Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
An examination of affective modulation, psychopathy, and negative schizotypy in college and community samples
(USC Thesis Other)
An examination of affective modulation, psychopathy, and negative schizotypy in college and community samples
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back o f the book. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. UMI A Bell & Howell Information Company 300 North Zeeb Road, Ann Aibor MI 48106-1346 USA 313/761-4700 800/521-0600 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. AN EXAMINATION OF AFFECTIVE MODULATION, PSYCHOPATHY, AND NEGATIVE SCHIZOTYPY IN COLLEGE AND COMMUNITY SAMPLES by Veronica Mejia A Thesis Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree MASTER OF ARTS (Psychology) December 1997 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 139 0004 UMI Microform 1390004 Copyright 1998, by UMI Company. AH rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. UMI 300 North Zeeb Road Ann Arbor, MI 48103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY O F SO UTHERN CALIFORNIA T H E G R A D U A T E SCHOOL U N IV E R S IT Y PARK L O S A N G E L E S . C A LIFO R N IA 9 0 0 0 7 This thesis, w ritten by V e ro n ic a Y v e tte M ejia under the direction of A . ® ?. Thesis Committee, and approved by all its members, has been p r e sented to and ac c ep te d by the Dean of The Graduate School, in partial fulfillment of the requirements fo r the degree of M a s te r o f A rts Dean r 2 5, 1997 THESIS C O M M IT T E E - C hairm an ' \ ' V O ----- L y 's Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGMENTS I would like to express my appreciation to my advisor, Dr. Michael E. Dawson for all his advice and support throughout the execution of this project. I would also like to thank Dr Anne M. Schell for her helpful suggestions throughout the execution and writing of this project. I gratefully acknowledge the generosity of Dr. Adrian Raine and especially his assistant Lori Lacasse for all their time and helpful input regarding the use of the Psychopathy Checklist and Schizotypal Personality Questionnaire data. I am also deeply thankful to Dr. Rand Wilcox for re-introducing me to the world of statistical analysis and for all his help and guidance through the data analysis phase of this project. To Bill Troyer, I would extend my thanks for all the technical support he has provided and especially for answering all of my naive questions regarding equipment. Additionally, I would like to thank my fellow students Serkan Oray, Kim Seljos, and especially Jonathan Wynn for all their constructive criticism, input, and general support throughout the writing of this project. Finally, I would like to extend my deepest gratitude to Dr. Eric Vanman for the initial development of this project, for providing a scoring program for the startle eyeblink data, but most of all, for introducing me to the study of psychophysiology and emotion and for always encouraging me to push myself further. This research was supported by NIMH grant MH40496 awarded to Dr. Michael E. Dawson and by NIMH grant MH50940 awarded to Dr. Adrian Raine. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Abstract This study attempted to clarify the nature of emotional deficiencies experienced by schizotypals and psychopaths through the examination of their affective startle modification patterns. Subjects in Experiment 1 and Experiment 2 were classified by the Schizotypal Personality Questionnaire (SPQ) Negative Symptom factor scores. In Experiment 2 only, subjects were also classified by Hare Psychopathy Checklist (PCL- Screening Version) scores. Present results replicated Patrick, Bradley, and Lang’s (1993) finding that only subjects high in both factors of the PCL (measuring emotional detachment and antisocial behavior) show reduced emotional modification of startle. Subjects scoring high on the SPQ Negative Symptom factor displayed enhanced affective startle modification compared to those scoring low. These startle patterns appear to illustrate the difference between a deficient affective state (suffered by psychopaths) and deficient affective expression (possibly suffered by schizotypals). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Tables 1. Experiment 1 Trimmed Means and Trimmed Standard Errors for Percent Change in Startle Modulation 2. Experiment 1 Trimmed Means and Trimmed Standard Errors for Skin Conductance Orienting Response Amplitudes 3. Experiment 1 Means and Standard Errors for Percent Change in Startle Modulation - Raw Analyses 4. Experiment 1 Means and Standard Errors for Skin Conductance Orienting Response Amplitudes- Raw Analyses 5. Experiment 1 Means and Standard Errors for Percent Change in Startle Modulation- Outlier-Identified and Replaced Analyses 6. Experiment 2 PCL-defined Group Trimmed Means and Trimmed Standard Errors for Percent Change in Startle Modulation 7. Experiment 2 PCL-defined Group Trimmed Means and Trimmed Standard Errors for Skin Conductance Orienting Response Amplitudes 8. Experiment 2 SPQ-defined Group Trimmed Means and Trimmed Standard Errors for Percent Change in Startle Modulation 9. Experiment 2 SPQ-defined Group Trimmed Means and Trimmed Standard Errors for Skin Conductance Orienting Response Amplitudes 10. Experiment 2 PCL-defined Group Means and Standard Errors for Percent Change in Startle Modulation - Raw Analyses 11. Experiment 2 PCL-defined Group Means and Standard Errors for Skin Conductance Orienting Response Amplitudes - Raw Analyses 12. Experiment 2 SPQ-defined Group Means and Standard Errors for Percent Change in Startle Modulation - Raw Analyses 13. Experiment 2 SPQ-defined Group Means and Standard Errors for Skin Conductance Orienting Response Amplitudes - Raw Analyses 14. Experiment 2 PCL-defined Group Means and Standard Errors for Percent Change in Startle Modulation - Outlier-Identified-and-Replaced Analyses 15. Experiment 2 SPQ-defined Group Means and Standard Errors for Percent Change in Startle Modulation - Outlier-Identified-and-RepIaced Analyses page 20 20 24 24 25 36 36 37 37 43 43 44 44 45 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V List of Figures page 1 . Emotional Modulation of Startle Circuit 5 2. Experiment 1: SPQ-defined Groups at 4500 ms Lead Interval 22 3. Experiment 2: PCL-defined Group Scatterplot 34 4. Experiment 2: PCL-defined Groups at 4500 ms Lead Interval 39 5. Experiment 2: SPQ-defined Groups at 4500 ms Lead Interval 40 6. Experiment 2: PCL-defined Groups at 300 ms Lead Interval 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A b Examination of Affective Modulation, Psychopathy, and Negative Schizotypy in College and Community Samples As any emotion researcher can affirm, emotions are complex states. Nevertheless, the human experience of emotion is often taken for granted. Most people unfamiliar with this area of research are unaware of the great difficulties associated with the study of emotion. Despite ongoing arguments about the existence of a certain number of “basic” emotions, the validity of specifying a one-to-one relationship between physiological responses and specific emotions, the role of cognition in emotion and, of course, the ever-present worry about the reliability of self-report data, the past 30 years have seen great progress in this field. Much of current research supports and extends the view that emotional reactions are not only learned, but also part of our biological inheritance. Fischer, Shaver, and Carochan (1990) have offered the following as a working definition of emotion, “Emotions include appraisals or appreciations, patterned physