Ellis & Hunt, 1993). Dichotic listening referred to the presentation of different auditory messages to each of the ears at the same time. In the dichotic Relationship between Attention and Social Functioning 25 listening task, the subject was asked to listen to one message presented to one ear and ignore the message from the other ear. In order to minimize shifting of attention between the "attended" and "unattended" ear, the subject was asked to shadow the message of the "attended" ear. Shadowing is a procedure in which the listener followed the message to the "attended" ear by repeating the message as it occurred. Researchers argued that a subject effectively shadowing one message could not switch attention to the other message. When the shadowing procedure was incomplete, it was assumed that attention had shifted and the subject might know something about the unattended message at that point (Broadbent, 1958; Ellis & Hunt, 1993; Sohlberg & Mateer 1989). Attenuator Models. Treisman (1964) agreed that attention filtered information prior to meaningful analysis. However, Treisman's (1964) "attenuator' model allowed for the processing of more than one input at a time and suggested that attention may be influenced by semantic as well as sensory information. Studies presenting compound sentences during the dichotic listening tasks challenged Broadbent's initial idea that unattended information was not processed at a higher level and therefore could not influence performance (Ellis & Hunt, 1993; Treisman, 1964). Compound sentencing involved the presentation of the first half of a message to the shadowed ear and the second half of the message to the nonshadowed ear. Simultaneously, another message was presented similarly, with the first half of the message directed to the nonshadowed ear and the second half of the message to the shadowed ear. Researchers found that as the meaning of the shadowed sentences switched to the nonshadowed ear, shadowing was disrupted. Subjects found it impossible to shadow the appropriate message consistently. This finding suggested that meaning of the second portion of the message in the nonshadowed ear was being processed enough to influence the subject's behavior. To account for this information, theorists proposed the attenuator model (Ellis & Hunt, 1993; Treisman, 1964). Attention still filtered Relationship between Attention and Social Functioning 26 information prior to meaningful analysis or pattern recognition. However, unlike the all-or-nothing proposition of the early selection model, the attenuator model allows for the processing of more than one input at a time. As with the early selection model, changes in the physical cues of input could serve as the basis for adjusting the attenuation of various inputs. Yet, unlike the early selection model, the meaning of previously analyzed material may also influence attention. Therefore, attenuator control is exerted by both sensory and semantic information (Kahneman & Treisman, 1984; Treisman, 1964). Late Selection Models. In contrast to the early selection model, the late selection model assumes that the attention filter occurs at the point of response selection, not at the sensory input level (Deutsch & Deutsch, 1963; Norman, 1968). In this model, attention operates after activation of memory representation such that "meaning" of the stimulus is preconscious. Thus, attention selects among the preconsciously activated meanings, allowing only a limited number of meanings to reach consciousness (Deutsch & Deutsch, 1963; Ellis & Hunt, 1993; Norman, 1968). The issue of preconscious selection is a critical difference between early- and late- selection models. Evidence to support the concept of preconscious selection was derived from studies of divided attention and the amount of information the subjects appeared to know about the unattended input (Ellis & Hunt, 1993; Norman, 1968; Posner & Rafel, 1987; Wickens, 1984). Late selection researchers argued that simply asking the subject to tell what they remembered after the dichotic listening and shadowing task used in early selection research inadequately measured preconscious selection. According to late selection models, unattended memory representations decay within seconds, thus during the delay between the presentation and testing of the information, the nonshadowed material is forgotten (Ellis & Hunt, 1993; Posner & Rafel, 1987). Relationship between Attention and Social Functioning 27 Capacity Models. The basic premise of capacity models is that psychological resources are limited and a certain amount of cognitive capacity is necessary to devote to various tasks. The number of activities that can be done simultaneously is determined by the capacity each requires (Ellis & Hunt, 1993; Green, 1993; Nuechterlein & Asarnow, 1989; Wickens, 1984). Automatic processing, or "automaticity", is another important concept of the capacity model. Automaticity refers to the apparent lack of central capacity requirements for a particular task (Duchek, 1991; Ellis & Hunt, 1993; Kahneman & Treisman, 1984; Nuechterlein & Asarnow, 1989; Schneider, Dumais, & Shiffrin, 1984; Sohlberg & Mateer, 1989; Wickens, 1984). The development of automaticity is important for normal functioning in everyday tasks. Although practice is a critical ingredient, the precise mechanisms underlying the development of automaticity are unknown (Ellis & Hunt, 1993; Nuechterlein & Asarnow, 1989; Wickens, 1984). Research on capacity models employed secondary task techniques to measure how resources are allocated to various inputs. The techniques involved having subjects perform a primary task, such as solving anagrams, while simultaneously performing a secondary task, such as pressing a button as rapidly as possible when a tone sounded. The response speed of the secondary task was taken as a measure of capacity required to perform the primary task. In accordance with the basic premise of the capacity model, it was hypothesized that the more difficult anagram would require more capacity to perform the anagram task, thus less capacity would be available for the tone detection task and the reaction to tone would be slower (Ellis & Hunt, 1993; Wickens, 1984). The outcome of these studies supported the notion that subjects would need to modulate the supply of resources between tasks in order to obtain the desired levels of performance (Ellis & Hunt, 1993; Wickens, 1984). Relationship between Attention and Social Functioning 28 Component Models. Component models of attention have developed from the need to address clinical issues involving individuals with cognitive impairments, and therefore have also been referred to as clinical models (Sohlberg & Mateer, 1989). These models are relatively new in the field and have been developed during the past 10-15 years. Although the literature is rich in normal attention capacity theory, it is not always obvious how capacity models might directly lead to treatment strategies (Posner & Rafal, 1987; Sohlberg & Mateer, 1989). Posner and Rafal (1987) and Sohlberg and Mateer (1989) argued that previous models stop at the level of signal detection and target selection, thus are i l l equipped to address the complex clinical phenomena related to attention deficits. Component models (Posner & Rafal, 1987; Sohlberg & Mateer, 1989) conceptualize attention as a multidimensional capacity critical to memory, new learning, and all other aspects of cognition. Although there is no generally agreed upon definition of attention, Posner and Rafal (1987) argued that three aspects of attention predominate in the literature. The three general aspects are alertness, selective attention and sustained concentration or vigilance. Alertness improves readiness to respond to selected information. Selective attention facilitates the processing of both sensory information, and information stored in memory (Posner & Rafal, 1987). The process of selective attention largely proceeds automatically and unconsciously; therefore a great deal of information about a signal is processed simultaneously. Selection of information for decision-making and response execution, associated with vigilance, requires conscious mental effort (Posner & Rafal, 1987). Sohlberg and Mateer (1989) described a hierarchical model for the theoretical understanding and assessment of attention. Each level of the hierarchy is viewed as more complex and as requiring effective functioning of the previous level. Table 1 offers a description Relationship between Attention and Social Functioning 29 of Sohlberg and Mateer's model of attention and measures proposed by Mateer and Mapou (1996) to assess the various levels of attention. The model of attention developed by Sohlberg Table 1 Sohlberg & Mateer's Model of Attention with Proposed Measures to Assess Levels of Attention Level Description Measures Focused attention Ability to respond discretely to specific visual, auditory, or tactile stimuli. Does not imply purposefulness of response. A l l following listed tests Sustained attention Ability to maintain consistent behavioral response during continuous and repetitive activity. Incorporates concept of vigilance and, at perhaps higher level, the concepts of mental control or working memory. Continuous Performance Test; Digit Span-Forward (WAIS-R); Trail Making Test, Part A; Test d2; Brief Test of Attention; Attentional Capacity Test (ACT) Selective attention Ability to maintain behavioral or cognitive set in the face of distracting or competing stimuli. Incorporates concept of "freedom from distractibility". Continuous Performance Test; Digit Span-Forward (WAIS-R); Trail Making Test, Part A ; Test d2; Brief Test of Attention; ACT Alternating attention Capacity for mental flexibility that allows individuals to shift their focus of attention and move between tasks having different cognitive requirements or requiring different behavioral responses, thus controlling which information will be selectively attended to. Incorporates concept of shifting an established "set" easily. Symbol Digit Modalities Test; Digit Symbol (WAIS-R); Paced Auditory Serial Addition Test (PASAT); Digit Span-Backward (WAIS-R); Arithmetic (WAIS-R); Consonant Tri grams Divided attention Ability to respond simultaneously to multiple tasks or multiple task demands. Two or more behavioral responses may be required, or two or more kinds of stimuli may need to be monitored. Symbol Digit Modalities Test; Digit Symbol (WAIS-R); PASAT; Digit Span-Backward (WAIS-R); Arithmetic (WAIS-R); Consonant Tri grams Relationship between Attention and Social Functioning 30 and Mateer (1989) was based on the experimental attention literature, clinical observation, and patients' subjective complaints. Benedict (1989) reviewed the literature investigating the efficacy of attention remediation interventions for individuals with brain injuries. Of the attention remediation interventions reviewed, Benedict (1989) noted that "the most promising intervention thus far" had been Sohlberg and Mateer's (1987) hierarchical approach to attention remediation. Sohlberg and Mateer's (1987) intervention was based on their hierarchical model of attention. The value of research and theory development in understanding a phenomena and guiding clinical reasoning has been documented in the field of rehabilitation (Krefting, 1985; Tammivaara & Shepard, 1990). Therefore, it is reasonable to assume that Sohlberg and Mateer's (1987) attention remediation protocol may be effective because it is based on a sound theoretical model of attention. Assessment of Attention Attention is a complex construct and how attention is conceptualized will determine the manner of how attention deficits are assessed and treated (Green, 1993). There exist a number of different clinical measures currently used to assess attention. According to the literature review, measures used to assess attention in individuals with schizophrenia included the: Attention Span Test (as cited in Freedman et al., 1998), Continuous Performance Test (Nuechterlein & Asarnow, 1987), Digit Span Distractibility Test (Wechsler, 1981), (Oltmanns & Neale, 1975), Stoop Color-Word Test (Golden, 1978), Symbol Digit Modalities Test (Wechsler, 1981), WAIS-R Arithmetic subtest (Wechsler, 1981), WAIS-R Digit Span Test (Wechsler, 1981), and WISC-HI Mazes (Wechsler, 1991). These measures are very different in their requirements, and are not theoretically based in any particular model of attention or even a component of such a model (Sohlberg & Mateer, 1989). Relationship between Attention and Social Functioning 31 Evaluation of attention is often equated with the "freedom from distractibility" subtests such as the Digit Span, Arithmetic, and Coding/Digit Symbol tests from the Wechsler intelligence scales, or the Speech-sounds Perception and Seashore Rhythm tests from the Halstead-Reitan Neuropsychological Battery (as cited in Mateer & Mapou, 1996). According to Mateer and Mapou (1996) and Sohlberg and Mateer (1989), the use of models to guide assessment of attention has been rare because: (a) clinicians have disagreed about the processes of attention measured by different tests; (b) most measures of attention were not developed within a particular conceptual model of attention; and (c) despite a long history of attention research among cognitive and experimental psychologists, there has been little crossover between the clinical and cognitive-experimental literature. Although researchers (Broadbent, 1958; Deutsch & Deutsch, 1963; Moray; 1967; Norman, 1968; Treisman, 1964) have proposed numerous theories to explain the mechanisms of attention, many of the theories are strongly tied to specific experimental tasks or paradigms, and do not identify ways in which they might directly lead to treatment strategies (Sohlberg & Mateer, 1989). Historically, the Continuous Performance Test (CPT) has been considered a test of "vigilance" or "sustained attention", and has been cited as a measure of attention functioning in numerous studies (Cornblatt & Keilp, 1994). The CPT is not a single measure but a family of measures that shares a number of features: (a) the rapid presentation of a long series of stimuli, (b) the requirement that a subject respond whenever a designated target or target sequence occurs in the series, and (c) a relatively low probability that the target will appear (Cornblatt & Keilp, 1994). Cornblatt and Keilp (1994) reviewed over 40 studies that used various versions of the CPT as the primary measure of attention. Although the ability to sustain attention over time is clearly essential to successful CPT performance, according to Cornblatt and Keilp (1994) the majority of studies that employed the CPT found that a decline in sustained attention over time was not the critical impairment in schizophrenia. Rather, deficient processing capacity, that is Relationship between Attention and Social Functioning 32 the amount of information that can be processed at any one time, was the overriding deficit in nearly all_cases (Cornblatt & Keilp, 1994). For instance, certain versions of the CPT, such as the degraded stimulus CPT (DS CPT) and the identical pairs CPT (CPT-IP), were designed to increase the information-processing load in order to detect subtle deficits in individuals who were at risk but not yet clinically affected by schizophrenia. Other measures have also been identified in the literature as appropriate means of assessing attention functioning. Sohlberg and Mateer (1989) defined cognitive rehabilitation as the therapeutic process of increasing or improving an individual's capacity to process and use information so as to allow increased functioning in every day life. They considered attention as "a multidimensional cognitive capacity critical to memory, new learning, and all other aspects of cognition" (pi20). Of the various measures available to diagnose attention impairments and assess treatment efficacy, Sohlberg and Mateer (1989) recommend the Paced Auditory Serial Addition Task (PASAT) developed by Gronwall (1977). The PASAT measures an individual's capacity or ability to process information by determining the time required to provide a correct response. Sohlberg and Mateer (1989) recommended the use of the PASAT because: (a) it is a neuropsychological test sensitive to attention deficits; and (b) it assesses divided attention. Divided attention requires the individual to process multiple stimuli. Reduced capacity for divided attention and complex information processing can result in significant impairments in daily life. In many common activities, individual are required to divide their attention among several subtasks such as driving while talking to a passenger, or listening to someone while performing a job. Assessment protocols must incorporate measures that adequately tap these higher levels of attention capacities (Sohlberg & Mateer, 1989). Measures to Assess Social Functioning A multitude of terms has been used to describe an individual's ability to express his or her thoughts and feelings. For the purpose of this review, the term social functioning will be Relationship between Attention and Social Functioning 33 considered synonymous with the various terms found in the literature such as: social skills; level of participation; social perception; social problem solving; social competence; and conversational skills. A large number of instruments and methods have been developed to assess aspects of social functioning in psychiatric populations (Munroe-Blum et al., 1996; Weissman, Sholomskas, & John, 1981). Instruments and methods often reflect different purposes including rehabilitation assessment, treatment planning, prediction of outcome in epidemiological surveys, discriminating groups or subgroups of patients, and evaluation of treatments (Munroe-Blum et al., 1996). Some of the methods used to assess social functioning include: (a) self-report questionnaires, (b) interviews, (c) naturalistic in vivo observations, (d) confederate tests, and (e) role play situations. Liberman (1982) reviewed the aforementioned methods used to assess social functioning and concluded that, while these methods have been used reliably and productively, they all have advantages and limitations. For example, many self-report questionnaires have been based on face validity and lack systematic analysis of their psychometric properties. Role-play tests of social functioning offer researchers a standardized, easily replicable assessment situation, yet the performance anxiety associated with the "testing" qualities of the role-play may interfere with the subject's full demonstration of social functioning (Liberman, 1982). Confederate tests and naturalistic in vivo observations can be costly and time consuming. Furthermore, the variation among naturalistic interactions in duration, content of task, stressors, and setting makes comparisons among studies difficult (Liberman, 1982). Certain studies that examined the relationship between cognition and social functioning in individuals with schizophrenia employed methods that required direct observation of behaviors. Behaviors often associated with social functioning included eye contact, ability to respond to questions, speech rate and fluency, pauses in speech, body posture, and restlessness (Kern et al., 1992; Massel et al., 1991; McKee et al., 1997; Penn et a l , 1995; and Penny et al., Relationship between Attention and Social Functioning 34 1995). Some researchers employed standardized tests of social functioning that identified and defined the behaviors to be monitored. For example, Penny et al. (1995) used the Social Interaction Test, Harvey et al. (1997) used the Social-adaptive Functioning Evaluation, and Lysaker et al. (1995) used the Work Personality Profile. Other researchers did not employ standardized tests. Rather, they selected behaviors that they deemed relevant to appropriate social functioning (Kern et al., 1992; Massel et al., 1991; McKee et al., 1997; Penn et al., 1995). Other studies that examined the relationship between cognitive and social functioning in individuals with schizophrenia employed methods that did not require direct observation of behaviors. Some of the measures used in these studies included: (a) Social Adjustment Scale II (Schooler, Hogarty, & Weissman, 1974); (b) Means Ends Problem Solving Test (Piatt & Spivack, 1974); (c) Social Interaction Scale (Tower, Bryant, & Argyle, 1978); (d) Personality Disorders Examination (Loranger, Susman, Oldham, & Russakoff, 1987); (e) Social Function Scale (Birchwood, Smith, Cochrane, Wetton, & Copestake, 1990); (f) Social Dysfunction Scale (Munroe-Blum et al., 1996). Comprehensive reviews of these instruments can be found in the literature (Linn, 1988; Wallace, 1986; Weissmann, 1975; Weissmann et al., 1981). The Social Dysfunction Scale (SDI; Munroe-Blum et al., 1996) was used to assess social functioning in this study. While no instrument is without limitations, the SDI meets the following criteria defined by Weissman et al. (1981) for an "appropriate" measure of social functioning: a) a review of the measure's item content, (b) objectively defined anchor points, (c) a clearly described method of obtaining information, (d) a review of psychometric properties, (e) an examination of the instrument's precision, (f) the cost of the instrument should be low, (g) clearly defined scoring procedures, and (h) instrument should reflect changes in traditional roles, especially among women. Relationship between Attention and Social Functioning 35 Conclusion Several of the studies reviewed suggested that attention is positively correlated with social functioning in individuals with schizophrenia or schizoaffective disorder. However, other studies reviewed suggested that alternative aspects of cognition are correlated with social functioning. Because there is no consensus as to which aspect of cognition is correlated with social functioning in individuals with schizophrenia or schizoaffective disorder, continued research in this area is warranted. Many of the studies reviewed lacked guiding hypotheses, employed a variety of cognitive measures, and conducted post hoc analyses of the data. As a consequence, some of the significant associations in this review may have been accidental. This study examined the relationship between only one aspect of cognition, namely attention, and social functioning in individuals with schizophrenia or schizoaffective disorder. The null and alternative hypotheses were identified in Chapter 1. Analysis of the data was a priori. The following chapter will describe the methodology of the study. Relationship between Attention and Social Functioning 36 CHAPTER 3 Method According to some researchers (Addington & Addington, 1999; Addington et a l , 1998; Dickerson et al., 1996; Green, 1996) the present study could be described as a "rate-limiting" study. Addington et al. (1998) described a rate-limiting study as "one that has examined associations between cognitive functioning and outcome measures that are believed to be relevant to functioning in the real world" (p. 64). In this study the rate-limiting cognitive factor was attention, as measured by PASAT. The outcome measure believed to be relevant to functioning in the real world was social functioning, as measured by the SDI. Rate-limiting studies can identify "core" cognitive deficits associated with social functioning, and are therefore necessary prerequisites to "generalization studies" (Addington & Addington, 1999; Green, 1996). According to Addington et al. (1998) generalization studies are designed to "test the effects of remediation of associated cognitive deficits on social functioning" (p. 64). Subjects Subjects were selected from the participants of the Schizophrenia Rehabilitation Day Program (SRDP) at U B C Hospital in Vancouver, B. C. The SRDP is a hospital-based, psychosocial rehabilitation program that provides a comprehensive outpatient service for individuals with schizophrenia or related disorders. SRDP participants range in age from 17 to 55 years of age, with the majority aged 22 to 33 years. The rationale for selecting subjects from the SRDP was that potentially extraneous factors resulting from psychiatric treatments could be documented. Individuals fulfilling all the following criteria were eligible to enter the study: 1. Male or female participants of the SRDP, from 17 to 55 years of age. Relationship between Attention and Social Functioning 37 2. Subjects had the diagnosis of schizophrenia or schizoaffective disorder as confirmed by the SRDP psychiatrist who conducted a semi-structured interview based on DSM-IV (American Psychiatric Association, 1994) diagnostic criteria. 3. Subjects were outpatients. 4. Subjects were English speaking and must have completed a Grade IX level of education. 5. Subjects were mentally stable, as indicated by a review of his or her medical record. 6. Each subject gave written informed consent before his or her participation. Individuals fulfilling any one of the following criteria were excluded from the study: 1. Serious physical illness that would interfere with participation in the study or interfere with cognitive functioning. 2. DSM-IV classified substance (alcohol or other drug) abuse or dependence within the 3 months prior to the initiation of the study. 3. Evidence of brain injury or mental retardation as evaluated with DSM-IV criteria. Individuals meeting inclusion criteria 1 to 5 were given a letter of initial contact (as approved by the U B C and Vancouver Hospital Ethics Committees) by a SRDP staff member other than the primary investigator, thus notifying individuals of the study. A l l individuals met inclusion criterion 6 prior to admission to the study. This study included individuals with a diagnosis of schizophrenia or schizoaffective disorder because: (a) the DSM-IV (American Psychiatric Association, 1994) identified poor occupational functioning and restricted range of social contact as features in both schizophrenia and schizoaffective disorder; (b) previous studies have indicated that schizophrenia and schizoaffective disorder were similar in their neuropsychological deficits (Dickerson et al., 1996); (c) several studies that examined the relationships between cognitive and social functioning included individuals with the diagnosis of schizophrenia and schizoaffective disorder (Addington et al., 1998; Kern et al., 1992; Lysaker et al., 1995; McKee et a l , 1997; Penn et al., Relationship between Attention and Social Functioning 38 1995; and (d) from a clinical perspective, a mix of individuals with the diagnosis of schizophrenia or schizoaffective disorder would be more representative of the client population treated at an outpatient service for schizophrenia and related disorders. Relationships among subjects' demographic information and SDI and PASAT scores To determine whether subjects' demographic information may have confounded the results of the study, comparisons of subjects' demographic information with SDI overall percent scores and PASAT mean time/responses scores were conducted. An alpha level of 0.05 was used for all statistical tests. Descriptive statistics for subjects' age, level of education and duration of illness are displayed in Table 2. Pearson correlation analyses were conducted to determine the relationships between the subjects' age, level of education and duration of illness on SDI overall percent scores and PASAT mean time/response scores. Pearson correlations were used to analyze relationships between these demographic variables and SDI and PASAT scores because Pearson correlations are more appropriate than t-tests when using continuous data (Jonathan Berkowitz, personal communication, February 13, 2001; Jonathan Money, personal communication, November 3, 2000). See Appendix C for SPSS printout of these Pearson correlation analyses. Pearson correlation analyses found no statistically significant relationships between subjects' ages and SDI overall percent scores (r = - 0.068, p = 0.698) or PASAT mean time/response scores (r = - 0.144, p = 0.408). Figure 3 displays scatter plots of the relationships between subjects' ages and SDI overall percent scores and PASAT mean time/response scores. The slope of the regression lines for each of the two scatter plots were not significantly different from zero, suggesting no relationship between subjects' ages and SDI overall percent scores and PASAT mean time/response scores (Jonathan Berkowitz, personal communication, February 13, 2001). Relationship between Attention and Social Functioning 39 Table 2 Descriptive Statistics of Subjects' Demographic Information Range of Scores Minimum Score Maximum Score Mean Standard Deviation Age (in years) 32 17 49 28.40 8.31 Years of Education 13 9 22 13.37 2.68 Months Diagnosed 339 1 340 66.54 9.52 Similarly, there was no statistically significant relationship between subjects' level of education and PASAT mean time/response scores (r = - 0.149, p = 0.394). However, there was a statistically significant relationship between the subject's level of education and SDI overall percent scores (r = - 0.370, p = 0.029). Figure 4 displays scatter plots of the relationships between subjects' level of education and SDI overall percent scores and P A S A T mean time/response scores. The slope of the regression line for the scatter plot displaying the relationship between subjects' level of education and PASAT mean time/response scores was not significantly different from zero, thus suggesting no relationship. However, the slope of the regression line for the scatter plot displaying the relationship between subjects' level of education and SDI overall percent scores was significantly different from zero, suggesting social dysfunction decreases with more years of education. (Jonathan Berkowitz, personal communication, February 13, 2001). Unfortunately, the scattering of data points around the regression line suggests a relatively large variability among subjects, limiting the ability to predict a direct one-to-one correlation between the level of education and SDI overall percent scores. Relationship between Attention and Social Functioning 40 Finally, there were no statistically significant relationships between the number of months subjects had been diagnosed with schizophrenia or schizoaffective disorder and SDI overall percent scores (r = - 0.057, p = 0.745) or PASAT mean time/response scores (r = - 0.022, p = 0.901). Figure 5 displays scatter plots of the relationships between the number of months diagnosed and SDI overall percent scores and PASAT mean time/response scores. The slope of the regression lines for each of the two scatter plots were not significantly different from zero, suggesting no relationship between the number of months diagnosed and SDI overall percent scores and PASAT mean time/response scores (Jonathan Berkowitz, personal communication, February 13, 2001). These results support the findings of previous research. Corrigan et al. (1993) also found that no statistically significant correlations between age, education, or chronicity on the Social Cue Recognition Test (SCRT). Similarly, Penn et al. (1995) found that the relationship between cognitive measures and social competence remained after controlling for chronicity. Table 3 summarizes Pearson correlation analyses between subjects' age, level of education, and duration of illness and SDI overall percent scores and PASAT mean time/response scores. Frequency counts for subjects' diagnosis, gender and medication dosage level are displayed in Table 4. Table 3 Pearson Correlation Analyses between Subjects' Demographic Information and SDI Overall Percent Scores and PASAT Mean Time/Response Scores Subjects' Demographic Information Age Level of Education Duration of Illness SDI overall percent scores r = -0.068,p = 0.698 r = -0.370,p = 0.029 r = -0.057,p = 0.745 PASAT mean time/response scores r = - 0.144, p = 0.408 r = - 0.149, p = 0.394 r = - 0.022, p = 0.901 Relationship between Attention and Social Functioning Table 4 41 Frequency Count of Subjects' Demographic Information Frequency Count Percentage Diagnosis Schizophrenia 26 74.3 Schizoaffective Disorder 9 25.7 Gender Male 25 71.4 Female 10 28.6 Medication Dosage Level Lower Level 27 77.1 Higher Level 8 22.9 Independent t-tests were conducted to determine the effect of the subject's gender, diagnosis, and level of medication on SDI overall percent scores and PASAT mean time/response scores. See Appendix D for SPSS printout of independent t-test analysis. A Levene's test of equality of variances was conducted on all independent t-test analyses. The Levene's test of equality of variances was used to test i f the spread of the groups differed (SPSS base 10.0, 1999). The null hypothesis of the test assumes that the two population variances are equal. According to the SPSS printout, Levene's test of equality of variances (SPSS base 10.0, 1999) suggested that equal variances should be assumed for all independent t-test analyses except for the independent t-test between diagnosis and SDI overall percent scores. The F-statistic for the independent t-test between diagnosis and SDI overall percent scores was 4.201 with a p-value of 0.048. Since the observed significance level for the F-statistic was low (less than 0.05), the hypothesis of equal variances was rejected and a pooled variance test not used. This suggested that a separate-variance t-test for means must be used (SPSS base 10.0, 1999) for the independent t-test between diagnosis and SDI overall percent scores. Therefore, results of Relationship between Attention and Social Functioning 42 the independent t-test between diagnosis and SDI overall percent scores were based on the assumption of non-equal variances as displayed in the SPSS printout. Analysis with an independent t-test indicated there was no significant effect of subjects' gender on SDI overall percent scores, t (33) = 1.504, p = 0.142. Similarly, there was no significant effect of subjects' gender on PASAT mean time/response scores, t (33) = -0.521, p = 0.606. Therefore, male and female scores on the SDI or PASAT measures do not differ. Figure 6 boxplots display data distributions of SDI overall percent scores and PASAT mean time/response scores by subjects' gender. The medians and interquartile ranges for the boxplots displaying SDI overall percent scores by subjects' gender are similar; as are the medians and interquartile ranges for the boxplots displaying PASAT mean time/response scores by subjects' gender. These results support the findings of previous research. Addington & Addington (1999) also found that independent t-tests indicated no differences between gender on any of the cognitive and social functioning measures. The staff psychiatrist diagnosed subjects as having either schizophrenia or schizoaffective disorder. Analysis with an independent t-test indicated there was no significant effect of diagnosis on SDI overall percent scores, t (29) = -1.434, p = 0.162. Similarly, there was no significant effect of diagnosis on PASAT mean time/response scores, t (33) = 0.481, p = 0.634. Figure 7 boxplots display data distributions of SDI overall percent scores and PASAT mean time/response scores by subjects' diagnoses. Although the distribution spreads of the boxplots displaying SDI overall percent scores by subjects' diagnoses are visually different, the variability of the values is not statistically significant. The distribution spreads of the boxplots displaying PASAT mean/response scores by subjects' diagnoses are similar. However, the distribution spread for individuals' diagnosed with schizoaffective disorder is positively skewed. To determine whether SDI and PASAT scores were associated with medication dosage, the staff psychiatrist identified subjects' relative dosages as high or low according to the Relationship between Attention and Social Functioning 43 recommended optimal dose ranges referenced in the Compendium of Pharmaceuticals and Specialties, 34 t h edition, 1999 (David Irwin, personal communication, September 28,1999). Independent t-test analysis revealed no significant effect of medication dosage level on SDI overall percent scores, t (33) = -0.852, p = 0.400. Similarly, independent t-test analysis found no significant effect of medication dosage level on PASAT Mean Time/Response scores, t (33) = 0.876, p = 0.387. Figure 8 boxplots display data distributions of SDI overall percent scores and PASAT mean time/response scores by subjects' medication dosage level. The medians and interquartile ranges for the boxplots of SDI overall percent scores by subjects' medication dosage level are similar. However, the distribution spread for the lower medication dosage level is positively skewed. Although the medians are similar for the boxplots of PASAT mean time/response scores by subjects' medication dosage level, the distribution spread for the higher medication dosage level is greater than distribution spread for the lower medication dosage level data. Table 5 summarizes t-test analyses between subjects' gender, diagnosis, and level of medication and SDI overall percent scores and PASAT mean time/response scores. These results also support the findings of other researchers. Addington and Addington's (1999) correlation analyses revealed that medication dose in chlorpromazine (CPZ) equivalents was unrelated to performance on any of the cognitive tests used. Similarly, Corrigan et al. (1994) found no significant correlations with neuroleptic medication levels and scores on the SCRT. Relationship between Attention and Social Functioning 44 Table 5 Independent t -test Analysis between Subjects' Demographic Information and SDI overall percent scores and PASAT mean time/response scores Subjects' Demographic Information Gender Diagnosis Level of Medication SDI overall percent scores t (33) = 1.504, p-0.142 t (29) = -1.434, p = 0.162 t (33) = -0.852, p = 0.400 PASAT mean time/response scores t (33) = -0.521, p = 0.606 t (33) = 0.481, p = 0.634 t (33) = 0.876, p = 0.387 Variables The two variables examined were subjects': (a) attention, & (b) social functioning. Attention. Attention was determined by the PASAT. The PASAT is a measure designed to identify individuals with attention impairments and assess the degree of progress during recovery (Gronwall, 1977). Sohlberg and Mateer (1989), Sohlberg (personal communication, November 25, 1998) and Mateer (personal communication, October 22, 1998) recommended that the PASAT be used to identify individuals with attention impairments. Although the PASAT was originally used for managing the rehabilitation of individuals with traumatic brain injuries, Sohlberg (personal communication, November 25, 1998) and Mateer (personal communication, October 22, 1998) argued that the PASAT would be an appropriate measure for individuals with schizophrenia since many of the cognitive deficits associated with schizophrenia are similar to the cognitive deficits associated with brain injuries. Tryssenaar and Goldberg (1994) included the PASAT in a single subject study that examined the efficacy of Sohlberg and Mateer's (1989) hierarchical approach to attention remediation in an individual with schizophrenia. Sohlberg and Mateer (1987) recommended the PASAT as an evaluation tool of attention impairments because it: (a) demonstrated sensitivity to attention deficits (from a sample of 320 individuals with head injuries, 99% scored one standard Relationship between Attention and Social Functioning 45 deviation below the mean of normal controls (Gronwall, 1977)), (b) has a strong normative base (n = 80), (c) has an inherent test-retest reliability (a practice effect was observed between first and second administration of the test, but further administrations produced negligible practice effects (Gronwall, 1977)), (d) provides an estimate of the subject's ability to register sensory input, rapidly process the information, and respond verbally, and (e) is compatible with Sohlberg & Mateer's (1989) theoretical understanding of attention. For example, the PASAT presupposes the existence of the first of the five levels of Sohlberg and Mateer's (1989) model of attention, namely focused attention, and depends heavily on the adequacy of sustained and selective attention. Many attention measures are contaminated by requirements for sometimes complicated verbal, mathematical, or numeric requirements in addition to the attention requirements. Weber (1986) criticized the PASAT on the basis that it required mathematical skills and correlated significantly with adding ability. However, Gronwall and Sampson (1974) found that serial addition correlated only 0.24 with arithmetic ability. Gronwall and Sampson (1974) argued that the low correlation was probably due to the very elementary level of arithmetic skill required for the PASAT, which is adding two single-digit numbers. The PASAT required the subject to listen to a sequence of numbers and add each number to the previous one in a continuous fashion. For example, i f the subject were given the sequence 5, 3, 7, 4, 2, 9, the subject would add the 5 and 3 together and say 8. Then the subject would add the 3 and 7 together and say 10. Next the subject would add the 7 and 4 together and say 11. The task was explained to the subject using a written example in order to clarify the concept of serial addition. The subjects were first given an 'unpaced' practice with the auditory form of the test, and then a 'paced' practice with stimuli presented at the rate of one number every 2.4 seconds. The numbers ranged from 1 through 9 and the duration of each digit was approximately 0.4 seconds (Gronwall, 1977). When necessary, further explanation, written Relationship between Attention and Social Functioning 46 demonstration, and unpaced practice were provided to ensure that the subject fully understood the task. Each of the four test trials involved the same 60 addition sums in the same order but at increasing speeds. The first trial presented numbers at the rate of one per 2.4 seconds, the second at 2.0 seconds, the third at 1.6 seconds, and the fourth at 1.2 seconds. The subject needed to respond before the next number was presented in order to be scored correct. At least 30-second intervals were allowed between trials. Total testing took between 15 and 20 minutes. The score was the total correct at each pacing speed. Average time per correct response was also computed to allow direct comparison of tests where different numbers of trials were given. This was done by dividing the total trial time by the number correct for each of the four pacing rates and averaging the results to give a composite score. Social functioning. An interview format to assess social functioning was selected over other formats, such as role plays, because it offered several advantages: (a) interviews are common experiences of psychiatric clients and therefore allow more realistic responses than may occur during a role-play (Penny et al., 1995); (b) an interview avoids problems identified with role-play tests such as the subject's understanding of the instructions, the situational content of the role-play, and the confederate's behavior influencing the role-play (Penny et al., 1995); (c) interviews can assess a subject's satisfaction of his/her social functioning, thus providing direction for therapeutic intervention; and (d) interviews can yield considerable information with an economy of clinical time. Subjects' social functioning was assessed using the Social Dysfunction Index (SDI). The SDI was developed to measure social dysfunction in individuals with schizophrenia and related disorders (Munroe-Blum et al., 1996). Unlike some older measures, the SDI reflects changes in traditional roles, especially with respect to gender (Munroe-Blum et al., 1996). The SDI can be used for: (a) clinical assessment and planning; (b) research investigating social functioning; and Relationship between Attention and Social Functioning 47 (c) as a measure of therapeutic response in treatment evaluation studies. The SDI is a semi-structured interview assessing the past month of functioning (Munroe-Blum et al., 1996). There are nine components of dysfunction assessed: (a) public self (behavior, appearance, social presentation); (b) independent living; (c) occupational functioning; (d) family relationships; (e) important relationships other than parental family; (f) community, leisure and recreational activities; (g) health maintenance activities; (h) communication (textual and contextual communication); (i) insight and expectations. Three items in each component make up the 27 items of the SDI. Table 6 presents the scoring format applied to each of the nine components. The SDI has certain advantages for the clinical setting in that it: (a) requires little professional time, (b) is acceptable to clients and their families, (c) is robust and has well-documented psychometric properties making it an ideal means to evaluate the effects of psychiatric and psychosocial treatment on social functioning, and (d) includes satisfaction items that allow for client-centered planning (Munroe-Blum et al., 1996). Probes and anchor points were included in the interview schedule. Appendix E offers an example of the component areas, the related example of probes and examples of serious social dysfunction for that component area. The SDI produced an overall social dysfunction score, a satisfaction score and a score for each of the nine separate components of dysfunction (Munroe-Blum et al., 1996). Suggested administration time was about 40 minutes (Munroe-Blum et al., 1996), however actual administration time averaged 50 to 60 minutes. Raters from a range of disciplines were quickly and effectively trained to administer the index as reported by Munroe-Blum et al. (1996). Relationship between Attention and Social Functioning 48 Table 6 Social Dysfunction Index Component Scoring (Munroe-Blum et al., 1996) 1. Dysfunction during past month 0 No major dysfunction 1 Minor dysfunction occasionally, but often manages adequately 2 Major dysfunction most of the time 3 Adequate functioning not evident and/or entirely managed by others 2. Number of areas of dysfunction 0 No major dysfunction 1 One or two areas of occasional to constant dysfunction 2 Three or four areas of occasional to constant dysfunction 3 Five or more areas of occasional to constant dysfunction 9 Not applicable/not known 3. Satisfaction with level of functioning 0 Satisfied 1 Not satisfied 9 Not applicable/not known The psychometric properties of reliability, validity, sensitivity to change and clinical utility of the SDI were tested in three separate studies of clients with schizophrenia or schizoaffective disorder (Munroe-Blum et al., 1996). Internal consistency was considered good for a scale of this length (alpha = 0.80). Inter-rater reliability was high (r = 0.96). Use of component scores was confirmed by factor analysis. Nine factors accounting for 72.2 % of the variance were extracted. Evidence of construct validity was demonstrated by findings of: (a) Relationship between Attention and Social Functioning 49 differences in social dysfunction between subjects who lived independently versus those who lived in boarding homes (30.10, SD 13.45 vs. 19.10, SD 12.57, t=3.42, df 58, p = 0.001); (b) differences between subjects who were unemployed versus subjects who were employed (F = 10.27, df 3,63, p< 0.01); and (c) independence of social functioning from both age and education (Munroe-Blum et al., 1996). Test Administration The primary investigator (PI), one research assistant, two nursing staff members, and one social work staff member of the SRDP were involved in administration and rating of the SDI. Except for the PI, all raters were blind to the study's hypothesis. The PI trained the other four raters to administer the instrument. Training of the raters involved: (a) an in-service about the purpose, components and psychometric properties of the SDI, and (b) practice scoring of the SDI while watching videotapes of a separate training sample of three SRDP clients whose consent to be interviewed and videotaped was previously obtained. Each client of the training sample was videotaped while the PI administered the SDI to him or her. After watching each of the training videotapes, the raters independently rated and then scored the training clients. The results of the separate training sample were not included in the study. Inter-rater reliability. During the training, inter-rater reliability was determined by using a weighted measure of percent agreement (Peter Liddle, personal communication, December 13,1999). For items that were rated on a scale of 0 - 3, the following weighted agreement scores were assigned: 1 when there was perfect agreement, 0.66 for a difference of 1 point, 0.33 for a difference of 2 points and 0 for a difference of 3 points. For items that were rated on a scale of 0 - 1, a 1 was assigned when there was perfect agreement and a 0 when there was disagreement. Percent agreement was calculated by dividing the total number of full and partial points obtained for the items by the Relationship between Attention and Social Functioning 50 total number of items scored. If the percentage agreement was not at least 90%, the PI and raters examined possible discrepancies in scoring and adjusted the instruction and training accordingly. To control for rater bias and drift (Kazdin, 1982) during the study, all subjects were jointly interviewed by the PI and one other rater. Each rater independently rated and then scored the SDI for each subject. After each subject was scored, and the scores recorded, the two raters examined inter-rater reliability using the weighted measure of percent agreement described earlier. Inter-rater rater reliability for all of the subjects was 80% or more using the weighted measure of percent agreement. Had inter-rater reliability been less than 80% for a particular subject, then that subject's SDI and PASAT scores would have been excluded from the study. Furthermore, the raters involved in the scoring of the excluded subject would have reviewed the training tapes until 90% agreement was reestablished. In situations where inter-rater reliability was 80% or better but the PI and the other rater did not agree on a particular score, then the average of the two scores was used to calculate the overall score the SDI. Inter-rater reliability of the SDI was confirmed at the end of the study by computing an intraclass correlation coefficient (ICC) and a Cohen's kappa using the five raters and 35 subjects. Unlike merely calculating the percentage of times two raters agree on their rating, ICC takes differences due to raters into account and controls the bias associated with the high levels of chance agreement whenever one or two categories predominate (Howell, 1997). Variability in the data may be due to differences between subjects, differences between raters, interaction between raters and subjects, and/or error (Howell, 1997). ICC takes the variance estimate for the effect in question, namely differences among subjects, relative to the sum of the estimates of several sources of variance. ICC is a method of measuring reliability within the context of inter-rater agreement that involves constructing ANOVA-type models for the observations in which the raters' scores are responses (SPSS, 1999). Therefore, ICC is basically subjects times raters A N O V A , with raters as a repeated measure. Relationship between Attention and Social Functioning 51 Another important statistic that measures inter-rater reliability, but uses contingency tables, is Cohen's kappa (Howell, 1997). Cohen's kappa is also a chance-corrected measure and considers the degree of disagreement by "weighting" larger differences between raters more severely than smaller differences between raters (Peter Liddle, personal communication, December 13,1999). Cohen's kappa involves calculating expected and observed frequencies. Examination of the formula for kappa reveals that in the numerator the number of agreements expected merely by chance is subtracted from the number of actual agreements. In the denominator the number of agreements expected merely by chance is subtracted from the total number of judgments. Therefore Cohen's kappa correlation is a ratio of the two chance-corrected values (Howell, 1997). Table 7 displays a summary of the ICC and Cohen's Kappa Correlation scores for the SDI raters. The SDI overall average ICC was 0.96, indicating very good inter-rater reliability (Howell, 1997). The SDI overall average Cohen's Kappa Correlation was 0.88, also indicating an excellent inter-rater reliability (Howell, 1997). Refer to Appendix F for a detailed description of the SDI ICC reliability analysis and Appendix G for a detailed description of the SDI Cohen's Kappa Correlation reliability analysis. An ICC or Cohen's kappa to determine reliability for each item of the SDI was not calculated before proceeding with the study because ICC and Cohen's kappa calculations would require substantial numbers of subjects. Furthermore, Munroe-Blum et al. (1996) tested the inter-rater reliability of the SDI by computing an ICC using 4 raters and 13 subjects that were diagnosed with schizophrenia or schizoaffective disorder. Munroe-Blum et al. (1996) found the SDI ICC to be 0.96, indicating good inter-rater reliability. Relationship between Attention and Social Functioning Table 7 52 Summary of ICC and Cohens Kappa Scores for the SDI raters Dysfunction during Past Month Number of Areas of Dysfunction Satisfac Fund tion with ioning ICC Cohens Kappa ICC Cohens Kappa ICC Cohens Kappa Public Self 0.980 0.948 0.979 0.913 1.000 1.000 Independent Living 0.963 0.864 0.924 0.731 1.000 1.000 Occupational Functioning 0.997 0.741 0.953 0.888 0.971 0.941 Family Relationships 0.971 0.859 0.938 0.841 1.000 1.000 Important Relationships 0.990 0.913 0.953 0.827 0.964 0.928 Community/Leisure Recreation 0.976 0.913 0.941 0.807 0.942 0.886 Acceptance & Adherence to Health Regiments 0.977 0.877 0.976 0.942 0.969 0.938 Communication 0.869 0.602 0.847 0.584 1.000 1.000 Insight/Accuracy of Information/ Expectations 0.993 0.960 0.928 0.799 1.000 1.000 Average Scores 0.968 0.853 0.938 0.815 0.983 0.966 To ensure reliable scoring practices of the PASAT throughout the study, the PI and another SRDP occupational therapist (OT) jointly administered the PASAT to the same training sample of SRDP clients that completed the SDI. Inter-rater reliability of the training sample was calculated by dividing the number of occasions on which the raters agreed on the rating of a Relationship between Attention and Social Functioning 53 particular item by the total number of items scored. This value was then multiplied by 100 to yield percentage agreement. During the training, i f the percentage agreement was not at least 90%, the PI and OT examined possible discrepancies in scoring and adjusted the instruction and training accordingly. To control for rater bias (Kazdin, 1982) during the study, PASAT testing for each subject was audiotape-recorded. The PI administered the PASAT to all subjects, but did not score the PASAT. The OT independently scored all of the recorded PASAT tests. A l l audiotape-recorded PASAT tests were stored in a locked cabinet at the end of the study. Design and Procedure The study employed a cross-sectional design, examining the relationship between attention and social functioning in a sample of 35 subjects with the diagnosis of schizophrenia or schizoaffective disorder. Dr. Marshall Arlin (personal communication, May 20, 1999) recommended a minimum sample size of 32 (30 degrees of freedom) in order to provide sufficient power for a Pearson product moment correlation between the overall PASAT and the overall SDI scores. It took approximately 12 months to recruit 35 subjects. An SRDP staff member other than the PI contacted individuals meeting all of the study's inclusion criteria and none of the exclusion criteria. When initially contacted about participating in the study, potential subjects were told that the study was about concentration and socialization and then given a letter of initial contact. Potential subjects were informed that they would be required to listen to a series of numbers and asked to do simple addition exercises such as 3 + 5 or 7 + 9. It was stressed that although the tasks themselves would be fairly simple, the real challenge would be to stay focused on the exercises. They would also be asked some questions about their thoughts regarding socializing and interacting with others. Finally, potential subjects were informed of the time commitment, their right to withdraw from the study at any stage, the Relationship between Attention and Social Functioning 54 need for a signed informed consent as required by the U B C Ethics committee, and the monetary honorarium for their participation. As subjects agreed to participate in the study, the SRDP staff psychiatrist, i f he had not already done so, conducted a semi-structured interview based on the DSM-IV (American Psychiatric Association, 1994) diagnostic criteria to ensure that subjects met the criteria for schizophrenia or schizoaffective disorder. After each subject's diagnosis was confirmed and informed consent obtained, a mutually convenient assessment period was negotiated between the subject and the primary investigator (PI). The SDI and the PASAT were administered to each subject on the same day during one assessment period. To control for possible order effects, administration of the SDI and PASAT were randomly determined for each subject by the PI prior to the subject's assessment period. The assessment period averaged approximately 85 to 100 minutes. Subjects were given a minimum 15-minute rest period between the SDI and PASAT. The SDI and PASAT were administered to each subject within a maximum time frame of five hours. The SDI and PASAT testing environments were consistent throughout the study for all subjects. In order to maintain a client-centered approach, the raters responded to individual needs with respect and sensitivity within the limits of a standardized procedure. Any comments made by the subjects regarding the tasks or their own thoughts about attention and social functioning were responded to as of interest to the rater. If the subject expressed concern about doing poorly, individually adjusted reassurance was given. After the assessment period was completed, each subject was thanked for his or her time and offered a $15.00 honorarium for participating in the study. Data Analysis A Pearson product moment correlation (r) was used to quantify the relationship between the SDI overall percent scores and PASAT mean time/response scores for the sample of 35 subjects. Interpretation of the coefficient was through significance testing at the 5% level, Relationship between Attention and Social Functioning calculation of the coefficient of determination, calculation of power, construction of a 95% confidence interval around the correlation coefficient, and calculation of sample size. Relationship between Attention and Social Functioning 56 CHAPTER 4 Results This chapter presents findings of several analyses: (a) Pearson correlation between SDI overall percent scores and PASAT mean time/response scores; (b) calculation of the coefficient of determination; (c) calculation of power; (d) construction of a 95% confidence interval around the correlation coefficient; (e) calculation of sample size and (f) Pearson correlation between the occupational functioning subcomponent of the SDI and PASAT mean time/response scores. The SDI overall percent scores and PASAT mean time/response scores were calculated for each subject and used in the following data analysis. The SDI overall percent score was calculated as a percentage of total possible scores without penalty for missing items (Munroe-Blum et al., 1996). The three subcomponents in each of the nine components of dysfunction made up the 27 items of the SDI. The 27 items were summed and divided by the total possible from the applicable items (total possible score was 63). The resulting value was then multiplied by one hundred to obtain a percentage value. The SDI overall percent scores ranged from 4.75 to 57.14, with a mean score of 34.78 and a standard deviation of 11.72. Prior to calculating the PASAT mean time/response scores, the total trial time for each pacing rate was determined. The total trial time was determined by multiplying the pacing speed for a particular trial (2.4, 2.0, 1.6, or 1.2 seconds) by sixty, which represented the total number of single digits per trial. The total trial time for a particular pacing rate was then divided by the number of correct responses for each of the corresponding rates. The results of the four pacing rates were then averaged to determine the PASAT mean time/response score (Gronwall, 1977). The PASAT mean time/response scores ranged from 2.46 to 22.43, with a mean score of 5.88 and a standard deviation of 4.38. Relationship between Attention and Social Functioning 57 Pearson correlation between SDI overall percent scores and P A S A T mean time/responses scores The purpose of this research was to determine whether a relationship existed between attention and social functioning in individuals with schizophrenia or schizoaffective disorder. A Pearson correlation analysis found a small, statistically insignificant relationship between the SDI overall percent scores and PASAT mean time/response scores (r = 0.086, p = 0.624) of the sample population. Figure 9 is a scatter plot displaying the relationship between the SDI overall percent scores and PASAT mean time/response scores. The slope of the regression line displayed in Figure 9 was not significantly different from zero (r2 = 0.0074), suggesting that a relationship between SDI overall percent scores and PASAT mean time/response scores was unlikely (Jonathan Berkowitz, personal communication, February 13, 2001). See Appendix Ff for SPSS printout of Pearson correlation analysis of the relationship between the SDI overall percent scores and PASAT mean time/response scores. Coefficient of determination One method of evaluating the importance of the correlation coefficient is to calculate the coefficient of determination (Domholdt, 1993). The coefficient of determination, r2, is an indication of the percentage of variance that is shared by the two variables and is determined by the square of the correlation coefficient, r (Domholdt, 1993). The coefficient of determination for the relationship between SDI overall percent scores and PASAT mean time/response scores was 0.0074 (r2 = 0.0862). Therefore, approximately 0.74% of the variability within one variable (SDI overall percent scores or PASAT mean time/response scores) can be accounted for by the other variable. The remaining 99.26% of the variability is due to variables not yet considered. Calculation of power To determine the probability of correctly rejecting a false null hypothesis (Ho: p = 0, where p or "rho" is the correlation coefficient in the population), a power calculation of the correlation between the SDI overall percent and the PASAT mean time/response scores was Relationship between Attention and Social Functioning 58 conducted (Jonathan Money, personal communication, January 18, 2001). This was done by using the Fisher's z transformation for the critical value of r and for the sample r. To describe the boundary of the rejection area, alpha (a) was set at 0.05. Thus, the critical value of r represented the score that cut off the lowest 5% of the distribution (Howell, 1997). The transformed values of r and sample r were z a and z respectively (Zar, 1984). The power of the test for Ho: p = 0 was 1- P(l), where (3(1) is the probability of the normal deviate. The equation to determine Zp ( )) was: Zp(i) = (z - z a)V n -3 , where n = number of subjects (Zar, 1984). The Pearson correlation (r) between the SDI overall percent scores and PASAT mean time/response scores was 0.086. According to the Fisher's transformation for correlation coefficients table in Zar (1984), the transformed r to z score was 0.0862. Based on the tables provided in Zar (1984), the critical value of the correlation coefficient (r o.o5(2),33) and the corresponding z 0.05 score were 0.334 and 0.3473 respectively. Using the above-mentioned equation to solve for Zp(i); Zp(i) = (0.0862 - 0.3473) V32 = -1.48. From the proportions of the normal curve table provided in Zar (1984), P (Z > | -1.48 |) = 0.0694 = p. Therefore, the power of the correlation between the SDI overall percent and the PASAT mean time/response scores for this sample of subjects was 1 - P = 1- 0.0694 = 0.931. A power of 0.931 suggested that a type II error was most likely avoided since there was a 93% probability of correctly rejecting a false null hypothesis. Confidence intervals for the correlation coefficient Although there was a 93% chance of detecting a correlation of 0.334 or more (Peter Liddle, personal communication, April 22, 2001), r is only an estimate of the population r because of the phenomenon of sampling error. Therefore to determine the meaningfulness of r, a confidence interval about r was calculated. Confidence intervals attempt to set limits that have a high probability of encompassing the true value of the population correlation coefficient (p) Relationship between Attention and Social Functioning 59 (Howell, 1997). Using Fisher's z transformation for correlation coefficient tables provided in Zar (1984), r was converted to z, and the 1 - a confidence interval was computed for (lower case Greek zeta which is a population parameter). The 95% confidence interval for £, = z ± to.05(2), ° ° cr z , where r = 0.086, z = 0.0862, t0.05(2) = 1.96, and oz = Vl/n-3 = 0.177 (Zar, 1984). Therefore the 95% confidence interval for C, = 0.0862 ± (1.96)(0.177) = 0.0862 ± 0.3469, or L, = -0.2552 and L 2 = 0.4078. In order to determine the 95% confidence interval for p, L i and L2 values computed for C, were transformed using tables in Zar (1984) that provided correlation coefficients, r, corresponding to Fisher's z transformations. The 95% confidence interval for p was L i = -0.2552 and L 2 = 0.4078. This confidence interval suggests a 0.95 probability that true value of p (the correlation in the full population from which the sample was drawn) lies between the values of -0.2552 and 0.4078. Since a considerable amount of 'noise' (influence of other factors) is associated with this type of data, a correlation approaching the end values of the confidence interval may still reflect a meaningful relationship (Peter Liddle, personal communication, April 22, 2001). Calculation of sample size However, because there was such large variability in SDI overall percent (s2 = 137.3) and the PASAT mean time/response (s =19.19) scores in the present sample of subjects, a sample size calculation was conducted to determined the required sample size to reject Ho: p = 0 with a specified correlation coefficient in the population of individuals with schizophrenia or schizoaffective disorder (Jonathan Berkowitz, personal communication, February 13, 2001; Jonathan Money, personal communication, January 18, 2001). The large variability in SDI overall percent and the PASAT mean time/response scores suggested that individuals with schizophrenia or schizoaffective disorder were not homogenous in regards to attention and social functioning. Relationship between Attention and Social Functioning 60 If the desired power is stated, (0.931) then the size of a sample required to reject Ho: p = 0 if it is truly false with a specified p * 0 can be determined. A sample size estimation requires the following equation: n = (Z p (i) + Z a / C,0)2 +3, where C,0 is the Fisher transformation of the p 0 specified, and the significance level a is two-tailed (Zar, 1984). Assume that the null hypothesis will be rejected Ho: p = 0.931 (93.1 %) of the time when |p| > 0.086, and the hypothesis is tested at the 0.05 level of significance. Therefore, p (i) = 0.069 and Z p(i) = 1.554 (according to the critical values of the t distribution tables provided in Zar (1984)); a (2) = 0.05 and Z a (2) = 1.9600; and for r = 0.086, z = 0.0862. Then, n = (1.554 + 1.9600 / 0.0862)2 + 3 = 1661. Therefore, a sample size of at least 1661 subjects should be used to confidently test a correlation coefficient of 0.086 between SDI overall percent and the PASAT mean time/response scores in a population of all individuals with schizophrenia or schizoaffective at a (2) = 0.05. Although a sample size of 1661 subjects could determine whether a correlation coefficient of 0.086 between SDI overall percent and the PASAT mean time/response is statically significant, the correlation coefficient could remain clinically insignificant (Jonathan Berkowitz, personal communication, February 13, 2001). Pearson correlation between the occupational functioning subcomponent of the SDI and PASAT mean time/response scores Addington et al. (1998) found a significant relationship between attention and the occupational functioning subcomponent of the SDI. However, when a Bonferroni correction was applied, the relationship was no longer significant (Addington et al., 1998). Similarly, a Pearson correlation analysis of the present study's data found no relationship between the total percent occupational functioning subcomponent scores of the SDI and PASAT mean time/response scores (r= 0.205, p = 0.240). The total percent occupational functioning subcomponent scores for each subject were determined by dividing the sum of all three occupational functioning Relationship between Attention and Social Functioning 61 subcomponents (dysfunction during past month, number of areas of dysfunction, and satisfaction with level of functioning) by seven that represented the total possible score. The resulting value was then multiplied by one hundred to obtain a percentage value. See Appendix I for SPSS printout of Pearson correlation analysis of the relationship between the total percent occupational functioning subcomponent scores of the SDI and PASAT mean time/response scores. Relationship between Attention and Social Functioning 62 CHAPTER 5 Discussion Several studies have examined the relationships between cognitive abilities and social functioning in individuals with schizophrenia. However, the relationships between cognition and social functioning are complex and there is little consensus as to which aspect of cognition is most relevant to social functioning (Green, Kern, Braff, & Mintz, 2000). This study examined the relationship between attention, as measured by the PASAT, and social functioning, as measured by the SDI. Although the results of this study suggested that a small but statistically insignificant relationship (r = 0.086, p = 0.624) existed between PASAT mean time/response scores and SDI overall percent scores, a closer examination of the relationships between attention and social functioning cited in previous research may provide a better appreciation of this study's findings. Appendix J reviews the relationships between attention and social functioning of ten studies cited in Chapter 2. These studies were selected because each provided statistical analysis data of the relationships between attention and social functioning. Appendix J provides a summary of the attention measure used, the aspect of attention measured, the social function scale used, number of subjects, and the strength of the relationship between attention and social functioning. Pearson correlations between attention and social functioning of these studies range from r = 0.02 (statistically insignificant, n = 80) to r = 0.8 (p < 0.001, n = 38). The beta correlation ((3) cited in two of the studies range from 0.01 (p < 0.04, n = 19) to 0.16 (p < 0.16, n = 205). Of the ten studies, six found statistically significant relationships between attention and social functioning, three found that some measures yielded statistically significant relationships between attention and social functioning while other measures did not, and only one previous study found no statistically significant relationship between attention and social functioning. Relationship between Attention and Social Functioning 63 Green et al. (2000) reviewed a number of studies that investigated the relationships between cognitive deficits and functional outcomes in schizophrenia and schizoaffective disorder. Using meta-analytic procedures for combining correlation coefficients, Green et al. (2000) summarized results across studies for four of the key cognitive constructs, namely secondary verbal memory, immediate memory, executive function and attention. Functional outcome constructs fit into three general categories: (a) assessment of success in psychological rehabilitation programs; (b) assessment of social problem solving ability in laboratory settings; and (c) assessment of broader aspects of behavior in the community and activities of daily living. Meta-analyses involved calculating the pooled estimate of r on the basis of weighted z values and tested for significant by reference to the normal curve. Appendix K summarizes Green et al.'s (2000) meta-analysis of key neurocognitive constructs with functional outcome constructs. The first column beneath the heading 'Functional Outcome Correlate' in Appendix K identifies the functional outcomes that were correlated with a particular aspect of cognition in four or more studies. The second column beneath the heading 'Functional Outcome Correlate' in Appendix K identifies the functional outcomes that were correlated with a particular aspect of cognition in only two or three studies. Box scores in Appendix K represent the number of positive findings compared with the number of null findings. Green et al. (2000) noted that a complete box score should also show "paradoxical" findings, that are significant and in the opposite direction. Since no paradoxical findings were found, Green et al. (2000) argued that this finding reflected "the rather consistent direction of relationships." Green et al.'s (2000) review summarized a number of findings that were reported in four or more studies. Secondary verbal memory was consistently related to all functional outcomes, and immediate memory was related to psychosocial skill acquisition. Executive functioning was associated with community outcomes, and attention (defined as vigilance) was linked to social problem solving/skill performance. Green et al. (2000) found that the aforementioned Relationship between Attention and Social Functioning 64 relationships between cognitive constructs and functional outcomes were highly significant with p values less than 0.000001. Green et al. (2000) contended that the studies had adequate power with sample sizes ranging from 188 to 1, 1002. The estimated pooled r's for the relationships ranged from 0.20 to 0.40, and the effect sizes ranged from small-medium to medium-large. In particular, estimated pooled r's for attention and functional outcomes was 0.20 with an effect size of small-medium (see Appendix K for details). Green et al. (2000) concluded that each of the four neurocognitive constructs have significant relationships with functional outcomes. In this study, the subjects' social functioning was assessed using the SDI. The SDI was designed to assess an individual's ability to maintain desired social roles and relationships within his or her community. Based on Green et al.'s (2000) categorization of functional outcome constructs, the SDI is a social measure that assesses the broader aspects of behavior in the community and activities of daily living. None of the studies in Green et al's (2000) review found a statistically significant relationship between attention and functional outcome measures that assessed broader aspects of behavior in community and activities of daily living. As Green et al. (2000) noted, functional outcomes such as community social functioning, are determined by a host of factors. Much of the current literature offers support for the notion that cognition is associated with social functioning. However, "rate-limiting" (Addington et al., 1998) studies, such as the present study, are not designed to determine how cognition is related to social functioning. Green et al. (2000) proposed that once relationships have been demonstrated, it would be reasonable to examine and identify mechanisms responsible for the relationships. A possible mediator of social functioning may be "learning potential" (Green et al., 2000). According to Green et al. (2000), learning potential focuses on latent capacity rather than on developed abilities. Green et al. (2000) stated that learning potential "is dependent on basic neurocognition and is related to psychosocial skill acquisition, but is not identical to either " (p. 131). The assessment of learning potential would require the application of "dynamic Relationship between Attention and Social Functioning 65 assessments" (Green et al., 2000). Essentially, dynamic assessment involves examining within subject variability though repeated assessment. With repeated assessment, influences on performance such as training and intervention can be studied. At present, few laboratories have developed dynamic assessment methods to study schizophrenia and further research is required (Green et al., 2000). Another possible mediator of social functioning may be "social cognition". Green & Nuechterlein (1999) postulated that "a cluster of mediating variables labeled social cognition" exist between basic neurocognition and functional outcomes. These authors suggested that components of social cognition include: emotional perception, social schema, insight into illness, and coping/attributional strategies (Green & Nuechterlein, 1999). According to Green & Nuechterlein (1999), "basic neurocognition is a prerequisite for social cognition, and social cognition, in turn, is a prerequisite for social functioning" (p. 315). Some research has supported this hypothesis. Penn, Spaulding, Reed, and Sullivan (1996) administered a battery of cognitive and social cognitive tasks to 27 inpatients with schizophrenia and rated their social function on various indices of ward behavior. Penn et al. (1996) found that deficits in social cognition were associated with impairments of ward behavior. However, relationships between social cognition and ward behaviors did not completely overlap with relationships between cognitive functioning and ward behavior. Thus, the relationship of social-cognitive deficits with ward behavior did not simply reflect the consequences of general cognitive impairment. Similar findings to Green et al. (2000) were observed from the studies reviewed as summarized in Appendix J. Three studies (Addington et al., 1998; Addington & Addington, 1999; Dickerson et al., 1996) used social functioning measures (SDI, SAS II, and SFS) that assessed broader aspects of social behavior in community and activities of daily living. None of the three studies found statistically significant relationships between attention and social functioning when the SDI, SAS II or SFS were used. Yet, Addington et al. (1998) and Relationship between Attention and Social Functioning 66 Addington and Addington (1999) found statistically significant relationships between attention and social functioning when the ALPSS was used to assess social functioning. The AIPSS is a measure of social problem solving. Unlike the SDI, SAS II, or SFS, the AIPSS assesses a specific aspect of social functioning. This study's finding of a statistically insignificant relationship between attention and social functioning appears to be congruent with the findings of other studies that used similar social functioning measures as the SDI. The remainder of this chapter will focus on: (a) possible factors that may have contributed to the lack of a statistically significant relationship between the PASAT and SDI; (b) strengths and limitations of the study; and (c) future research and implication of findings. Factors associated with PASAT and SDI that may have contributed to findings Factors that may have contributed to the lack of a statistically significant relationship between the PASAT and SDI will be reviewed in two parts. The first part will examine factors associated with the PASAT and the second part will examine factors associated with the SDI. Factors associated with the PASAT. One factor that was examined for it's contribution to the lack of a statistically significant relationship between the PASAT and SDI was the possibility that the PASAT could not detect attention impairments in the sample population. It was presupposed that, i f the research subjects' scores were not significantly different than the scores of the normative control group used in the development of the PASAT, then the PASAT would not be able to detect attention impairments in the sample population. To determine whether the research subjects diagnosed with schizophrenia or schizoaffective disorder scored differently on the PASAT from normal control subjects, the PASAT mean time/response scores were compared with the scores of the normative control group used in Gronwall's (1977) study. A comparison of the demographic characteristics between the present research subjects and Gronwall's (1977) control subjects was made to identify potential confounds that may influence PASAT mean time/response scores. Relationship between Attention and Social Functioning 67 Demographic information of the subjects used in the development of the PASAT is limited. The control group used in Gronwall's (1977) study included ten individuals from each of the age groups: 14 to 16, 17 to 25, 26 to 40 and 41 to 55 years. Additional data from forty first-year university students and ten naval personnel were included. Based on this information, the bulk of the subjects would have been between 17 and 25 years of age. The ratio of males to females and their levels of education were not specified. In the present study, there were 25 male and 10 female subjects, ranging in age from 17 to 49 years of age (mean = 28.4 years; mode = 21 years). Thus, the present study's sample population was similar in age to Gronwall's (1977) control group sample population. Subjects' level of education in the present study ranged from 9 to 22 years, with a mean education level of 13.4 years and a modal education level of 12 years. Half of the subjects in Gronwall's (1977) control group were first-year university students, suggesting a minimum of 13 years of education for those subjects. However, the education levels of the remaining subjects in Gronwall's (1977) study were not stated. Therefore, an accurate comparison of subjects' level of education between the present study and Gronwall's (1977) study was not possible. Although demographic characteristics of the subjects of the present study and the control subjects of Gronwall's (1977) study were not identical, a t value for testing the difference between two means was calculated in order to determine whether the research subjects scored differently on the PASAT than the normal control subjects. Table 8 displays the descriptive statistics of the present study's subjects and Gronwall's (1977) control group subjects. Since the smaller sample came from a population with the larger variance (n 2 = 35 & S 2 2 = 19.19) the probability of committing a Type I error is greater than a (Zar, 1984). The comparison of two means from normal populations with unequal variances is known as the "Behrens-Fisher problem," and one of the easiest, yet most reliable solutions is the "Welch's approximate t" test statistic (Zar, 1984). The test statistic is t = X i - X 2 / V (si2/ni + s2 2/n2). Relationship between Attention and Social Functioning 68 Therefore, t = 3.2 - 5.88 / V (0.063 / 80 + 19.19 / 35) = -3.5935. The critical value is Student's t with degrees of freedom of v = (si2/ni + S2 2/n 2) 2 / ((si 2/ni) 2 / ni - 1 + (s2 2/n2) 2 / n 2 - 1) (Zar, 1984). Therefore, v = (0.063 / 80 + 19.19 / 35) 2 / ((0.063 / 80) 2 / 80 -1 + (19.19 / 35) 2 / 35 -1) = 34.1. Using the critical values of the t distribution tables in Zar (1984), t 0.05 (2), 34 = 2.032. In a two tailed test, the null hypothesis is rejected i f 111 > t o.os (2), v Since 111 = 3.5935 is greater than t o.os (2), v = 2.032, the null hypothesis is rejected and a significant difference between Gronwall's (1977) normal subjects and the present study's subjects exists. It can be concluded that the PASAT mean time/response scores of the present study's subjects were significantly lower than those of Gronwall's (1977) normal subjects. Thus, the PASAT appeared capable of detecting attention impairments in the present study's population. Table 8 Descriptive Statistics for the Present Study Subjects and Gronwall's (1977) Control Group Descriptive Statistics Gronwall's (1977) control groups subjects Present study's subjects Number of subjects = n ni = 80 n 2 = 35 Degrees i f freedom = v v, =79 v 2 = 34 Mean of PASAT mean time/response scores = X X , =3.2 X 2 = 5.88 Standard deviation = S Si =0.25 S 2 = 4.38 Variance = S 2 S\ = 0.063 S 2 2 = 19.19 Another variable that could potentially influence the outcome of the study is the manner in which attention was conceptualized and assessed. Measures of attention can be presented in either a visual or auditory manner. Appendix L provides a summary of frequently cited measures of attention and the manner in which they are usually presented. The PASAT is a Relationship between Attention and Social Functioning 69 measure of attention that is presented in an auditory manner. Of the studies reviewed in Chapter 2, five used measures of attention that were presented in an auditory manner (Addington et al., 1998; Cornblatt et al., 1992; Dickerson et al., 1996; Freedman et al., 1998; and Kern et al. 1992). Addington et al. (1998) found no statistically significant relationships between the attention measure (CPT- auditory version) and the social function measure (SDI). Similarly, Dickerson et al. (1996) found no statistically significant relationships between attention measures (WAIS-R Digit Span and WAIS-R Arithmetic) and the social function measure (SFS). The other three studies (Cornblatt et al., 1992; Freedman et al., 1998; and Kern et al. 1992) cited statistically significant relationships between attention and social function measures. However, it is important to note that only Addington et al. (1998) and Dickerson et al. (1996) used social function measures that assessed broader aspects of social functioning. Other studies reviewed in Chapter 2 used measures of attention that were presented in a visual manner. The visually presented attention measures included the CPT, the SPAN, the Stroop Color-Word Test, the Symbol Digit Modalities Test, and the WISC-ffl Mazes (see Appendix L). Some of the findings across studies employing visually presented attention measures were inconsistent. Addington and Addington (1999) employed the CPT and SPAN to measure attention and the SFS and AIPSS to assess social function. Corrigan et al. (1994) also used the CPT and S P A N to measure attention, but used the SCRT to assess social function. Addington and Addington (1999) reported statistically significant findings between the CPT and the AIPSS, but not the SPAN (neither the CPT nor SPAN were significantly correlated with the SFS). Corrigan et al. (1994) reported statistically significant findings between the SPAN and SCRT, but not the CPT. According to Green et al.'s (2000) categorization of social function measures, the AIPSS and SCRT are defined as "laboratory assessments of instrument skills and social problem-solving." Although Addington and Addington (1999) and Corrigan et al. (1994) Relationship between Attention and Social Functioning 70 used visually presented attention measures and social function measures that were of the same category, their results were inconsistent. The majority of studies reviewed in Appendix J utilized some version of the CPT. Only two studies used the CPT and a social function measure similar to or the same as the SDI (Addington et al., 1998; Addington & Addington, 1999). Addington et al. (1998) employed an auditory version of the CPT, whereas Addington and Addington (1999) employed a visual version of the CPT. Neither of the studies found a statistically significant relationship between the CPT and the social function measure. Regardless of whether the attention measure was administered in an auditory or visual manner, there was not a consistent correlation between attention and social function. However, there seems to be evidence suggesting that relationships between attention measures and broad based social function assessments such as the SDI are often statistically insignificant. Therefore, the manner in which the attention measure was presented may not have influenced the outcome of this study. Additional evidence to support the argument that the manner in which attention measures are administered may not bias the results of the measure is provided by Trysennar and Goldberg (1994), who examined the efficacy of cognitive training of attention defects for an individual with schizophrenia. Trysennar and Goldberg (1994) employed the following visual and auditory-based attention measures: (a) PASAT, an auditory numerical measure; (b) Performance on the Rey Complex Figure, a test of figural attention; and (c) Selective Reminding Task, a auditory verbal attention measure. A l l three measures detected impairments in attention prior to treatment and detected improvement in function after treatment. Different attention measures assess various aspects of attention. Table 1 and Appendix J display various attention measures and corresponding aspects of attention assessed by each measure. It can be argued that i f the attention measure assesses a level of attention that is too difficult for most subjects, then the majority of attention scores will be skewed towards the lower Relationship between Attention and Social Functioning 71 end of the distribution and the possibility of establishing any correlation between attention and social function maybe difficult (Lyn Jongbloed, personal communication, April 4, 2001). To determine whether the level of attention measured may have influenced the outcome of a relationship between attention and social function, a comparison between this study and studies cited in Appendix J was made. Only three of the studies cited in Appendix J utilized a community social function measure similar to or the same as the SDI. Two studies (Addington et al., 1998; Addington & Addington, 1999) used the CPT and/or S P A N and one study used the WAIS-R Digit Span and WAIS-R Arithmetic tests (Dickerson et al., 1996). The CPT and SPAN simultaneously assess sustained and selective attention. Like the PASAT, the WAIS-R Digit Span and WAIS-R Arithmetic tests used by Dickerson et al. (1996) simultaneously assess all levels of attention, namely: sustained, selective, alternating and divided attention (Sohlberg & Mateer, 1989). Similar to this study, none of the three studies found a statistically significant relationship between attention and social function. Therefore, the level of attention measured did not appear to influence the outcome of the relationship between attention and social function. A criticism of the PASAT is its reliance on basic arithmetic skills to assess attention. Gronwall and Sampson (1974) argued that serial addition correlated only 0.24 with arithmetic skills. However, i f the majority of individuals assessed in the present study had severe arithmetic skill impairments, the PASAT results could potentially be biased. Dickerson et al. (1996) used two attention measures: the WAIS-R Digit Span and the WAIS-R Arithmetic tests. Appendix K briefly describes the two tests. In short, the WAIS-R Arithmetic test requires mathematical skills, whereas the WAIS-R Digit Span test does not. Dickerson et al. (1996) found that neither attention measure was significantly correlated with social function. Dickerson et al.'s (1996) study suggests that arithmetic-based attention measures do not unduly bias attention scores. There has been limited use of attention measures requiring arithmetic skills in research examining the relationships between cognitive and social functioning in individuals Relationship between Attention and Social Functioning 72 with schizophrenia or schizoaffective disorder. Further study may be necessary to determine whether reliance on basic arithmetic skills actually confound the results of attention measures for individuals with schizophrenia or schizoaffective disorder. Factors associated with the SDI Psychometric properties of reliability, validity, sensitivity to change, and clinical utility of the SDI were tested in three separate studies of individuals with schizophrenia or schizoaffective disorder (Munroe-Blum et al., 1996). Table 9 displays information about the studies and corresponding mean SDI scores. Larger SDI scores suggest a greater degree of social dysfunction. Mean SDI scores ranged from 28.86 to 38.21 and standard deviation scores ranged from 11.96 to 12.92. The mean SDI score for the subjects involved in the present study was 34.78 with a standard deviation of 11.72. The mean and standard deviation scores of the present group of subjects were comparable to the mean and standard deviation scores of the groups of subjects used in the development of the SDI. Furthermore, the subjects used in the development of the SDI had similar demographic characteristics to the subjects used in the present study. In general, subjects used in the development of the SDI were in their mid-twenties to mid-thirties, primarily male, and had high school with some college education (Munroe-Blum et al., 1996). Subjects used in the present study averaged 28.4 years of age, were primarily male, and had 13.4 years of education. Since the degree of social dysfunction and demographic characteristics of the present study's subjects were similar to that of the subjects used in the SDI's development, it is likely that the SDI was a reliable and valid measure of social dysfunction for the present study's sample. A possible explanation for the lack of a statistically significant relationship between PASAT mean time/response scores and SDI overall percent scores is that the aspect of social functioning assessed by the SDI may not be related to attention. Similar to the SFS, the SDI is designed to assess a relatively global social functioning spectrum. Addington and Addington Relationship between Attention and Social Functioning 73 (1999) argued that social functioning as assessed by the SFS was insignificantly related to cognitive functioning because the areas of social function assessed by the SFS may be disrupted only by relatively gross cognitive impairments. Although subjects in this study exhibited some impairment in attention, their cognitive functioning may not have been sufficiently impaired to affect their community social functioning as measured by the SDI. Another way of examining the SDI involves the notion of "microsocial" and "macrosocial" domains of functioning (Addington & Addington, 1999; Penn et al., 1995). Penn et al. (1995) described social problem solving and interaction skills as belonging to a microsocial domain and community social functioning as belonging to a macrosocial domain. Table 9 Studies involved in the development of the SDI Study 1 (Munroe-Blum's study as cited in Munroe-Blum et al., 1996) Study 2 (Babiski's study as cited in Munroe-Blum et al., 1996) Study 3 (Munroe-Blum and McCleary's study as cited in Munroe-Blum et al., 1996) Study Design Longitudinal Cross-sectional Randomized Control Trial Number of Subjects 33 67 113 Mean SDI Scores 38.21 28.86 37.28 Standard Deviation 12.92 13.99 11.96 They also suggested that functioning in the microsocial domain might be independent of functioning in the macrosocial domain. Addington and Addington (1999) argued that the microsocial and macrosocial domains might be opposite ends of a continuum. Thus, different measures of social functioning may assess different aspects of functioning along this continuum. Relationship between Attention and Social Functioning 74 For example, measures such as the SFS and SDI may assess more of the macrosocial domain, while AIPSS may assess more of the microsocial domain. Addington and Addington (1999) also suggested that social problem-solving measures (e.g. AIPSS) could be conceived as social information processing tasks and, as such, are closer to cognitive measures than are measures of community social functioning (e.g. SDI). Therefore, an association between cognitive functioning measures (as measured by the PASAT) and the macrosocial domain (as measured by the SDI) would be less likely than an association between cognitive functioning measures and the microsocial domain (Addington & Addington, 1999). Green (1996) argued that social function within the community was heavily dependent upon skill acquisition and social problem solving. Green (1996) stated that the relationship between attention and both skill acquisition and social problem solving had "a certain face validity". He theorized that individuals who were better able to distinguish signal from noise in computerized tests would also be better able to separate relevant from irrelevant information in the flow of continually changing social situations. However, at the time of Green's (1996) review, none of the studies involving community social function measures evaluated attention. Green et al.'s (2000) more recent review included studies that examined the relationship between attention and community social functioning. Although none of these studies found a statistically significant relationship between attention and community social functioning, Green et al. (2000) argued that "community social function can be considered the sum total of the acquisition and performance of key life skills" (p. 131). The present study did not find a statistically significant relationship between attention and community social functioning as measured by the PASAT and the SDI. Unfortunately, no comment regarding a relationship between attention and social problem solving or skill acquisition can be made since this study did not use social problem solving or skill acquisition measures. As Addington and Addington (1999) remarked, measures of community social functioning such as the SDI assess "how much" an individual is doing, (i.e., Relationship between Attention and Social Functioning 75 how many friends they have, how often they see those friends, how they spend their leisure time) but only indirectly assess the "quality" of social interactions and social problem solving. Assessments such as the SDI may lead the researcher to assume that i f an individual has many friends and leads an active social life, he or she likely has good social problem solving and interaction skills. Strengths and limitations of the study Statistical analysis suggested that subjects' demographic information did not appear to confound the results of the study. With the exception of a statistically significant correlation between the subject's level of education and SDI overall percent sores, Pearson correlation analyses did not reveal statistically significant relationships between subjects' age, level of education, and duration of illness and SDI overall percent scores or PASAT mean time/response scores. Similarly, independent t-test analyses did not reveal statistically significant effects of subjects' gender, diagnosis, and level of medication on SDI overall percent scores or PASAT mean time/response scores. To control for rater bias and drift during the study, inter-rater reliability procedures were instituted during the administration of the SDI and PASAT. During the training of the SDI administration, inter-rater reliability was determined by using a weighted measure of percent agreement. Inter-rater reliability of the SDI was confirmed at the end of the study by computing an ICC and Cohen's kappa. The SDI overall average ICC (0.96) and overall average Cohen's kappa (0.88) indicated that very good inter-rater reliability was achieved. To ensure reliable scoring practices of the PASAT throughout the study, the PI and a SRDP OT maintained a minimum 90% agreement during training. To control for experimenter bias, the SRDP OT independently scored all the recorded PASAT tests. The present study was based on the null hypothesis that there would be no relationship between attention and social functioning in individuals with schizophrenia or schizoaffective Relationship between Attention and Social Functioning 76 disorder, and the alternate hypothesis that there will be a significant positive relationship between attention and social functioning in individuals with schizophrenia or schizoaffective disorder. Unlike many of the studies investigating relationships between cognition and social functioning in individuals with schizophrenia (Green, 1996), this study had a guiding hypothesis. The lack of hypotheses in previous studies meant that most of the analyses were conducted in a post hoc fashion. Post hoc analyses capitalize on chance; so some of the significant associations in many previous studies may have been random occurrences (Green, 1996). Even though many studies found significant relationships between cognitive and social functioning measures, several of the correlation coefficients (Addington & Addington, 1999; Addington et a l , 1998; Dickerson et al., 1996; Lysaker et al., 1995; McKee et al., 1997; Penn et al., 1993; Penn et al., 1995; Penny et al., 1995) would fall within the low to moderate range (0.26-0.49 and 0.50-0.69 respectively) (Domholdt, 1993). Green (1996) argued that many of the studies were "underpowered" and "overanalyzed" since they attempted to analyze the results of several different measures with a limited number of subjects. A power calculation of the present study's results revealed that a type II error was most likely avoided since there was a 93% probability of rejecting a false null hypnosis. The Pearson correlation between the SDI overall percent scores and PASAT mean time/response scores was r = 0.086, p = 0.624, thus making critical r = 0.334. Consequently, there existed a 93% chance of detecting a correlation of 0.334 or more. Since the Pearson correlation r is only an estimate of the population r, a confidence interval was calculated to determine the limits that had a high probability of encompassing the true value of the population correlation coefficient (p). The confidence interval suggested a 0.95 probability that the true value p laid between -0.2552 and 0.4078. A significant correlation approaching the end values of the confidence interval may still reflect a meaningful relationship between the SDI and PASAT in individuals with schizophrenia Relationship between Attention and Social Functioning 77 or schizoaffective disorder. Further research with larger sample sizes would be required to test this possible correlation. This study differed from other studies that examined the relationships between cognition and social functioning in individuals with schizophrenia or schizoaffective disorder in that attention was the only aspect of cognition assessed. The use of a single cognitive test avoided the problem of applying a correction for multiple comparisons that would have been necessary had different aspects of cognitive function been measured (Peter Liddle, personal communication, December 14, 1999). Such a correction would have weakened the power to detect meaningful correlations. Unfortunately, utilizing a single cognitive measure limits the types of conclusions that can be drawn. For instance, although a Pearson correlation suggested that no statistically significant relationship between the PASAT and SDI existed, conclusions about whether or not impaired social functioning is related to other types of cognitive impairments cannot be made (Peter Liddle, personal communication, December 14, 1999). Furthermore, like many other attention measures, the PASAT measures not only aspects of attention, but also other aspects of cognition such as immediate memory and arithmetic ability. Consequently, precise claims about the nature of the information-processing deficit impairing social functioning may be limited. Suggestion for future research and implication of findings Attention is a complex construct. The conceptualization of attention will determine how attention deficits are assessed and treated (Green, 1993). A number of different clinical measures are currently used to assess attention; most are not theoretically based on any particular model of attention and are very different in their task requirements. The CPT has been considered a test of "vigilance" or "sustained attention", and has been cited as a measure of attention in numerous studies (Cornblatt & Keilp, 1994). Vigilance performance, that is the ability to respond rapidly to target stimuli presented over an extended time period, is a common Relationship between Attention and Social Functioning 78 measure of choice in the assessment of attention. Cornblatt and Keilp (1994) reviewed over 40 studies that used various versions of the CPT as the primary measure of attention. Although the ability to sustain attention over time is clearly essential to successful CPT performance, according to Cornblatt and Keilp (1994) the majority of studies that employed the CPT found that a decline in sustained attention over time was not the critical impairment in schizophrenia. Rather, deficiencies in processing capacity, that is, the amount of information that can be processed at one time, was the overriding deficit in nearly all cases (Cornblatt & Keilp, 1994). For instance, certain versions of the CPT, such as the degraded stimulus CPT (DS CPT) and the identical pairs CPT (CPT-EP) were designed to increase the information-processing load in order to detect subtle deficits in individuals who were at risk but not yet clinically affected by schizophrenia. The PASAT has been recommended as an effective evaluation tool of attention impairments and a good measure of information processing capacity (Gronwall, 1977; Sohlberg & Mateer, 1989). The PASAT is based on the concept that attention deficits will be manifest when the amount of information to be processed per available time exceeds the capacity of the system. Since the PASAT requires subjects to add 60 pairs of digits on each trial, the PASAT incorporates demands of vigilance as well as demands on divided attention. Gronwall (1977) demonstrated the PASAT's sensitivity to attention deficits. From a sample of 320 individuals with head injuries, 99% scored one standard deviation below the mean of normal controls. Weber (1986) noted that the PASAT has been used in a wide range of studies such as evaluating the effects of anesthetic drugs, mountain sickness and an antidote drug, night duty on doctors, and migraine. Tryssenaar and Goldberg (1994) used the PASAT to assess improvement in attention in an individual with schizophrenia. From a clinical perspective, the PASAT has certain advantages over other attention measures such as the CPT. The PASAT is relatively inexpensive, readily accessible to a variety Relationship between Attention and Social Functioning 79 of clinical practitioners, does not require extensive training or expensive equipment, and is relatively quick and simple to administer. The CPT however, costs considerably more, requires a computer to administer, and clinicians must have a Masters or Doctoral level of education to acquire and administer it. These are very pragmatic considerations for clinicians in a health care environment where resources are limited. Since the PASAT mean time/response scores of the present study's subjects were significantly lower than those of Gronwall's (1977) normal subjects, the PASAT appeared capable of detecting attention impairments in the present study's population. However, future studies can strengthen this finding by randomly administering the PASAT and more frequently used measures, such as the CPT, to a matched sample of normal control subjects and subjects diagnosed with schizophrenia. In matched assignment, subjects with and without the diagnosis of schizophrenia would be matched on certain characteristics thought to be potentially biasing such as age, gender, level of education, etc. This type of study would allow a direct comparison of the relative sensitivities of the PASAT and CPT to detect attention impairments in individuals with schizophrenia. Green et al.'s (2000) review found that attention was linked to social problem solving/skill performance. Hence, it may be appropriate to use a problem solving/skill performance measure such as the AIPSS with the PASAT when examining the relationship between attention and social functioning in individuals with schizophrenia or schizoaffective disorder. Occupational therapists often address the difficulties individuals with schizophrenia experience in social, occupational, and self-care functioning. Occupational therapists also are aware of the cognitive impairments that limit the functioning of individuals with schizophrenia, yet they rarely remediate cognitive deficiencies (Allen, 1995; Averbuch & Katz, 1992; Tryssenaar & Goldberg, 1994). The reluctance to remediate cognitive deficiencies in individuals Relationship between Attention and Social Functioning 80 with schizophrenia may be related to the limited amount of research investigating direct treatment of these cognitive deficits by occupational therapists. In today's economic environment of increasingly limited health care resources, the need for evidence-based practice is imperative. Clinical reasoning needs to be based on theory development that is supported by research. Attention and social functioning impairments are complex problems. However, a better understanding of the attention deficits individuals with schizophrenia experience and how these deficits interfere with functional behavior can offer direction for more comprehensive and effective interventions (Ducheck, 1991; Green, 1993; Green, 1996; Penn et al., 1995; Spaulding, 1992). Conclusions To date, research in this area of study has underscored the complexity of cognitive and social functioning in individuals with schizophrenia and schizoaffective disorder. Researchers have conceptualized attention and social function in a variety of ways. The lack of commonly agreed upon definitions restrict coordinated research efforts designed to provide theoretical understanding of complex concepts. In particular, it has resulted in research that has employed a multitude of attention and social function assessments and, at times, reported inconsistent findings. Thus, to ensure continued theory development and sound evidence-based practice in the clinical field, researchers may need to be more rigorous and consistent when defining concepts such as attention and social function. Evidence has suggested that attention may be correlated with social skill acquisition and social problem solving, but not with community social functioning. Although Green et al. (2000) have argued that community social function can be considered the sum total of skill acquisition and social problem solving, there appear to be additional factors that potentially influence an individual's community social function. This notion seems to be supported by the fact that many studies investigating the relationships between cognition and social function Relationship between Attention and Social Functioning 81 revealed correlation coefficients that fell within the low to moderate range. Given that attention is only weakly correlated with community social function, the degrees that skill acquisition, social problem solving and other factors related to community social function are potential areas of future investigation. Much of the literature investigating relationships between cognition and social functioning has lacked guiding hypotheses. Future investigations could better facilitate the theoretical understanding of cognitive and social functioning by using guiding hypotheses to identify and examine factors related to community social functioning. The findings of these investigations would be of direct benefit to mental health professionals, individuals with schizophrenia, and their families. Mental health professionals need to understand the factors related to community social functioning before efforts to remediate deficits or develop compensatory strategies can be considered. Individuals with schizophrenia and their families also need to understand what and how these factors affect activities of daily living. Heightened understanding can demystify the illness and promote a sense of empowerment in individuals with schizophrenia. Empowerment promotes active participation of individuals and their families in the treatment of schizophrenia. Publication of research findings in peer-reviewed journals is necessary to provide mental health professionals with the knowledge to help their clients. To ensure that individuals and their families have access to recent advances in the understanding of schizophrenia, researchers and mental health professionals may also need to share their findings in community based forums, such as public presentations. Although the relationships between cognition and social functioning in schizophrenia are complex, the combined efforts of researchers, mental health professionals, individuals with schizophrenia and their families will eventually unravel the complexities of cognition and social functioning. Relationship between Attention and Social Functioning 82 References Addington, J. & Addington, D. (1998). Facial affect recognition and information processing in schizophrenia and bipolar disorder. Schizophrenia Research, 32, 171-181. Addington, J. & Addington, D. (1999). Neurocognitive and social functioning in schizophrenia. Schizophrenia Bulletin, 25(1), 173-182. Addington, J., McCleary, L., & Munroe-Blum, H. (1998). Relationship between cognitive and social dysfunction in schizophrenia. Schizophrenia Research, 34, 59-66. Allen, C. K. (1995). Occupational therapy for psychiatric diseases: Measurement and management of cognitive disabilities. Boston/Toronto: Little, Brown and Company. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4 t h ed.). Washington, DC: Author. Ancil, R. L. , Holliday, S., & Ffigenbottam, J. (1994). Schizophrenia: Exploring the spectrum of psychosis. Toronto: John Wiley & Sons. Andreasen, N . C. (1982). Negative symptoms in schizophrenia: Definition and reliability. Archives of General Psychiatry, 39, 784-788. Averbuch, S., & Katz, N . (1992). In Katz, N . (Ed.), Cognitive rehabilitation: Models for intervention in occupational therapy, (p. 219-239). Toronto: Andover Medical Publishers. Bellack, A . S. (1992). Cognitive rehabilitation for schizophrenia: Is it possible? Is it necessary? Schizophrenia Bulletin, 18 (1), 53-50. Bellack, A. S., Morrison, R. L., Wixted, J. T., & Mueser, K . T. (1990). An analysis of social competence in schizophrenia. British Journal of Psychiatry, 156, 809-818. Benedict, R. H . B . (1989). The effectiveness of cognitive remediation strategies of victims of traumatic head-injury: A review of the literature. Clinical Psychology Review, 9, 605-626. Relationship between Attention and Social Functioning 83 Berg, R., Franzen, M . , & Wedding, D. (1987). Screening for brain impairment: A manual for mental health practice. New York: Springer Publishing Company. Birchwood, M . , Smith, J., Cochrane, R., Wetton, S., & Copestake, S. (1990). The social functioning scale: the development and validation of new scale for social adjustment for use in family intervention programmes with schizophrenic patients. British Journal of Psychiatry, 157, 853-859. Bowen, L., Wallace, C. J., Glynn, S. M . , Nuechterlein, K . H. , Lutzker, J. R., & Kuehnel, T. G. (1994). Schizophrenic individuals' cognitive functioning and performance in interpersonal interactions and skill training procedures. Journal of Psychiatric Research, 28(3), 289-301. Braff, D. L. , Saccuzzo, D. P., & Geyer, M . A. (1991). Information processing dysfunction in schizophrenia: Studies of visual backward masking, sensorimotor gating, and habituation. In Steinhauer, S. R., Gruzelier, J. H. , & Zubin, J. (Eds.), Neuropsychology, psychophysiology and information processing, (p. 303-334). New York: Elsevier. Brenner, H . D., Hodel, B., Roder, V. , & Corrigan, P. (1992). Treatment of cognitive dysfunctions and behavioral deficits in schizophrenia. Schizophrenia Bulletin, 18 (1), 21-26. Broadbent, D. E. (1958). Perception and communication. Oxford, England: Pergamon Press. Canadian Association of Occupational Therapists (1997). Enabling occupation: An occupational therapy perspective. Ottawa, Ontario: CAOT Publications A C E . Cornblatt, B. A. , & Keilp, J. G. (1994). Impaired attention, genetics, and the pathophysiology of schizophrenia. Schizophrenia Bulletin, 20 (1), 31-46. Cornblatt, B. A. , Lenzenweger, M . F., Dworkin, R. H. , & Erlenmeyer-Kimling, L. (1992). Childhood attentional dysfunctions predict social deficits in unaffected adults at risk for schizoprhenia. British Journal of Psychiatry, 161, 59-64. Relationship between Attention and Social Functioning 84 Corrigan, P. W., Green, M . F., & Toomey, R. (1994). Cognitive correlates to social cue perception in schizophrenia. Psychiatry Research, 53, 141-151. Cramer, P., Bowen, J., & O'Neil, M . (1992). Schizophrenics and social judgement: Why do schizophrenics get it wrong? British Journal of Psychiatry, 160, 481-487. Crow, T. J. (1985). The two-syndrome concept: Origins and current status. Schizophrenia Bulletin, 11 (3), 471-485. Cullum, C. M . , Harris, J. G., Waldo, M . C , Smernoff, E., Madison, A. , Nagamoto, H. T., Griffith, J., Adler, L. E., & Freedman, R. (1993). Neurophysiological and neuropsychological evidence for attentional dysfunction in schizophrenia. Schizophrenia Research, 10 (2), 131-141. Davison, G. C , & Neale, J. M . (1982). Abnormal psychology: an experimental clinical approach (3 r d ed., pp. 391-413). Toronto: John Wiley & Sons, Inc. DePoy, E., & Burke, J. P. (1992). In Katz, N . (Eds.), Cognitive rehabilitation: Models for intervention in occupational therapy, (p. 240-257). Toronto: Andover Medical Publishers. Deutsch, J. A . & Deutsch, D. (1963). Attention: Some theoretical considerations. Psychological Review, 70, 80-90. Dickerson, F., Boronow, J. J., Ringel, N . R., & Parente, F. (1996). Neurocognitive deficits and social functioning in outpatients with schizophrenia. Schizophrenia Research, 21, 75-83. Domholdt, E. (1993). Physical therapy research: Principles and applications. Toronto: W. B. Saunders Company. Duchek, J. (1991). Cognitive dimensions of performance. In C. Christiansen & C. Baum (Eds.), Occupational therapy: Overcoming human performance deficits, (p. 285-303). Thorofare, NJ: Slack. Dworkin, R. H. (1992). Affective deficits and social deficits in schizophrenia: What's what? Schizophrenia Bulletin, 18(1), 59-64. Relationship between Attention and Social Functioning 85 Dworkin, R. H. , Bernstein, G., Kaplansky, L. M . , Lipsitz, J. D., Rinaldi, A. , Slater, S. L., Cornblatt, B. A. , & Erlenmeyer-Kimling, L. (1991). Social competence and positive and negative symptoms: A longitudinal study of children and adolescents at risk for schizophrenia and affective disorder. American Journal of Psychiatry, 148(9), 1182-1188. Ellis, H . C. & Hunt, R. R. (1993). Fundamentals of cognitive psychology. Madison, Wisconsin: Brown & Benchmark. Freedman, L. R., Rock, D., Roberts, S. A. , Cornblatt, B. A. , & Erlenmeyer-Kimling, L. (1998). The New York high-risk project: Attention, anhedonia and social outcome. Schizophrenia Research, 30, 1-9. Golden, C. J. (1978). Stroop color and word test: a manual for clinical and experimental uses. Chicago, IL: Stoelting Company. Green, M . F. (1993). Cognitive rehabilitation in schizophrenia: Is it time yet? American Journal of Psychiatry, 150 (2\ 178-187. Green, M . F. (1996). What are the functional consequences of neurocognitive deficits in schizophrenia? American Journal of Psychiatry, 153 (3), 321-330. Green, M . F., Kern, R. S., Braff, D. L., & Mintz, J. (2000). Neurocognitive deficits and functional outcome in schizophrenia: Are we measuring the "right stuff. Schizophrenia Bulletin, 26(1), 119-136. Green, M . F. & Nuechterlein, K. H . (1999). Should schizophrenia be treated as a neurocognitive disorder? Schizophrenia Bulletin, 25(2), 309-318. Gronwall, D. M . A . (1977). Paced auditory serial-addition task: a measure of recovery from concussion. Perceptual and Motor Skills, 44, 367-373. Gronwall, D.M.A. , & Sampson, H. (1974) The psychological effects of concussion. New Zealand: Auckland University Press. Relationship between Attention and Social Functioning 86 Harvey, P. D., Sukhodolsky, D., Parrella, M . , White, L., & Davidson, M . (1997). The association between adaptive and cognitive deficits in geriatric chronic schizophrenia patients. Schizophrenia Research, 27, 211-218. Howell, D. C. (1997). Statistical methods for psychology. Scarborough, Ontario: Wadsworth Publishing Company. Ito, M . , Shiragata, M . , Kanno, M . , Mori, Y . , Hoshino, K . -y., & Niwa, S. - i . (1998). Social cue perception in Japanese schizophrenic patients. Schizophrenia Research, 34, 113-119. Jackson, H. J., Minas, I. H. , Burgess, P. M . , Joshua, S. D., Charisiou, J., & Campbell, I. M . (1989). Is social skills performance a correlate of schizophrenia subtypes? Schizophrenia Research, 2, 301-309. Kahneman, D. & Treisman, A. (1984). Changing views of attention and automaticity. In R. Parasuraman & D. R. Davies (Eds.), Varieties of attention, (p. 29-61). Toronto: Academic Press. Kazdin, A. E. (1982). Single-case research designs: Methods for clinical and applied settings. New York: Oxford University Press. Kelly, J. A. , Urey, J. R., & Patterson, J. T. (1980). Improving heterosocial conversational skills of male psychiatric patients through a small group training procedure. Behavior Therapy, ! ! , 179-180. Kern, R. S., Green, M . F., & Satz, P. (1992). Neuropsychological predictors of skills training for chronic psychiatric patients. Psychiatry Research, 43, 223-230. Kielhofner, G. (1985). A model of human occupation: Theory and application. Baltimore, M D : Williams & Wilkins. Kietzman, M . L. (1991). Information processing and schizophrenia. In Steinhauser, S. R., Gruzelier, J. H . & Zubin, J. (Eds.), Neuropsychology, psvchophysiology and information processing, (pp. 227-239). New York: Elsevier. Relationship between Attention and Social Functioning 87 Krefting, L. H . (1985). The use of conceptual models in clinical practice. Canadian Journal of Occupational Therapy, 52, (4), 173-178. Lezak, M . D. (1976). Neuropsychological assessment. New York: Oxford University Press. Liberman, R. P. (1982). Assessment of social skills. Schizophrenia Bulletin, 8(1), 62-83. Liberman, R. P. & Green, M . F. (1992). Whither cognitive behavioral therapy for schizophrenia. Schizophrenia Bulletin, 18(1), 27-35. Liberman, R. P., DeRisi, W. J., & Mueser, K . T. (1989). Social skills training for psychiatric patients. Toronto: Pergamon Press. Liddle, P. F. (1987). The symptoms of chronic schizophrenia: A re-examination of the positive-negative dichotomy. British Journal of Psychiatry, 151, 145-151. Linn, M . W. (1988). A critical review of scales used to evaluate social and interpersonal adjustment in the community. Psychopharmacology Bulletin, 24(4), 615-621. Loranger, A . W., Susman, V. L., Oldham, J. M . , & Russakoff, L. M . (1987). The personality disorder examination: A preliminary report. Journal of Personality Disorders, 1 (1), 1-13. Lysaker, P. FL, Bell, M . D., Zito, W. S., & Bioty, S. M . (1995). Social skills at work: Deficits and predictors of improvement in schizophrenia. The Journal of Nervous and Mental Disease, 183 (11), 688-692. Mandal, M . K. , Pandey, R., & Prasad, A. B. (1998). Facial expression of emotion and schizophrenia: A review. Schizophrenia Bulletin, 24(1), 399-412. Massel, H. K. , Corrigan, P. W., Liberman, R. P., & Milan, M . A . (1991). Conversation skills training of thought-disordered schizophrenic patients through attention focusing. Psychiatry Research, 38, 51-61. Relationship between Attention and Social Functioning 88 Mateer, C. A. & Mapou, R. L. (1996). Understanding, evaluating, and managing attention disorders following traumatic brain injury. Journal of Trauma Rehabilitation, 11(2), 1-16. McKee, M . , Hull, J. W., & Smith, T. E. (1997). Cognitive and symptom correlates of participation in social skills training groups. Schizophrenia Research, 23, 223-229. Moray, N . (1967). Where is capacity limited? A survey and a model. Acta Psychologies 27. 84-92. Morrison, R. L. (1990). Interpersonal dysfunction. In A . S. Bellack, M . Hersen, & A. E. Kazdin (Eds.), International handbook of behavior modification and therapy (2 n d ed., 503-522). New York: Plenum Press. Munroe-Blum, H. , Collins, E., McCleary, L., & Nuttall, S. (1996). The social dysfunction index for patients with schizophrenia and related disorders. Schizophrenia Research, 20, 211-219. Norman, D. A . (1968). Toward a theory of memory and attention. Psychological Review, 75(6). 522-536. Nuechterlein, K . H. & Asarnow, R. F. (1987). Manual and computer program for the U C L A continuous performance test (version 1). Unpublished manual and program, University of California Los Angles. Nuechterlein, K . H. & Asarnow, R. F. (1989). Contributions of the psychosocial sciences. In H. I. Kaplan & B. J. Sadock (Eds.), Comprehensive Textbook of Psychiatry. Baltimore: Williams & Wilkins. Oltmanns, T. F., & Neale, J. M . (1975). Schizophrenic performance when distracters are present: attention deficit or differential task difficulty? Journal of Abnormal Psychology, 84, 205-209. Relationship between Attention and Social Functioning 89 Pan, A. & Fisher, A . G . (1994). The assessment of motor and process skill of persons with psychiatric disorders. The American Journal of Occupational Therapy, 48(9), 775-780. Parasuraman, R. & Davies, D. R. (1984). Preface. In R. Parasuraman & D. R. Davies (Eds.), Varieties of attention, (p. xi-xvi). Toronto: Academic Press. Perm, D. L. , Mueser, K . T., Spaulding, W. D., Hope, D. A. , & Reed, D. (1995). Information processing and social competence in chronic schizophrenia. Schizophrenia Bulletin, 21 (2), 269-281. Penn, D. L. , Spaulding, W., Reed, D., & Sullivan, M . (1996). The relationship of social cognition to ward behavior in chronic schizophrenia. Schizophrenia Research, 20, 327-335. Penn, D. L., Van der Does, W., Spaulding, W. D., Garbin, C. P. Linszen, D., & Dingemans, P. (1993). Information processing and social cognitive problem solving in schizophrenia: Assessment of interrelationships and changes over time. The Journal of Nervous and Mental Disease, 181 (1), 13-20. Penny, N . H , Mueser, C., & North, T. (1995). The Allen cognitive level test and social competence in adult psychiatric patients. The American Journal of Occupational Therapy, 49 (5), 420-427. Perry, W., & Braff, D. L. (1994). Information-processing deficits and thought disorder in schizophrenia. American Journal of Psychiatry, 151 (3), 363-367. Piatt, J. J., & Spivack, G. (1974). Means of solving real life problems: Psychiatric patients versus controls, and cross-cultural comparisons of normal females. Journal of Consulting and Clinical Psychology, 2, 45-48. Posner, M . I., & Rafal, R. D. (1987). Cognitive theories of attention and the rehabilitation of attentional deficits. In M . J. Meier, A. L. Benton & L. Diller (Eds.), Neuropsychological rehabilitation (p. 182-201). New York: The Guildford Press. Relationship between Attention and Social Functioning 90 Rund, B. R. (1998). A review of longitudinal studies of cognitive functions in schizophrenia patients. Schizophrenia Bulletin, 24 (3), 425-435. Schneider, W., Dumais, S. T., & Shiffrin, R. M . (1984). Automatic and control processing and attention. In R. Parasuraman & D. R. Davies (Eds.), Varieties of attention, (p. 1-27). Toronto: Academic Press. Schooler, N . , Hogarty, G., & Weissman, M . M . Social Adjustment Scale II (SAS II) (1979). In W. A. Hargreaves, C. C. Attkisson, & J. E. Sorenson (Eds.), Resource materials for community mental health program evaluators. ( A D M Publication No. 79-328, pp. 290-330. Sohlberg, M . M . & Mateer, C. A . (1989). Introduction to cognitive rehabilitation. New York/London: The Guilford Press. Spaulding, W. D. (1992). Design prerequisites for research on cognitive therapy for schizophrenia. Schizophrenia Bulletin, 18(1), 39-42. SPSS Base 10.0 Applications Guide. (1999). Chicago, IL: SPSS Inc. Tammivaara, J., & Shepard, K. F. (1990). Theory: The guide to clinical practice and research. Physical Therapy, 70(9), 578-582. Tower, P., Bryant, B., & Argyle, M . (1978). Social skills and mental health. Pittsburgh: University of Pittsburgh Press. Treisman, A. M . (1964). Selective attention in man. British Medical Bulletin, 20(1), 12-16. Tryssenaar, J. & Goldberg, J. (1994). Improving attention in a person with schizophrenia. Canadian Journal of Occupational Therapy, 61(4), 198-205. Van Zomeren, A. H. , & Brouwer, W. H. (1994). Clinical neuropsychology of attention. New York: Oxford University Press. Wallace, C. J. (1986). Functional assessment in rehabilitation. Schizophrenia Bulletin, 12(4), 604-630. Relationship between Attention and Social Functioning 91 Wallace, C. J., Nelson, C. J., Liberman, R. J., Aitchison, R. A. , Lukoff, D., Elder, J. P., & Ferris, C. (1980). A review and critique of social skills training with schizophrenic patients. Schizophrenia Bulletin, 6(1), 42-63. Weber, A . M . (1986). Measuring attention capacity. Unpublished doctoral dissertation, University of Victoria, Victoria, Canada. Wechsler, D. (1981). Wechsler adult intelligence scale-revised. Psychological Corporation, Cleveland, OH. Wechsler, D. (1991). Wechsler intelligence scale for children (3 r d edition). Psychological Corporation, San Antonio, TX. Weissman, M . M . (1975). The assessment of social adjustment. Archives of General Psychiatry, 32, 357-365. Weissman, M . M . , Sholomskas, M.A. , & John, K. (1981). The assessment of social adjustment: an update. Archives of General Psychiatry, 38, 1250-1258. Wickens, C. D. (1984). Processing resources in attention. In R. Parasuraman & D. R. Davies (Eds.), Varieties of attention, (p. 63-102). Toronto: Academic Press. Wolery, M . , & Harris, S. R. (1982). Interpreting results of single-subjects research designs. Physical Therapy, 62, 445-452. Zar, H . J . (1984). Biostatistical Analysis (2 n d edition). Prentice Hall, Englewood Cliffs, New Jersey. 00 a c O o GO T3 c c o CD > C '"5 CO c2i O QH • CD 4 3 o © 53 W "?3 CZ) Q CD CO S3 O D H o cd H-J S3 o CD <D >~. CD O S3 CD S « <-3 jj 43 4H 3 O 5-1 > 1 CD S3 g CO O P H 4 3 CO CD 0 0 S3 O P H X CD *c3 • I-H o Ct-H < u 42 & IS C o CD co S3 o P H CO CD l-c CD •a i-(D S3 CD 0 0 4 3 CD cd CD CM O CO CD o G CD 31 CT CD co S3 O - o • CD CO - C D 4 0 cd 31 o £ 13 CD OT Relationship between Attention and Social Functioning 93 EARLY ATTENTION DEFICITS INABILITY TO PROCESS INTERPERSONAL INFORMATION interpersonal cues/communication = most complex and subtle information to be processed in environment ACTIVE AVOIDANCE OF OTHERS REDUCED STRESS SYMPTOM CONTROL ATTEMPTED INTERPERSONAL INTERACTION INCREASED STRESS SYMPTOM EXACERBATION Figure 2. Hypothesized role of attention deficits in the development of social dysfunctions among persons with a genetic susceptibility to schizophrenia (Cornblatt & Keilp, 1994) Relationship between Attention and Social Functioning Relationship Between Subjects' Ages & SDI Overall % Scores Rsq = 0.0046 Relationship Between Subjects' Ages & PASAT Mean Time/Response Scores 10 Rsq = 0.0208 30 A G E Figure 3. Scatter plots of the relationships between subjects' ages and SDI overall percent and PASAT mean time/response scores Relationship between Attention and Social Functioning Relationship Between Level of Education & SDI Overall Percent Scores 95 Rsq = 0.1369 10 12 14 16 18 20 22 24 Years of Education Relationship Between Level of Education & PASAT Mean/Time Response Scores 30 1 Rsq = 0.0221 8 10 12 14 16 18 20 22 24 Years of Education Figure 4. Scatter plots of the relationships between subjects' level of education and SDI overall percent scores and PASAT mean time/response scores Relationship between Attention and Social Functioning Relationship Between Months Diagnosed & SDI Overall Percent Scores 50 o ,y 40 D CO • n • • • • • • • • n • Rsq = 0.0033 Months Diagnosed Relationship Between Months Diagnosed & PASAT Mean Time/Response Scores 30 D • • • • • r-rUS D P u n B> A ° CD ° u a • -100 0 100 200 300 Months Diagnosed Figure 5. Scatter plots of the relationships between the number of months subjects diagnosed and SDI overall percent scores and PASAT mean time/response scores Relationship between Attention and Social Functioning Boxplots of SDI Overall Percent Scores by Subjects' Gender 70 (U 50 Q W 20 10 °J . N= 25 10 Male Female G E N D E R Boxplots of PASAT Mean Time/Response Scores by Subjects' Gender 30 Figure 6. Boxplots displaying data distributions of SDI overall percent scores and PASAT time/response scores by subjects' gender. Relationship between Attention and Social Functioning Boxplots of SDI Overall Percent Scores by Subjects' Diagnoses 70 Boxplots of PASAT Mean Time/Response Scores by Subjects' Diagnoses 30 °J . . N = 26 B Schizophrenia Schizoaffective Diagnosis Figure 7. Boxplots displaying data distributions of SDI overall percent scores and PASAT time/response scores by subjects' diagnoses Relationship between Attention and Social Functioning Boxplots of SDI Overall Percent Scores by Medication Dosage Levels 99 70 60 Q> 50 <n 20 10 °J . . N • 27 a Lower Level Higher Level Medication Dosage Level Boxplots of PASAT Mean Time/Response Scores by Medication Dosage Levels Figure 8. Boxplots displaying data distributions of SDI overall percent scores and PASAT mean time/response scores by medication dosage levels o o '3 CD P -O N 13 o 00 03 3t T3 "> -3 ,3 p cxD P '3 O '•+-» o e o o 00 P o o U p CD CD 40 CO P H C o 'HH "CD CD 43 H-» -o CD P X w t3 43 -4—» CO CD '•3 35 CO HH o cd =3 CO O H cd >H <u 43 H 13 o '5b o • "o 43 o . co PH T3 CD -H> CO <D > HH O H O to T3 S3 <D O co ^ 'C c O . cu a & O ID O " f t S3 O 3 o cD IS -S SO CU T3 S3 cd •e <D > 1) > "o co CD O ' CD CM} S3 CD S3 ^ g 43 >H HH cq o « 43 cd T3 ~ T3 CO - - H « s O „ '+3 co S3 33 !-. co « CD .S3 13 & cd o W o . „ c^d ^ ^ -b o £ P< 2 o o cd CD S o cd i n X » -4-1 13 O H S3 CO \n <u c ID 3 o . 3 £ 43 O s co co • 13 SH <D & CD O . 6 .S PH £°1 fx, - 3 < n -3i CD r3 •5 c2 " .5 8 y .5 c<D 3 u PH O 60 PH 1$ cd S i= i2 3 « CO 13 cd 3 o O CO CD •3 CO cd 43 O CD 4? 3 co CD CD J3 HH o s > •<§ £ .13 bo co 3 '3 2 C T 43 O *^ cd f3 -a 2 CD M H-» 3 13 43 S3 •3 -2 cS t3 * s •Z? CD ft 0 ^ $S .J2 <3 8 4-4? .3 3 43 <" T3 o cd 'S £ CD f _ i ft co O H CO O — N cd 13 c "3 O T3 T3 CD cd H 3 E 3 o 2 bb u T3 f CD k-H 3 CO cd CD t>0 S3 '3 o H-» o c o o ICO co CD to 2 as 00 ON 1 S3 '3 o H-» o S3 c3 ft S3 4 ? t "3 -S cd I ^ £ & HH CO 3^ S C HH - « S O 3 £ « cu P co O H " „ CU & O H >H S3 F » O . 3 ^ 0 13 J* ^ > o S3 cu 43 S t"1 Cj_i co cu O RT 43 . « CO £? .S3 H-< qj ^JJ CU ^ 3 40 S3 '" 4^ O 3 CO IH cu 43 O S3 cd > O &J0 IS o CO eg o ^H S3 cd cu '71 <H 8 I CO > 00 S3 '3 O 4H CD ,1 CD > 16 CO l l l3 ro 3 cd cd CD l-H 2 ' S Q '3 CD ° Id > o ft 5 c3 id N 00 E« 43 ft IH <D , , CQ pd O U Q PQ co . r~] O > . I) Si O 43 O co S3 co co < CN © M a '3 o o C O 00 13 a cd a o c 1> 43 a . IS CO c o Pd P p • i-H -4—> PI O o P CU p. . 4> cd j 00 g '3 o o c o o loo =5 o cd 3 oo £ S3 CU ^ P 4^  o co s cd ^ H cd S3 -2 H - H rp "o o C P cu O H .2 A > cu cd a , 43 cd <u Oc rO >. <D (D +J cd 'C P, o JH 00 P. c a . -ft id 13 co g cd O r~ & H H CO o cu " c/T P (H cd 13 P 2 cd cd 14 43 S ° 2 P •jz! co co CD co p ft O 00 P CU 5 "2 • O P co P - 2 >•> O cd CO <U P Hocd -H> P O o -HH CCl b - . - P 4V <D •g ^ ft CO & i S • H U H P H H H O H O cd s > cd 43 -cu P 4= - a cu R > cd P 8 I 1 T3 O o p cu P 00 g 'B o P CU > o IU <3 |3 cd 00 p" <D cu I > H P O N CU O N W C o 42 45 co O H H 43 O as o o tH 43 P 13 £ P H 5 CD 2 £ f c2 -2 ^ is 00 cd .3 IB •3 e cd £> -P O H HH y CU 40 p cd o •p 3 cd i cd •£ ° > 00 !> g 13 p B O cd p s c3 8 P CD * H IT 3 -P o +3 S S W O. 3 s CD .2 x -P <a P „ <D S- P fD cd CD y ^ - f l CD CO "p 13 •2 S o. e '•P ° P P cd g O. HP 43 .P CO P O P o P CD cd CD > t5 o H-» PH '3 3 P <43 O CD p cu 13 o y P 43 CD " P M H3 >• CO 13 cu > O l-o u .g > H-> cd 00 <D P 13 P cd P O •2 i3 cd > O H -5 cd 13 13 cd H H P 3 ° £ r- - r n -J3 C HH o cd 4-* a cu cd > 13 P cd cu 00 J3 CD cd P " co *i co cd S s H - CU -& ft c 3 o cd b to* cd £ > s •P o 5 Id .« 6 oo iT i 3 S 3 3 O co cd 43 T3 •3 3 3 ^ oo y 33 & P o ffi U NH C7N O O N 8 o cd co CD P cd co P H p o 00 o § .3 ° .3 3 S cd 2 •& P O H .2 2^ cd |2 CO O H CO O 00 a 'a o o a PH "O 00 C cd C O & CO C o <D PI • l-H » PI o o •3 CD 2 P H o N 4 3 o 00 cn 13 pi -a ' > PI • l-H OX) PI • i-H P! O • i H H H o Pi pi ."2 '3 o 00 PI o a o U c CD CD CD 4 3 Ui P H • l-H 4 3 cn PI O 'jS ID 4 3 -4—» T3 CO C I X W <B (D "•3 =3 H—> 00 <HH o & E S pi 00 00 p 'S o o c o o loo M l P 'S O -*-» o ,1 > i3 H «8 l l |3 cd o 00 P PS 3 £ .2 <*} w O cd 3 m PH M ^ " cj o 3 '5 ?! o o oo C J 43 cd J-P - o , M O > C 3 " . „ S M O » to (U ™ O cd S "3 S o o o cd O . . m T3 ^ C 00 O c/0 cd w _« o o oo PH OO r-, H. JO O w *g 4V PH 3 C OO cn O <L> ' > ^ X - ^ o ^ «•> o c3 £ C J t o o 05 H 6 g CD Cd O W P S £ . 2 * 11 § o p. O O cd cn >H C J 43 < §b o CO Q c S • •—< -+-» • 2 <L> P . -5 cd ^ 2 § o M S J3 43 « 5 . 2 ^ <U cn - C 6 o H  rH o -o •S s O p ^ SH O <U 1.1 11 > !3 P « C g HH H • 3^ u u S U 42 u 00 *H 3 O lH P , cn co O O O cci 00 •S ^ -5 .> 42 C o i - H cd cd cd ^H S 3^ T3 cd w £ O0 g u HH o CO 1 ' ^ CO OO 0 1 Q cd h j CH c« o o > oo 00 oo td « P 4) 42 C? o « HH o cd o 4^  ' ^ OH 00 i HH 00 •^ H HH . £ P < oo w 3 fe" 51 i 4> — T3 c? o S x — — (L> P . CO ET co • S3 u t O cn J3 lo 2 43 „ o cd <u o 00 I—I cd •H2> cd i3 N S '<3 cn H cn cd oj 00 cn (H cn O cd T3 4J M cd S3" '^11 <0 N pi s 1 i H — O M » CH O O °3 43 -t-" cn O H e 3 u oo 6 E u ^ 00 ^ 00 C M • S 3 " ^ .5 — O § P3 «5 „ 'M 5 2 i n S C O u ^ o .S o s CO . ^ , , P cd O 3 60 R 'S u u >H o oo c^S U Id > o p o O 'JH ^ j3 -a . . co U O JO M w CO cd u kT -rt co [3 " a 3 ^ ^H-p, cd < 3 S 2 P . S °» P . T 3 ID <U cn S> cn P 8 C cd H cn o p 3 1) cn .—, HH H< co o o p o tH o PQ c o co l-p cj H< cd PH <$ Co 00 P Os o cj o > oo ' s ^ ob cd o Q P4 o o 3 CJ SP cn p — CJ .2 "5 & PH O O0 f 43 3 o o 60 a 'a o o a 3 PH o 00 - a a cd a o 42 &• IS CO c o -a CD P » p o o p CD OH OH < b0| R 'S o o p o 15 s o p • l-H R HH o P CJ > al is i l i3 00 -a , - lH 8 co M U . P •2 ^ ° cd CO • S H ( ) 3 ^ 3 1 s a «. C O M _ L R » CJ C O H ^ O P H cn 3 U ret 2 B o C+H T3 CJ > >H j p cd r t3 > C+H cj O § .2 • S CJ co -5 cj cd S ' fa O cd >H > P H TD co P P cd o P ^ a co O CD ' P tH P C/~l CD ' S £ CD P T3 P CHH Cd CD nb 3 H= C O ,J3 R cd cd ^ « CD y P T3 w > oi o cd cd fa 00 -ti p o R CD P I < CD H P CD P O O .-P. P - p •?• bO CD CD CD ^ rr H •P ^ o 6 n a j H C O C O ^ H P cd C O o CD CD CD " cd ^ CD cd H P > ^ cd .2 cd CD o l-H co r p P HP O CD U C O _ P fa CD p , 0 C N C O CD 5 ~ cd cd P , O • 00 00 C O 5 3 .fa p ^ 'cd CD t ! P^ P to H ON CD fa ON HH O cd "o 'S •H CD • S O N T3 -5 IP P cd o cd P H 00 oo W " CD C O CD g £ B B bo O C <HP P 2 .2 B P CD p z a § O CO O cd co X> CD co -P 5 H fe CD H P H co PL, PH W O <HP CD IH CD •a o Q , M P H o « co '+3 CD r-cd cjp B % s i 3 § C J ° t cj cd T 3 p . H5 6 - H CJ g S i C O ^ ~ I I P ^ co - S i f cd t>0 P -P :s. CD UH P=H H . U CD 1^ >^  H CD P CD C J \ ^ s .2 - p ^ g bO cd S is Jxi C J PH o o co •P cd" ' p « •S I cn O oo g '3 o o a s PH o 00 c o HO CH IS co a o CM T 3 CD P p • l-H -4—» p o o < X ' l-H p CD P H c3 • l-H P CD P H O N • l-H HP o 00 > • i-H T3 _P M P 'p H o p p UH o o 00 60 P 3 HH o ,1 cj l eg 60 R '3 o HH o ,1 CD > 15 H in l-H l l |3 00 c > o oo H P o CD HP 3 ^ CD cd O CD £ 4 . 1 t> P lH PH » 00 CD ^ co 00 CD HH ft O 03 00 rR r/1 T3 p CD p p 0 8 P T3 5? ci P P HH 3 c R CD R O o co CD a CD .1 CD0 K > HO PH 1 3 CD 2 o k T3 ^ CD U co CD CD > ^ -3 13 co S3 s § « o3° a g ^£3 co > P , CD O rP CD ^ 53 . id id -a CD g in CD O 4= CO +-» S .3 « 13 3=3 ,£ ) r/i O ' 1 S R o 5 CO CO CD CD (H .—I CD O f£ ccj CO CO CD CD - P ; p co O p o "3 CO 03 CD y Pi 3 R P R a CD P PH 00 6 C o CD CD t ! PH -O cd I d CD CO CO HH CD CO CO CD CO •<-. °* co cd cd -7; £ °< CD CD 45 o O R >-£ T d £ 8 & p* CO <*J £P CD •3 -a cd CJ S co w R R o S P -CJ A 8 « _ R cd O ' o CD O T d co cd Id 3 43 CD I f 13 cj •ii co • cd u co i d H lH CD O f g CO 43 HH CD P 'cd O co i d CD CO cd rP cj 00 8 .S R 4 ^ 0 y O O CD 1 >-! CO • g p cj o CD OH ^3 Id C J 3 in cd g PH — > 00 "S co CD CO H—• CD 03 p >-CO (H co 7 ^ CD ^ 3 CD P CD O p —' 3 00 CD C &3 R cd O 43 O CO op .3 ' P 43 CD CJ 60 .ti ^ fe CD > cd co 43 o PQ CD CO o P c ' 0 r3 ^ C IT) N O o 00 c a o a 3 O O 00 T 3 a cd e o P < CL> a, IS CO c o -tH "u ed -a p a o o p O H 9 00 g '2 o H H CJ ,1 l5 00| g 'S H H o ,1 CD > al |3 3 l l |3 oo •a | CO H P > . O > P cd C •ea 43 /»l CJ Mrf ° cd X co o T3 b HH O P H O -P 2« e -«-: 0 3 ft C D _ o S S 'P P cd rt C D ' CO C D VH CO X! C D cd ° 4? CD . „ CD CO P CD CD b CO p in cd P H < £ CD 'in P o 1H CD cd T3 3 § 2 P O HH 00 o " « M « u 3 o 8 & p -S p o T 3 CD cd >-> 43 J3 P Id O C 9 cd r-* *H P CD ft & cS HH •2 & « M l s I I I < D CD -ft cd cd .3 T3 cd cd 2 O 00 p cd O HH o C D C D IH 43 C D +-> S CO cd P O 45 cd <D P C D HH "P M l ^ + H * - > co "P P l - H C D O 3 p - g C D w cd > T3 x> a - a CN • « _..<?> ,2 o CJs HH O UH W CD w - cd 4> P P3 co P 'S cd > HH n; a ^ • ^H CHH CO o H? CO cd C D 5 >H p o •p p cd H? o s CD 6 HH S ° c •2 2 o .p E ' P O cd •ft * P O H UH cdCDO H B CD O H CD y ed P TJ O a 8 B B CD >-> CD >H co J H 00 2 .S B CD cd 43 O > > +3 lH CO C O H O H £ CD o * ~ o 3^ cd -a 'P P cd P O -4-» cd 3 C D IH < D ^ • 2 IH O 1 — 1 (D CD cd P H £j W cu cd CD 3 o 43 o CD +z O . CO O FT p — w g cd C D o co" 00 CD p CD CD T3 cd H & " P O s 8 P 43 (3 00 cd C D CD " o cd co — CD O •p H o G • ft CO H-» " I H CO CD « B .B-g ed u oo i . U to 00 CD x ' H—» HH CD CO CO CD - H t—i cd P ^ 2 o P m a r Pd .23 4H P H C D — O ,H o; 5 b S o P H J I _ CD CD O H 1 § o s 2 2 H U S -S o 00 a '3 *H—> o a 3 PH o o 00 T3 a cO e o e < CD C J £ H—» cu 42 a. 13 CO a o CD P . p » S3 o o X 53 CD PH . CD 00 p 'p o H-» o p o o loo 60 p 'p o H-» o ,1 CD > |3 l l p o o p , 3 „ 00 co <I 5S •P © H • 2 oo CD o "2 "3 ^ 52- e " P ^ CO o - P T T .S 43 < O T3 CD 43 43 T3 CD co CO CD CO 60 P oo oo ° P P O CD £ P3 CD ~ £ CD •s « c o T3 CD CO CO 42 CO CO P CD O VH • p « CO * o .2 a .2  oo % < p oo eg * 1 3 ° l-H CO y p £ ° o -jr. -a P< p "C CO o „ CO « *J CJ T3 S CD CO > O P CD » H CD fe on Q -a 2 o i—i co 00 . P T3 co • r co O co >-H CO P. CO p o PH "a CO • H CJ > pi CO g O cj c2 . « 2 CD g § p O a . CO "P U 2 CD -C O 3 J H u O CT P3 PH CD § 6 P p ' cd CJ > PQ 42 P CD O O Pi <~> . P / - N CD OO CD w 42 P O ' l-H o P CO CO CD O PH — O CO N ^ ^ O O 00 0 0 oo o 6 0 S3 '3 o o c 3 o CVO c cd 13 O cu < a CD & c/l a o ed T3 cu _S3 £3 O O X 53 CD O H . <D &0 g '3 o •4—> o ,1 o o I CO bO g '3 o H-> o ,1 cu > al icS H CO l-H 13 C JD CTs cd ,g ON m (3 > O cd ^ w!-S '-g '3 § § a § | - H O cd N c e o • - 3 • r 3 L l TO I N 3 3 « o u < < Z co &o O N O 60 g '3 o U H o o GO a C3 a o 43 & 4 3 CO S 3 s 1 . p S3 o CD < X • l-H -a S3 C U S3H < 2 '2 <D J3 SOH o N 4 3 O C O cd 31 • O • i-H > 4 3 S3 • l-H 0 0 S3 '3 _ o 'H—» CD S3 P HH CD o C O 7 3 S3 O a o U S3 CD CD CD 43 CO S3 O • l-H 43 "cD PH CD 4 3 T3 CD S3 X w td 43 -U-» co CD T 3 3i HH C O 3 co . CD 04 CO C D 00 S3 'S 'H-> O ,1 o o I c o so S3 3 O o § PH C D & C O cd cd -B C O CD S ^ cd O 4^ 2 v P s ^ r -SPH M F T , C D to & ft3 S ' V P .3 § C D •S3 cdcd • .2 o o c o UH O ft C D C O cd J3 "cd > cd 4H - o P 1 1 3 P cd h C D C O 3 .3 bO p 3 o + J CD S3 , 3 bO S3 'Si •2 ' p o ,1 CD > a |5 i l |3 CD 43 H |5 C D as C D 3 ON cd -a O p P — o 0 3 S ' - P C D S .3 > p o J3 u cd ^ PH Q - P c C D 43 • C • £ P - -£3 P - b o o S cd N C D Sb'C IS IE? co f j P o S3 CD P cd co 43 ,2 H 1 Cw • P P P O • l-H CO C O " S3 9 -2 C O ' ' — } H C T • O £ • p o =a co T3 -*H C D C D O H - ON cd PH <, CD cd co • H p o C D T3 i3 U H-» V <D -rf 43 1 , 3 M-H O S3 C D H P . o o cd « 2 ID g Id * co cd P *H co C D « a i l P <D * l 43 > 2 - p HH C D CD- T;, p cd • P 3 > cd c O o £ > "*H P! -3 cd P o > <D O ' o > S3 O HH cd co P cd cd p 43 co Z " P i*^ CD . " P co 3 (V> i—I 43 & ?D J3 ;^ co P 43 O 3 O co C O 3 C D -p J D P H 43 "cd „ o o < U C D co JH C D 43 O P H P HH C O C O H ^ 1 o IH C D seal an seal era C D " ub en 3 b0 H^ C O o ion 43" HH PH H-» bO cd P C D " C O C D N < D > »-H cd • P bO P o C D " P C D O U 3j S3 " o cd UH C D . g pea: stur ms S3, C D C D cd 00 •t3 C O p p O - P ft co O 8 ^ 4 3 3 ?P4^  C D P •- .3 O C - P l . i g 43 t; •l-H W Q J H CD > 3 ^ P », .2 UH "ft ° 2 CD 3 » co 'cd T3 o S.< CD > co-> . C O 43 < D . U H " P o C D o P C O H CLf U - P S-l O UH I 8 P S3H P H C O w o C O C O cd 4 3 & P " * C O H C D c o ^ C D b0 | 43 P O 2 o ON S3 CJ C D < - p H H ? "5 ^  > P3 co > U H o HH a O -T3 P £ . cd 3^ S cd CD co 2 H- . O ° ft CD P< 44 c„ C D 43 T3 a o o 6 u Id ., P P o M CD P3 > HH > ' cd << cd g cu co .5 o' CD P 4H « O C D C O 4 3 n O 3 HH -a ° 3 7J <D S 3 HH 3 ^ P C 4^1 C O C O cd - P C D cd C D CD o p C D C O 4 3 cd C D 43 H T3 C D C O cd C D U H O C D • P C O p O HH o C D O +H C D P i • P bO P • l-H 43 CD HH C D 3 > 43 • g '£ a C D 43 o HH cd co C D cd o P X S o £ 3 P T3 ft < D ° " P C D cd « 43 3 - s C D C D C D 4 ? § ff.a 5 C O O co o cd ^3 bO 43 S3 ra '5b I D c •3 *" a is O CD U M s—> . p 43 P . •3H I X2 cd C N O CD C O O ON I uJ < UH C D 4 3 bO 4 3 cd C D > C O a O CD P 43 U H I D PH 43 t - 1 bO ^ 4 3 43 T3 bO C D 43 3 w C D 00 c p • C D cn c 43 CD < < U O c o c o U H C D C O ' — -C D > 0 3 ON U 4 ? CD > (D C D ^ _ > cd •3 ' o a oo U S3 P * -2 « r-J 4) P <; p C D HH cd PH cd C D o p C D H ID SPH 43 P CD O C O U PH 60 g '8 o a PH o o 00 • a o ca a o & H-< CD X> 9* IS CO a o T3 OJ p p '•4—» P o o "3 P P H . CD Pi 00 p '3 o -t—' o ,1 o o loo P 'fi o CJ ,1 CJ > i5 fr-l =3 l l l l to CJ cj 3 CJ CO 'F? ca - P co tp 13 g o p .2 1H r r* PH CJ V 1 CD Cd " J p ' CJ cd o CJ til -H» O cd u • rt P-. - a P . P cd 4 ^ ip cj .P O g "S -s O Cfl rrt 0 0 O oo o i3 'S cd R u, ° • P P H CJ co HP CJ H-» in o H ^ « CD P •c .2 c3 tJ cd .HP .5 ' w co cd i s 3 "3 S § T3 CJ co CO CJ CO PH PH £ H2 C<-H cd 3 HH CO ,CJ CD ^ P = « ' sap co "> J- 13 - P I 8.3 P H e e P H cd o c p l l fe p o CO H* £ CJ CO CJ T 3 ..M P cd CJ . - HP O CD " P p o co ' p CO P . £ O " I I lH O O : CJ o O cp N Q C(H P H_> cd R CD 13 £ CQ ffi o 2? O P H P . JH Si P H o ° Relationship between Attention and Social Functioning 111 Appendix B Limitations of Studies Examining Relationships between Cognition and Social Functioning in Individuals with Schizophrenia Authors Limitations Brenner, Hodel, Roder, & Corrigan (1992) • Did not explain how cognitive and social functions were measured or evaluated. • No statistical analysis was provided. Massel, Corrigan, Liberman, & Milan (1991) • Neglected to describe how the subjects were assessed for attention deficits. • Although the sets of conversational skills had been socially validated in previous research with individuals with schizophrenia, Massel et al. (1991) had not explained how deficits in these sets conversational skills related to impaired attention/information processing skills. • Did not provide a theoretical explanation as to how Attention Focusing Procedure (AFP) enhanced attention. • Did not provide procedural details of the Social Skills Training (SST), whereas they did with AFP. • Inappropriately performed A N O V A s on data from a single subject research design study. Wolery & Harris (1982) noted, "when serial dependency is present, conventional statistical procedures are considered inappropriate for analyzing the results of single-subject research" (p.448). • The lack of counterbalanced or alternating treatment designs prevented an unambiguous interpretation of the results due to possible interaction of SST followed by AFP. Kern, Green, & Satz(1992) • Did not provide operational definitions for task related behavior. • Did not discuss how inter-rater reliability was achieved. • Limited power of statistical tests due to small number of subjects. McKee, Hull, & Smith (1997) • Did not provide detailed descriptions or operational definitions of the behaviors assessed by the 'participation' rating scale. • Inter-rater reliability was not established. • Did not provide a rationale for selecting certain behaviors to define social functioning. • Did not explain the relationship between cognitive measures and social functioning behaviors. • Small sample size and lack of control group. • Did not prove validity & reliability of measure used to assess level of participation. • Study did not account for possible effects of antipsychotic medication. Relationship between Attention and Social Functioning 112 Appendix B continued Limitations of Studies Examining Relationships between Cognition and Social Functioning in Individuals with Schizophrenia Authors Limitations Dickerson, Boronow, Ringel &Parente(1996) • Neglected to describe how inter-rater reliability between evaluators of the Social Functioning Scale (SFS) was obtained and maintained. • Omitted psychometric information of the SFS such as reliability, validity, and sensitivity. Cornblatt & Keilp (1994) • Did not determine how many of the high risk children actually developed schizophrenia. • No objective evidence that the high-risk individuals actually exhibited behavior associated with social dysfunction. Freedman, Rock, Roberts, Cornblatt, & Erlenmeyer-Kimling(1998) • Did not provide sample questions from the Personality Disorder Examination (PDE). • Did not determine whether high-risk individuals actually developed schizophrenia. • PDE was intended for individuals with personality disorders not with psychosis. Penn, Van der Does, Spaulding, Garbin, Linzen, & Dignermans (1993) • Means Ends Problem Solving test (MEPS) may be a measure of imagination, rather than an indicator of social function. Penn, Mueser, Spaulding, Hope, & Reed (1995) • Global ratings of the unstructured role play performance lacked objectively defined terms. • Did not provide operational definitions or "anchors" of the behaviors associated with the specific components of social functioning. • Assessment of information processing did not include a direct measure of memory functioning. • Did not include assessment of extrapyramidal or medication side effects in mediating the relationship between cognition and behavior. Cramer, Bowen, & 0 ' N e i l . (1992) • No mention of how intra-rater reliability was obtained. • Did not explain how potential biases such as rater drift were controlled for. • Did not use a neuropsychological measure Corrigan, Green, &Toomey(1994) • Did not directly assess how impaired social cue recognition actually affected individuals' ability to express certain facial emotion and function socially. Relationship between Attention and Social Functioning 113 Appendix B continued Limitations of Studies Examining Relationships between Cognition and Social Functioning in Individuals with Schizophrenia Authors Limitations Addington, McCleary, & Munroe-Blum (1998) • Neglected to provide definitions of the social functioning domains associated with the Social Dysfunction Index (SDI) & Social Adjustment Scale II (SAS -IT) (indirect assessment). • Did not provide detailed descriptions of how subjects' role-played behaviors were assessed with the Assessment of Interpersonal Problem Solving Skills (AIPSS) (direct assessment). • Small sample size may have limited the power to detect statistical significance between neuropsychological scores and SDI & SAS-II scores. Addington & Addington (1999) • Did not describe how rater bias was controlled for during the administration of the SFS. • Did not offer a rationale for excluding the role-play portion of the AIPSS. • Cognitive deficits exhibited by their subjects may not have been severe enough to be assessed by the SFS. Harvey, Sukhodolsky, Parrella, White, & Davidson (1997) • Neglected to mention the protocols of how and when the raters observed subjects' behaviors. • The resulting 'global' social function score prevented identification of the aspects of social functioning more sensitive to cognitive impairments. • Did not explain what aspect of cognitive functioning was measured by each cognitive test. • Cognitive tests used were highly correlated with one another. Penny, Mueser, & North (1995) • The Allen Cognitive Level Test (ACT-90) was a "global" measure of cognition, and appeared to lack to the capability of differentiating which of the elements of cognition were most impaired. • Only 20% of the Social Interaction Test and ACL-90 measures were monitored for inter-rater reliability. Lysaker, Bell, Zito, & Bioty (1995) • Did not provide details of how the Work Personality Profile (WPP) was conducted. • Did not provide information to support reasoning for re-evaluation of subjects after 10 weeks. • Neuropsychological measures assessed global, not specific cognitive processes. • Terms of the WPP were not given operational definitions. Relationship between Attention and Social Functioning 114 Appendix C SPSS printout of Pearson correlation analyses Pearson correlation between subjects' age & SDI overall percent scores Case Processing Summary Cases Valid Missing Total N Percent N Percent N Percent AGE * SDI Overall % Score 35 100.0% 0 .0% 35 100.0% Symmetric Measures Value Asymp. Std. Error3 Approx. f 3 Approx. Sig. Interval by Interval Pearson's R -.068 .146 -.391 .698° Ordinal by Ordinal Spearman Correlation -.082 .175 -.470 .642° N of Valid Cases 35 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis, c Based on normal approximation. Pearson correlation between subjects' age & PASAT mean time/response scores Case Processing Summary Cases Valid Missing Total N Percent N Percent N Percent AGE * PASAT Mean Time/Response 35 100.0% 0 .0% 35 100.0% Symmetric Measures Value Asymp. Std. Error3 Approx. i° Approx. Sig. Interval by Interval Pearson's R Ordinal by Ordinal Spearman Correlation N of Valid Cases -.144 -.084 35 .099 .156 -.838 -.484 .408° .632° a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis, c Based on normal approximation. Relationship between Attention and Social Functioning Appendix C continued SPSS printout of Pearson correlation analyses Pearson correlation between subjects' level of education & SDI overall percent scores Case Processing Summary Cases Valid Missing Total N Percent N Percent N Percent Years of Education * SDI Overall % Score 35 100.0% 0 .0% 35 100.0% Symmetric Measures Value Asymp. Std. Error3 Approx. T5 Approx. Sig. Interval by Interval Pearson's R -.370 .167 -2.288 .029° Ordinal by Ordinal Spearman Correlation -.263 .172 -1.568 .126° N of Valid Cases 35 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis, c Based on normal approximation. Pearson correlation between subjects' education level and PASAT mean time/response scores Case Processing Summary Cases Valid Missing Total N Percent N Percent N Percent Years of Education * PASAT Mean Time/Response 35 100.0% 0 .0% 35 100.0% Symmetric Measures Value Asymp. Std. Error3 Approx. "F*3 Approx. Sig. Interval by Interval Pearson's R -.149 .196 -.864 .394 c Ordinal by Ordinal Spearman Correlation -.280 .171 -1.674 .104 c N of Valid Cases 35 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis, c Based on normal approximation. Relationship between Attention and Social Functioning 116 Appendix C continued SPSS printout of Pearson correlation analyses Pearson correlation between the number of months diagnosed and SDI overall percent scores Case Processing Summary Cases Valid Missing Total N Percent N Percent N Percent Months Diagnosed * SDI Overall % Score 35 100.0% 0 .0% 35 100.0% Symmetric Measures Value Asymp. Std. Error9 Approx. f 3 Approx. Sig. Interval by Interval Pearson's R -.057 .102 -.328 .745° Ordinal by Ordinal Spearman Correlation -.219 .177 -1.289 .206 c N of Valid Cases 35 a- Not assuming the null hypothesis. b. Using the asymptotic standard error assuming the null hypothesis, c Based on normal approximation. Pearson correlation between the number of months diagnosed and PASAT mean time/response scores Case Processing Summary Cases Valid Missing Total N Percent N Percent N Percent Months Diagnosed * PASAT Mean Time/Response 35 100.0% 0 .0% 35 100.0% Symmetric Measures Value Asymp. Std. Error9 Approx. f 5 Approx. Sig. Interval by Interval Pearson's R -.022 .096 -.125 .901° Ordinal by Ordinal Spearman Correlation .375 .151 2.324 .026° N of Valid Cases 35 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis, c Based on normal approximation. 00 e '3 o o a 3 PH o co B O C CO e <: a CO CO CO ft CO cd •3 p CD P H 3 CO CD CO j>N H-» CO CD <4H o H—> p O •c P H 00 00 PH 00 co CD t-O o 00 CD > o Q oo -p IH CD -o P CD o p CD CD H—» CD H—» CO CD H P CD T P P CD P H CD T 3 (A O CO HH rs H—' CO ft 3 o <5 o w g CO C O C O O O C O C O C O C O C D C O "tf CNJ C O C N C D C D Tj- C D C O i -co i*. C D C O T - O T - O C M C O co m C D i -C O C O un o CN T -Q) H (A 0) ft E re co H—' c CD •o c CD ft CU T3 c o CO 111 CO >_ CD ts-g ,CD 03 I - > CO M — Q) ° § .£ - • a 1 0) O 0 £ Q iu s • CD 2 co b CD O C C CO CO I I CO CO C D C O CO CN L O L O C O C M C O C M C O C 0 o C O o C O C D CM 0 0 O O "3-o L O I— O O L O C O to CD o c CO ' 1— co 3p; > 92 ho I n r3 5. co 5. H -I C T CO L T o LU CO UJ C oo 00 a '3 o c 3 PH O 00 •a a C 3 a o "3 CD p < e CD CD H3 CO a o CD Pi T3 CD P _ P -4—» P O o CO CD CO § -*H CO CD o -a p CD P H P O H—• P •c P H OO 00 PH 00 co CD lH o CD O O CD P O CH CO CD CD B H <; 00 < PH -a tH CD p CD a p CD CD CD HO H Cfl CD H P CD P CD PH CD -a (A U P W CD HH (0 Q . 3 o 0 o "J I •a 2 •4-' 00 00 ^ CJ) r^ - a> v - o in CD o cn iri i- 1 in a> i-~ CD •* o oo o CN CD CM CD iri m o CN T -CD s " CO o CD a. !00 g 0- I— in (A Q) a. E ra (0 HH c 0) •a c a> Q . O T 3 C o TO HI co £ g co "cu o p >l co ±i CD O a) S £ $ O 5 J) ss £ 5 So -E CD uu 2 • CD CO Q CD O c c CD CD b 00 OH CO 10 CD o c CO CO "5 > « 10 -= E •= io co 3 co ro 2. co 2. -~ i cr co CT o HI co HI c CD CO o CD Q. CD 60 a 'S o a PH o O oo « « _o a CD < a CD CD IS CO a o -4—> CM T3 CD c • l-H » c o o -3 c CD CH CH < co CD Vi >> H-» CO CD =3 O c •a CH 00 00 PH 00 CO CD )H o CD 00 >H CD > Q oo co O a a • l-H Q c CD CD CD HO CD H C CD T3 C CD CH CD (A O '•5 <£ •p re •4—' CO Q. 3 o (3 w g CO co to o c s* co CN CN h - CD i n T-co oo CD O CO N -CD r CN CD m co h - h -m CM 0 0 o d CO 0 0 co a> CM o c_> to CO (A a) (-in a> a E re to 4H c 0) T3 C a> Q. cu •D c o £\ TO Ol HI CO i . <u 2 g CO • £ 05 CO I- > co "cu CD • £ CD O c _ CD 5 5 § g HI ^ • CD 2 *= co b CD O C C CO CD to to CN CO 0 0 m m -*" i n ->ct-0 0 CD o oo CD 0 0 oo •<-: oo o CD CD CN i n o CD CD o 0 0 0 0 0 0 0 0 o 0 0 -cf o o CN CO CD O c re CO ^ | > g . . "CO 3 "CO CO ^ CO 3 H i D" CO CT o [ID CO 111 C CD O o to o CN 00 g ' 8 o o C 3 U H U o 00 - a a « -t—> a CD e < a CU CD 43 B-43 CO a o CD cd CD 3» S3 -J—» S3 o CD X -3 3! CD O H O H < CO CD "c3 CO CD CH-H O H-J O H-» c O H 00 00 PH 00 CO CD I H o o 00 CD co S3 o S3H co CD CD a < o o < PH 1 3 co ' c o O S> cd S3 CD CD CD 43 H-» CO CD H S3 CD T3 S3 CD a CD T3 4 3 U) o H-H (fl H3 re H H CO a. 3 o i_ C3 o • d 2 -4—' CO co o & CD a> o h- CD co CN CD •<-CO O T -00 CM CO 00 N-CO a> T -CM CO oo OO LO 00 CD CD CO CN o o CO CO CD ca o CD a. L ! s CO g Q. I— (A 0) I -CO fl» a E re CO H—< c a> "O c a> Q. a> T 3 c 1— CD o CD CO LO  CD LU CD 00 OO ffer T — o ffer CN CO Q CD al LU CD CD CD O T — T c LO LO CD CD CO CO CD | Differ CM CM 2 | Differ OO 00 (0 i_ CD O C5 H-H CD CO c <»£ CO H -"(U O c >-CD ±2 S 1 CD £ co O cz _ CO 21 c LO m CM o CD CO a> Q . O) i^ -O O D CO CM CO •tailed) oo CD CO CO cn In CO u. to CD O c CD ,5 5 CD 3 CD CD 3 CO 2. HH . CT CO L T o LU CD LU c CO CD o CD Q. co g Q- I— a '3 o c 3 o o 00 -a a a o c CD < a CD CD £ CD -O & H3 CO a o ed 13 CU 3t .53 "-4—» a o CD 13 S3 CD OH CO CD CO >> "3 s H CO CD 3 o H a •a OH 00 00 PH 00 CD l-H o CD 00 \ ° cd U H CD > o l - H Q oo r - H CD > CD CD 00 cd CO O Q 53 O • i-H • a o • l-H 13 (U 53 CD CD CD X) H CO CD H 1 H 53 CD 13 53 CD OH CD 13 •5 00 o H H CO H H to C O a. 3 O i_ o t c -o" 2 CO CO CO o Q c o "co o •o CD CM CO a> CM CD CO •rt CD CD CO CO CO T - h-CM o OO CO CO oo c-CM LO c\i co co co •D- oo CM CD CD a> OO 0 0 CO 1^ 0 0 CO oo CM CD CO I X in CD h-uo a> a. E to CO H H C CD T 3 C a> a. a> 73 c o & "co 51 l u co i_ CD £ g CD CO »- > CO K -"CD ° c CD CD £ CD - I 21 H ^ a _^ CO r^- o 0 co oo a_ o CM Q. o D CD t~~ LO •* <- CD LLI 2 • CD "O It CO Q CD o c c CO CD b CO CO LO a> CD co CO CM CM O OO LO 0 0 1^ CM CO 0 0 0 0 o o o 0 0 0 0 CM LO 0 0 1^ oo CO CO -ct X3 CD p -* -CO 3 CO CO 2. <o 2. • c r if) c r o I H I CO H I c CD 1 o CJ CO C N C N 60 g '3 o c 3 PH O OO T 3 c « a o P <: CD ft B O CD CH" T3 CU 3 P o o X • i-H P cu O H < CO cu co p CC5 H—» CO CU o H-» o -4—» P •a O H OO 00 PH 00 O o 00 CU co p o P H 4) £ cu H < 00 <: PH T3 "cu > CU H J CU M 03 CO O Q p _o H—» O '•3 CU p CU cu CD H P CD H p CD T3 P CU O H CU T3 cn u co a. 3 O CD o t c 2 5 55 CD > CD _ l <D. CO co o Q g H CO o CD CD CO co r*-<N CM •<- CO 0 0 T -CO CD in co oo CO CM CO •>- CD cn in OO CM CO o CM in T~ O 0 0 0 0 CM CD CD co CM CD CD is s CO o CD Q . CO CD CO g 0- h-0) (A a> a E (0 CO H c CD T 3 c CD a. CD •a c o CO 3 CT| HI co i_ CD o o -~ CO CO - c ,CD CO I - > 10 4— "CD ° s & CD O CD •O £ c l O g CDi ss - 5! us 2 CD 0 3 . CD I D fD • CD 2 co b CD o C C CO CD b co co CM OO in o in m o CD o CM CO CD m oo CD CD 1^ 0 0 CM o in in oo oo oo oo CD 0 0 0 0 oo CN CD CD CO CD O c CO CO " O | > g - CO I CO 3 CO CO 5. to 2. -•-CX CO CT O LD CO I D c CD to c o CL to CD 0£ IW g a. i -The Relationship between Attention and Social Functioning 123 Appendix E Sample Item of the Social Dysfunction Index Independent living (maintaining accommodation, shopping, meals, season-appropriate clothing, upkeep of clothing, finances, self-care other than covered under social presentation of self) Examples ofprobes/means of assessment: 'Describe your living arrangement?' 'What does it look like?' 'How much do you participate in cooking, cleaning, washing, shopping?' What is your source of income?' 'Do you have debts?' 'Do you run out of money before your next cheque comes in?' If yes, 'Did that happen in the past month?' ' Do you need staff/family to tell you what to do, help you get up, get places, etc.?' 'What is the atmosphere like in your home?' 'Enough space?' 'Privacy?' ' Are you satisfied with your living arrangement?' Example of serious social dysfunction: (i) accommodation is extremely overcrowded, disorganized, dirty and/or atmosphere is tense and restrictive; (ii) does not secure basics for self (household items, food) on regular basis; (iii) unsafe household procedures; (iv) no time management; (v) extreme difficulty managing finances. Note: Is not serious i f client lives in a setting where others do the basic tasks as long as everyone is satisfied with the arrangement. CN 3 o 13 G cu O H H H O — ' o 00 T3 B a c _o c co < a CD IO c o CL> ed 43 H co V -CD -*H cd p-l l-H Q 00 H. S3 CD • H C D CHH CD O U PJ o CD b o U C O C O "o cd )H C O 15 Is "CD PH: Alpha 0.9799 0.9788 1.000 0.9629 0.9235 1.000 0.9970 0.9531 0.9705 £P bo 0.0000 0000 0 0.0000 0.0000 0.0000 0.0000 0.0000 Degrees of Freedom •ri-ro tn ro CO T ro ro m ro F value 49.8235 47.1818 26.7692 13.0690 338.443 21.3030 33.9412 Upper Limit 0.9899 0.9893 1.000 0.9811 0.9614 1.000 0.9985 0.9763 0.9851 Lower Limit 0.9602 0.9580 1.000 0.9260 0.8484 1.000 0.9941 0.9070 0.9416 Confidence Interval m os 5? tr, os s? Os 5? in OS m OS 5? m OS 5? m OS 5? m OS 5? tr, OS Intraclass Correlation 0.9799 0.9788 1.000 0.9626 0.9235 1.000 0.9970 0.9531 0.9705 Dysfunction during past month Number of areas of dysfunction Satisfaction with level of functioning Dysfunction during past month Number of areas of dysfunction Satisfaction with level of functioning Dysfunction during past month Number of areas of dysfunction Satisfaction with level of functioning Public Self Independent Living Occupational Functioning cn CN g '3 o s PH o 00 T3 c c o c CU ti < C cu CD £ CD J O & &n C o CD CD H T3 CD C • i-H •+-» d o CD X '-3 d CD CH C/3 •H CD -a Q oo CD - i H CD MH CD O U c o • H -4—* _ctj "3 t! o U CO 0 0 r2 "o SH CO j>s "cS ' l-H "cD Alpha 0.9714 0.9375 1.000 0.9904 0.9527 0.9636 0.9758 0.9411 0.9428 M 00 0.0000 0.0000 0.0000 0.0000 0.0000 0000 0 0.0000 Degrees of Freedom CI PO ro CO f> ro co ro Ti-ro F value 49.8235 16.0000 104.1875 21.1471 27.4706 41.3429 16.9714 17.4848 Upper Limit 0.9856 0.9685 1.000 0.9952 0.9761 0.9816 0.9878 0.9703 0.9711 Lower Limit 0.9433 0.8762 1.000 0.9808 0.9063 0.9279 0.9521 0.8833 0.8867 Confidence Interval %S6 as s? OS IT, OS S? Os IT) OS 5? m Os 5? m OS S? m OS Intraclass Correlation 0.9714 0.9375 1.000 0.9904 0.9527 0.9636 0.9758 0.9411 0.9423 Dysfunction during past month Number of areas of dysfunction Satisfaction with level of functioning Dysfunction during past month Number of areas of dysfunction Satisfaction with level of functioning Dysfunction during past month Number of areas of dysfunction Satisfaction with level of functioning Family Relationships Important Relationships Community/ Leisure/ Recreation SO CN a •a o o 00 -a a cd e o CD e < e CD CD a. e o CD PH CD 43 H cu =3 a o o H H 5H •3 cu CO >H <D H-» cd I—I Q 00 >H c2 -4—» S3 cu • i-H O c+3 cu o U o <u b o O CO CO cd "o cd I H CO > N * U CM Alpha 0.9765 0.9763 0.9688 0.8686 0.8465 1.000 0.9929 0.9284 1.000 60 bo 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 Degrees of Freedom co CO CO co CO CO CO T CO CO F value 42.4857 42.1765 32.0588 7.6115 6.5163 141.5882 13.9706 Upper Limit 0.9881 0.9880 0.9843 0.9337 0.9225 1.000 0.9964 0.9639 1.000 Lower Limit 0.9534 0.9530 0.9382 0.7397 0.6960 1.000 0.9860 0.8582 1.000 Confidence Interval £ in OS <n OS in os 5? m Os -5 ox m OS m OS 5? in OS 5? m OS 5? m OS Intraclass Correlation 0.9765 0.9763 0.9688 0.8686 0.8465 1.000 0.9929 0.9284 1.000 Dysfunction during past month Number of areas of dysfunction Satisfaction with level of functioning Dysfunction during past month Number of areas of dysfunction Satisfaction with level of functioning Dysfunction during past month Number of areas of dysfunction Satisfaction with level of functioning Acceptance to Health Regiments Communication Insight/ Accuracy of Information The Relationship between Attention and Social Functioning Appendix G Cohen's Kappa Output for SDI Raters Public Self-dysfunction during month -PI * Public Self-dysfunction during month -IR Crosstabulation Count Public Self-dysfunction during month -IR Total .00 1.00 2.00 Public Self-dysfunction .00 13 13 during month -PI 1.00 19 19 2.00 1 2 3 Total 13 20 2 35 Symmetric Measures Value Asymp. Std. Error9 Approx. I*3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .948 35 .051 6.700 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. Public Self-areas of Dysfunction-PI * Public Self-areas of dysfunction-IR Crosstabulation Count Public Self-areas of dysfunction-IR Total .00 1.00 2.00 3.00 Public Self-areas .00 12 12 of Dysfunction-PI 1.00 14 2 16 2.00 6 6 3.00 1 1 Total 12 14 8 1 35 Symmetric Measures Value Asymp. Std. Error3 Approx. T*3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .913 35 .059 7.906 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning Appendix G continued Cohen's Kappa Output for SDI Raters Public Self-satisfaction with function-PI * Public Self-satisfaction with function-IR Crosstabulation Count Public Self-satisfaction with function-IR Total .00 1.00 Public Self-satisfaction .00 17 17 with function-PI 1.00 18 18 Total 17 18 35 Symmetric Measures Value Asymp. Std. Error3 Approx. i3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases 1.000 35 .000 5.916 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. Independent Living-dysfunction during month-PI * Independent Living-dysfunction during month-IR Crosstabulation Count Independent Living-dysfunction during month-IR .00 1.00 2.00 3.00 Total Independent .00 12 12 Living-dysfunction 1.00 1 15 1 17 during month-PI 2.00 1 4 5 3.00 1 1 Total 13 16 5 1 35 Symmetric Measures Value Asymp. Std. Error3 Approx. T6 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .864 35 .075 7.160 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning Appendix G continued Cohen's Kappa Output for SDI Raters Independent Living-areas of dysfunction-PI * Independent Living-areas of dysfunction-IR Crosstabulation Count Independent Living-areas of dysfunction-IR Total .00 1.00 2.00 Independent Living-areas .00 10 10 of dysfunction-PI 1.00 1 14 15 2.00 5 5 10 Total 11 19 5 35 Symmetric Measures Value Asymp. Std. Error3 Approx. f 3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .731 35 .098 6.189 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. Independent Living-satisfaction with function-PI * Independent Living-satisfaction with function-IR Crosstabulation Count Independent Living-satisfaction with function-IR .00 1.00 Total Independent Living-satisfaction with function-PI .00 1.00 30 5 30 5 Total 30 5 35 Symmetric Measures Value Asymp. Std. Error3 Approx. I*3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases 1.000 35 .000 5.916 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning Appendix G continued Cohen's Kappa Output for SDI Raters iccupational Functioning-dysfunction during month-PI * Occupational Functioning-dysfunctior during month-IR Crosstabulation Count Occupational Functioning-dysfunction during month-IR Total .00 1.00 2.00 3.00 Occupational .00 1 1 Functioning-dysfunction 1.00 1 11 12 during month-PI 2.00 3 12 15 3.00 2 5 7 Total 2 14 14 5 35 Symmetric Measures Value Asymp. Std. ErroP Approx. I*5 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .741 35 .096 6.484 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. Occupational Functioning-areas of dysfunction-PI * Occupational Functioning-areas of dysfunction-IR Crosstabulation Count Occupational Functioning-areas of dysfunction-IR Total .00 1.00 2.00 Occupational .00 1 1 Functioning-areas 1.00 1 20 21 of dysfunction-PI 2.00 1 12 13 Total 2 21 12 35 Symmetric Measures Value Asymp. Std. Error3 Approx. 1° Approx. Sig. Measure of Agreement Kappa N of Valid Cases .888 35 .076 5.931 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning Appendix G continued Cohen's Kappa Output for SDI Raters Occupational Functioning-satisfaction with function-PI * Occupational Functioning-satisfaction with function-IR Crosstabulation Count Occupational Functioning-satisfactio n with function-IR .00 1.00 Total Occupational .00 Functioning-satisfaction with function-PI 1 - 0 0 14 1 20 15 20 Total 14 21 35 Symmetric Measures Value Asymp. Std. Error3 Approx. T6 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .941 35 .058 5.578 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. :amily Relationships-dysfunction during month-PI * Family Relationships-dysfunction during month-IR Crosstabulation Count Family Relationships-dysfunction during month-IR .00 1.00 2.00 3.00 Total Family .00 18 18 Relationships-dysfunc 1.00 2 8 10 tion during month-PI 2.00 1 5 6 3.00 1 1 Total 20 9 5 1 35 Symmetric Measures Value Asymp. Std. Error3 Approx. T 5 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .859 35 .077 7.271 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning Appendix G continued Cohen's Kappa Output for SDI Raters Family Relationships-areas of dysfunction-PI * Family Relationships-areas of dysfunction-IR Crosstabulation Count Family Relationships-areas of dysfunction-IR Total .00 1.00 2.00 Family .00 17 17 Relationships-areas 1.00 2 14 16 of dysfunction-PI 2.00 1 1 2 Total 19 15 1 35 Symmetric Measures Value Asymp. Std. Error3 Approx. T^  Approx. Sig. Measure of Agreement Kappa N of Valid Cases .841 35 .086 5.600 .000 a- Not assuming the null hypothesis. b. Using the asymptotic standard error assuming the null hypothesis. Family Relationships-satisfaction with function-PI * Family Relationships-satisfaction with function-IR Crosstabulation Count Family Relationships-satisfact ion with function-IR .00 1.00 Total Family Relationships-satisfa ction with function-PI Total .00 1.00 24 24 11 11 24 11 35 Symmetric Measures Value Asymp. Std. Error3 Approx. I*3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases 1.000 35 .000 5.916 .000 a- Not assuming the null hypothesis. b. Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning Appendix G continued Cohen's Kappa Output for SDI Raters Important Relationships-dysfunction during month-PI * Important Relationships-dysfunction during month-IR Crosstabulation Count Important Relationships-dysfunction during month-IR .00 1.00 2.00 3.00 Total Important .00 Relationships-dysfunc 1 QO tion during month-PI 2 00 3.00 Total 14 1 15 9 9 1 1 2 8 8 14 10 1 9 34 Symmetric Measures Value Asymp. Std. Error9 Approx. T*3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .913 34 .059 8.084 .000 3- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. Important Relationships-areas of dysfunction-PI * Important Relationships-areas of dysfunction-IR Crosstabulation Count Important Relationships-areas of dysfunction-IR Total .00 1.00 2.00 Important .00 12 1 13 Relationships-areas 1.00 1 11 12 of dysfunction-PI 2.00 2 8 10 Total 13 14 8 35 Symmetric Measures Value Asymp. Std. Error3 Approx. I*3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .827 35 .081 6.904 .000 a- Not assuming the null hypothesis. b. Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning 134 Appendix G continued Cohen's Kappa Output for SDI Raters Important Relationships-satisfaction with function-PI * Important Relationships-satisfaction with function-IR Crosstabulation Count Important Relationships-satisfact ion with function-IR Total .00 1.00 Important .00 9 9 Relationships-satisfa 1.00 ction with function-PI 1 25 26 Total 10 25 35 Symmetric Measures Value Asymp. Std. Error3 Approx. I*3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .928 35 .071 5.503 .000 a- Not assuming the null hypothesis. b. Using the asymptotic standard error assuming the null hypothesis. Community/Leisure/Recreation-dysfunction during month-PI * Community/Leisure/Recreation-dysfunction during month-IR Crosstabulation Count Community/Leisure/Recreation-dysfunction during month-IR Total .00 1.00 2.00 3.00 Community/Leisure/ .00 6 6 Recreation-dysfuncti 1.00 1 14 1 16 on during month-PI 2.00 12 12 3.00 1 1 Total 7 14 13 1 35 Symmetric Measures Value Asymp. Std. Error3 Approx. f 3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .913 35 .060 7.826 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning Appendix G continued Cohen's Kappa Output for SDI Raters Community/Leisure/Recreation-areas of dysfunction-PI * Community/Leisure/Recreation-areas of dysfunction-IR Crosstabulation Count Community/Leisure/Recreation-areas of dysfunction-IR Total .00 1.00 2.00 3.00 Community/Leisur .00 7 7 e/Recreation-area 1.00 1 18 2 21 s of dysfunction-PI 2.00 1 5 6 3.00 1 1 Total 8 19 7 1 35 Symmetric Measures Value Asymp. Std. Error3 Approx. f 3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .807 35 .091 6.933 .000 a- Not assuming the null hypothesis. D- Using the asymptotic standard error assuming the null hypothesis. Community/Leisure/Recreation-satisfaction with function-PI * Community/Leisure/Recreation-satisfaction with function-IR Crosstabulation Count Community/Leisure/Re creation-satisfaction with function-IR Total .00 1.00 Community/Leisure/ .00 17 2 19 Recreation-satisfact 1.00 ion with function-PI 16 16 Total 17 18 35 Symmetric Measures Value Asymp. Std. Error3 Approx. I*3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .886 35 .078 5.276 .000 a- Not assuming the null hypothesis. D- Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning Appendix G continued Cohen's Kappa Output for SDI Raters Acceptance & Adherance to Health Regimes-dysfunction during month-PI * Acceptance & Adherence to Health Regimes-dysfunction during month-IR Crosstabulation Count Acceptance & Adherence to Health Regimes-dysfunction during month-IR .00 1.00 2.00 Total Acceptance & Adherance to Health .00 1.00 2.00 24 1 1 6 24 Regimes-dysfunction during month-PI 3 5 6 Total 25 3 7 35 Symmetric Measures Value Asymp. Std. Error3 Approx. "I*3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .877 35 .081 6.865 .000 a. Not assuming the null hypothesis. b. Using the asymptotic standard error assuming the null hypothesis. Acceptance & Adherance to Health Regimes-areas of dysfunction-PI * Acceptance & Adherence to Health Regimes-areas of dysfunction-IR Crosstabulation Count Acceptance & Adherence to Health Regimes-areas of dysfunction-IR Total .00 1.00 2.00 Acceptance & Adherance .00 21 21 to Health Regimes-areas 1.00 1 12 13 of dysfunction-PI 2.00 1 1 Total 22 12 1 35 Symmetric Measures Value Asymp. Std. Error3 Approx. T 5 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .942 35 .057 6.072 .000 a. Not assuming the null hypothesis. b. Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning Appendix G continued Cohen's Kappa Output for SDI Raters Acceptance & Adherence to Health Regimes-satisfaction with function-PI * Acceptance & Adherence to Health Regimes-satisfaction with function-IR Crosstabulation Count Acceptance & Adherence to Health Regimes-satisfaction with function-IR .00 1.00 Total Acceptance & .00 Adherence to Health Regimes-satisfaction 1 .oo with function-PI Total 12 1 13 22 22 12 23 35 Symmetric Measures Value Asymp. Std. Error3 Approx. T 5 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .938 35 .061 5.559 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. Communication-dysfunction during month-PI * Communication-dysfunction during month-IR Crosstabulation Count Communication-dysfunction during month-IR Total .00 1.00 2.00 Communication-dysfun .00 6 2 8 ction during month-PI 1.00 3 11 2 16 2.00 2 9 11 Total 9 15 11 35 Symmetric Measures Value Asymp. Std. Error3 Approx. i1 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .602 35 .116 4.977 .000 a- Not assuming the null hypothesis. b. Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning Appendix G continued Cohen's Kappa Output for SDI Raters Communication-areas of dysfunction-PI * Communications-areas of dysfunction-IR Crosstabulation Count Communications-areas of dysfunction-IR Total .00 1.00 2.00 Communication-areas .00 6 2 8 of dysfunction-PI 1.00 3 14 4 21 2.00 6 6 Total 9 16 10 35 Symmetric Measures Value Asymp. Std. Error3 Approx. f 5 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .584 35 .119 4.935 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. Communication-satisfaction with function-PI * Communication-satisfaction with function-IR Crosstabulation Count Communication-satisf action with function-IR Total .00 1.00 Communication-satisf .00 21 21 action with function-PI 1.00 14 14 Total 21 14 35 Symmetric Measures Value Asymp. Std. Error Approx. "T Approx. Sig. Measure of Agreemer Kappa N of Valid Cases 1.000 35 .000 5.916 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning Appendix G continued Cohen's Kappa Output for SDI Raters Insight/Accuracy of Information/Expectations-dysfunction during month-PI * Insight/Accuracy or Information/Expectations-dysfunction during month-IR Crosstabulation Count Insight/Accuracy or I nformation/Expectations-dysfunction during month-IR Total .00 1.00 2.00 3.00 Insight/Accuracy of .00 3 3 Information/Expect 1.00 11 1 12 8 ations-dysfunction 2.00 8 during month-PI 3.00 12 12 Total 3 11 9 12 35 Symmetric Measures Value Asymp. Std. Error3 Approx. "T*3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .960 35 .040 9.107 .000 a- Not assuming the null hypothesis. b. Using the asymptotic standard error assuming the null hypothesis. Insight/Accuracy of Information/Expectation-areas of dysfunction-PI * Insight/Accuracy of Information/Expectation-areas of dysfunction-IR Crosstabulation Count Insight/Accuracy of Information/Expectation-areas of dysfunction-IR Total .00 1.00 2.00 Insight/Accuracy of .00 3 3 Information/Expectation- -| go 12 1 13 areas of dysfunction-PI 2 00 3 16 19 Total 3 15 17 35 Symmetric Measures Value Asymp. Std. Error3 Approx. I*3 Approx. Sig. Measure of Agreement Kappa N of Valid Cases .799 35 .096 5.806 .000 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning 140 Appendix G continued Cohen's Kappa Output for SDI Raters Insight/Accuracy of Information/Expectations-Satisfaction with function-PI * Insight/Accuracy of Information/Expectations-satisfaction with function-IR Crosstabulation Count Insight/Accuracy of .00 Information/Expect ations-satisfaction 1.00 with function-PI Total Insight/Accuracy of I nformation/Expectatio ns-satisfaction with function-IR .00 21 21 1.00 14 14 Total 21 14 35 Symmetric Measures Value Asymp. Std. ErroP Approx. f* Approx. Sig. Measure of Agreement Kappa N of Valid Cases 1.000 35 .000 5.916 .000 a- Not assuming the null hypothesis. b. Using the asymptotic standard error assuming the null hypothesis. The Relationship between Attention and Social Functioning 141 Appendix H SPSS printout of Pearson correlation analysis of the relationship between SDI overall percent and PASAT mean time/response scores Case Processing Summary Cases Valid Missinq Total N Percent N Percent N Percent SDI Overall % Score * PASAT Mean Time/Response 35 100.0% 0 . 0 % 35 100.0% Symmetric Measures Value Asymp. Std. Error3 Approx. i° Approx. Sig. Interval by Interval Pearson's R .086 .197 .495 .624° Ordinal by Ordinal Spearman Correlation .025 .187 .142 .888 c N of Valid Cases 35 a- Not assuming the null hypothesis. b- Using the asymptotic standard error assuming the null hypothesis. c- Based on normal approximation. The Relationship between Attention and Social Functioning 142 Appendix I SPSS printout of Pearson correlation analysis of the relationship between the total percent occupational functioning subcomponent scores of the SDI and PASAT mean time/response scores Case Processing Summary Cases Valid Missing Total N Percent N Percent N Percent Total Occupational Functioning subscale * PASAT Mean Time/Response 35 100.0% 0 . 0 % 35 100.0% Symmetric Measures Value Asymp. Std. Error9 Approx. T*3 Approx. Sig. Interval by Interval Pearson's R .204 .110 1.197 .240 c Ordinal by Ordinal Spearman Correlation .200 .171 1.172 .250 c N of Valid Cases 35 • a. Not assuming the null hypothesis. b. Using the asymptotic standard error assuming the null hypothesis. c- Based on normal approximation. -3-I o a 3 P H o o 00 T3 a CO c o < a CU 40 a, 43 w a o CU Pi CU 43 H • r H p CD C H -a o l-H cd CD CO CD HH CO =3 O "> CD tH CM _ P OX) P 'S 'H—» CD p o o co P O 'H (3 CD P CD CD CD HO co OH 43 CO p O CH O 43 to p CD tH H—» 00 p o 'M "CD o U C+H o 43 H-> 00 P CD HH + H 00 cd CD 00 p 'H-» o S3 P PH •1H ' C D O 00 c o "5 p £ .2 CO Jj 'P C HH <D s s - P cd < o H-» • l H CD -S CD P C H £ co +3 I 2 P co CD cd MH O CD H-> 4P O C CD I f £ CO CD HP CD & CD CO CD Pi -P 5 co cd II g HH -S P o 'H—» CD P cS CO Q Q 00 X CD T 3 H P P cd o <HH ;|I CO p p CD r — • s H P < ^ 0 0 cd CD 'o Id O o 00 oo co ,—I =3 o M O w V 00 v H _ C P H ' ' c o <—T CO <~1 CD m 8 © a . CO 4*i CO 00| P -3 P CD co '-O 'fO P ! H P o " P CD CD C > co CD CD £ ^  £ 00 co cd co H—• P CD cd CD f ° cd 'S CD JH C H o N O § ) 4 p —, cd .-p 4P — 0 0 o °^  <3 "S 3 § CO CD O ^ O S)| ii "a HH .5 00 p P o cd CD 00 00 IT") o ^ o V C H CO " 001 II <•-> " CO HH • CD > I—H cd P CO • 1—I > °3 T3 CD J> P P > O co CD CD 00 co cd t—1 u o" §>l 00 PH 00 Ml P '> in rp CO CO 00 P o o CD i CN HH o ^ ,—; co 4^  CO 4*1 "2 CO P H „ CD OO co P "—' • —i CO *—1 co O o o o 1 a . cd o • i H H-» CO a CO Q J—I • 1-- H * CO "cd .S «« T 3 CD CD .5 W CD .5 « 00 co P ^ ,2 ^ S < CD P H 43 00 CD ' P fe 4ai cd O H P - cd oo < H C+H o CO H-» p CD cd CD H ^ S § a o .2 cd P CD J3 a o 4P • C J . H s CD 0 0 T3 ? co P O H-» 00 p P O to ON ON < : < c "3-g '3 o o a PH o T3 c a c o H—> a < a <u ID 43 a, o O H U 43 H T3 CD P .S3 P o o p CD OH OH < 4 3 o CD CO CD VH co P O "> CD HH OH 00 P ' 3 'H—> CD 3 a • r H o o co p o 'H—» p CD P CD CD CD 43 CO OH P o "CD HH <HH o 43 H 60 (3 CD HH H—» 0 0 S3 o ID fa o U <H-H o 43 -4—» 00 S3 CD t H H—» 0 0 cd CD OO S3 o o 13 P IX, "c3 • >H o o o o 13 _ o "cd S3 - . 2 co H ' p s .5 CD 8 3 43 cd < O CD +3 CD S3 PH £ CO H . 2 2 • 2 1 3 P CO CD cd <+H O h CD H-» HO CD 5 'x? >7 3 AH CO HH CD 4 3 o cd1 CD co CD PH. cd "P ^ cd o CD O co O O CO CO CO > O CO P OH r H CD ^ P CO O 4P o ii 8 "cd ^ eg . 2 -s ' o o O ^ HH . S HH .1 S 3 » o £ co PH Q ^ >^ 2 .•P. - i 3 P 3 cd cd a 5 3 - p e I CD X PH m T3 cd 3 _ H co > cd O <3 P • H P o cd CD - g H - * CD P co 45 CD P CD * i OO 0 0 cd CD o 0 0 O CHH . . cd co O C+H J? • H 3 CHH ° -^ 00 CHH 00 co P ^ co Q P c d . t i r j " co  V N co 1—1 o 2 ^ ^ - H § p , ^ _ y 2 c W i a ' p O X O H - H ^ H— J H „ HH P ' T D P P H P H P H P r r t H t r ^ r s H - ^3 < 43 o b o o CD S cd s j j , . JS cd 42 cd ' P cd 3 § 3 • O 2H T 3 <+-< J3 - r <4H .EH O > O T3 4^ O T3 ^ -p co N p co CD — -cl £ — -cl c £ 9 s O * H o HH i t j > CD cd 5 2 S3 o m c d c r o c d c d ' P o o o cd P •a •a p CD cd CN ON 40 ON 6 w U Id II * © g V CD q OH > ° CN X J ^ / O r - l ON co •H—» Cd CO o ir> "+H CN -P ZH C HH • H CO CD H p CD O .2 o CD CD o u OO P< CD CD P > CO CD CD C#0 co cd PH 00 T3 CD CD '3 O co CD r 5 » 0 0 co H PH U P CD • i-H . 3 4 P so ' J CN > cd . * <H M-H r H 0 CD 1 & cd 43 T3 co CD p cd O N 00 O N O _ : U Id cd o cd CO OH CD H P CO CO HH CD co o , HH CHH Cd 53 CDn CD TD »*H P o <H JH p " - a ' S o § « o '55 rX ° b P CD c_; o P H . P t3 o o PH 00 o P - T3 •> 00 CD ^ CD" . 3 3, p . S t 3 > . 2 i S O B ^ P CO CD CD T3 CD 5 u p c s 00 co cd cd cd 00 • l-H Q o o H > PH| i> 00 p< 0 0 < • c co .2 g § •s a O H . 2 o - 5 o o - r o o > cd "8 CD J3 OH O N c^ '-8 C+H CD O co CD P ^ O s o 2 °^  CD ON o w Q 15 a o P H o O C/D T 3 a cd C o B CD a < c CD CD CD -S a o CD ed CD 43 H CD . g '-4—* a o o c CD CM CM < 4 3 O CD CD 35 O "> CD !-< CM 3! • i-H (3 O • i-H -4—> o I ' o o CO X) S3 o • l-H -4—» $3 CD -4—» H—» cd 3! CD CD CD 43 co CM 43 CO S3 O • T—( -4—» _cd 13 C+H o cd) 31 CD M +-» 0 0 c o a3 tl o U + H o 4 3 +-» c CD IH H 0 0 CD 13 O 0 0 S3 o H—> o S3 3i PH ' o O 0 0 S3 _o "cd S3 & .2 CO J J a CD 3*8 C+H o CD 13 CD S3 OH ^ I 2 •2 * S3 co CD cd £ J CD < 2 CD +H o O (2H S CD 4 3 o c3 CD CO CD p-1 o cd C+M o cn CD 3" C+H O o O || co t d VO" © C+H C O . o B © (D CD • i-H CM , CO 1 3 Q c a ^ P H Q P H co T 3 > id <8 S3 .12 o +H (!) S3 co 4 H CD 3 D tj 0 0 0 0 cd CD H 3 'CO . C  <+H «-> 2 0 0 ° co CD a C+H co £ g g O g m ~ * - CD .2 . - •§ >H CO H -3 ° Jr! P H o OJ > CD < Q . CO Cd +-> +H r9> C+H '5b to O 00 O -J3 CD co S3 O- co Q £ CD O ^ CD i 3 g M So S H co r j > U > < H > 0 0 -2 cd J2 cd 3 1 " O C T3 H (H ^ H (H > o - d ^ O - 0 ^4 T 3 S 3 C O N S 3 C O C D > H + --a "g — 1 'C g m (3 C+H - H l -H O < > + H * o •£> +H M*T T c d c o T r c d c d T 3 ' - H > *3 S3 cd oo f3 as •3 °^  ID O CD )H _ ; PH cd CN o © V OH IT) © S co O cd CO " ! H cr CD J? ^ 3 3 CD 3 a? 0 0 a cd 31 co • r H °3 "S £ S3 o .£ o H fll SH co T - 3 CD 33 CD ± 3 0 0 OO cd H P H O O o o V OH o" OO c^ 4 3 CD 4 3 T 3 CD "cd ^ CD CO cd CD H - r 3 o -4—> 3! 1 3 CD S3 cd HH CO 33 0 0 - T 3 M CD CD . > t d o 6 CD CD co cd T 3 UD I § cd | S3 • — c3 HO C H 'HS • i -H ^3 • f-H • i-H fJJ Q Q H cd -3 .2 S3 § o OH ' CD I H N CD O 43 43 2 T 3 i-H |> co m CN C H T 3 a> o s SO 00 g 'S o PH O GO "O a cd a o < a u u P H CO a o CD PH CD 4= H CO P • I-H -4—> P o o X '•3 a CD 4 3 CD CD co P O '> O IH OH c w p • i-H p o • r H H CD (3 cS • IH o o CO G O » p CD H 13 S3 CD CD £ H CD 4 3 CO PH • i-H 4 3 CO S3 o 'I IH C+H o 00 S3 CD IH +H 00 S3 o H +H j d a3 C o U +H o 4 3 tb S3 CD rH +-» 00 cd o 00 S3 _o '+-» o S3 P PH 13 • r H CD o 00 S3 o 't3 CD CO S3 < o C+H O CD 35 CD S3 a «^  co +3 I 2 •3 ^ S3 co CD cd C+H O Sri co CD + H rO CD p u rZ CO CD 4 3 CD c3 CD CO CD o © V PH O O II ca CO •s 9-CD O g 5b •S P CD S3 < f H cd 31 co > 'd -CD S3 S3 O 'cd +3 +-» S3 co CD 3i t i 00 cd co S3 " O O O V O H cn © 0 2 . r ^ l CO S3 to •5 I1 •2 S i 13 Q OH M "3 .S +3 S3 1 •§ PH +3 co H CD ' S f l C - 1 H o b Q CD i P £ - :- ^ P. M O 3 3 00 co OD -e r2 f t « o P. o ^ CD 3S co ±3 CD O P O ^ cd < -3 .£ o r - H g oo 13 I H o ^ + H O co 2 o O O CO P ° .2 S) 13 .2 P H T 3 .S 4 3 cd '2 CD J3 P H O N as • £ <HH i - H 3> O co CD CD Os CD OS o S Id o © V PH cn CO 4 3 2 * HH CO CO T3 CD P H V oo p s c3 > 3 o 2 00 CD J> P > H3 O g co CD cD . S u e 00 co cd 0 CD P H cd • H CO P 13 CD P PH cd Q PH CD _> ' + H CD C+H <+H IH cd CD O T3 o J 3 .E^rO .S . ^ 4 3 .S g 8 i> co i - H (5 to "O c i cd cn as as cd rO o "Hb W CD in o O P o £ PH £ 00" 8 CN rt o .2 cd co O CD o a r2 " P H CD 'o s s "<=> cd cd OD -P f H IH CD 5 p rO O + H , 1 co y cd O f l g • r H r H ^ CD .3! M 0 13 c 00 2 P cd P CO > CD CD P P > O OT CD <u 00 co cd PH C+H 00 O r4 cd O S u H PH U cd P en OT +-t CD H en cd PH PH O cd _ p 4 3 H-* N f in O C O CO CD _ > 'H—» O C+H <+H „ Cd CD O T3 * -P 3 .-P rP OT OT T3 cd +-» CD ^ r H S Os PH W 00 g '3 o UH o GO C CO C o +3 < a CD £ CD a. co 3 o CD P i CD JS H 1 3 S3 CD C H O 2 + - » CO a o CD CD _ > + H o o o IH 3f CD S3 > . CD C+H O CO '53 X "3! a3 S3 03 I cd +-» CD O o CN CD S3 CD CD IH o C+H o & S B 3 00 p value <-U.UUUUUl <0.000001 <0.000001 O.OOOOOl Effect size Medium Medium-large Small-medium Small-medium Standard error 0.039 0.077 0.033 0.040 Pooled estimate r 0.29 0.40 0.23 0.20 Total sample size r-fN r-00 00 r H 1002 CN OO V O Box Scores Number of studies Total 00 l - H <N < S r--r H Box Scores Number of studies Null results in i—l l - H 00 Box Scores Number of studies Positive results m V) i - H C\ Functional Outcome Correlates Two to three studies Social problem solving Psychosocial skill acquisition Psychosocial skill acquisition Functional Outcome Correlates Four or more studies Community/ daily activities Social problem solving Psychosocial skill acquisition Psychosocial skill acquisition Community/ daily activities Social problem solving Neurocognition Secondary verbal memory Immediate verbal memory Card sorting Vigilance The Relationship between Attention and Social Functioning 148 Appendix L Review of frequently cited measures of attention Auditory Based Test The Digit Span subtest of the WAIS-R is often used as a test of verbal immediate memory. The test has two general sections, both consisting of seven pairs of random sequences of numbers. In the Digits Forward segment, the examiner reads aloud number sequences that are from three to nine digits long, and the patient must repeat each segment exactly as it is heard. The Digits Backward portion of the subtest operates in much the same fashion, the major difference being that the patient must say the digits read by the examiner in reverse order. Each section of the Digit Span discontinues when the patient fails to repeat both sequence of a pair of equal length (Berg, Franzen, & Wedding, 1987). The Arithmetic subtest of the WAIS-R consists of 14 items. The simplest item calls for block counting. Item range form simple to more difficult arithmetic problems. Some of the items pose more conceptually difficult problems. For example an item may read, "Four men can finish a job in eight hours. How many men will be needed to finish it in a half hour?" Arithmetic items have time limits ranging from 15 seconds on the first four to 120 seconds on the fourteenth. A subject can earn raw score bonus points for particularly rapid responses on the last four items (Lezak, 1977). The Relationship between Attention and Social Functioning 149 Appendix L continued Review of frequently cited measures of attention The Digit Span Distractibility Test is an audiotape measure of selective attention that requires the subject to attend selectively to a female voice and ignore a male voice. In the neutral condition, a female voice says 6 different digits at a rate of 1 per 2 seconds, and the subject is instructed to write down the digits in correct order after the tape recorder is stopped. In the distracting condition, the female voice says 5 different digits at the rate of 1 per 2 seconds. However, in this condition, a male voice says 4 digits during each of the intervals between digits said by the female voice. The subject is instructed to ignore the male voice, and write down the correct order of digits said by the female voice (Kern et al., 1992). The Continuous Performance Test (CPT), auditory version requires the individual to listen to a random series of spoken letters presented at one-second intervals over 3 minutes. The stimuli are presented by taped recordings of 300 letters include 30 target letters. The subject presses a key each time he or she hears the letter ' A ' followed by the letter ' X ' . The final score is the number of correct responses out of a possible 45 (Addington et al. 1998). Visually Based Tests The WISC-II Mazes is a timed test of attention and planning. Subjects must trace a maze without crossing lines or going into blind alleys (McKee et al., 1997). The Relationship between Attention and Social Functioning 150 Appendix L continued Review of frequently cited measures of attention The CPT (visual version) is a measure of visual sustained attention. This involves monitoring a quasi-random series of stimuli (numbers) as they are presented briefly one at a time, in a continuous sequence, and pressing a response button each time that a predesignated stimulus occurs (Addington & Addington, 1998). The SPAN measures the efficiency of early iconic memory and read-out stages of visual information processing relatively independently of active short-term memory. The measures presents arrays of 3 or 12 letters that contain either a 'T ' or an 'F ' . Responses are scored to obtain the number of correct detections per array size (Addington & Addington, 1998). The Stroop Test requires the individual to state the color of the ink that words are written in. The words themselves are color names that are often different to the ink color in which they are written. For example, the word "blue" may be written in red ink (Berg, Franzen, & Wedding, 1987). The Symbol Digit Modalites Test is a timed visual search and scanning test requiring sustained, focused attention and directed visual shifting. Subjects substitute numbers for symbols using a key printed at the top of the page (McKee et al., 1997). 

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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0090139/manifest

Comment

Related Items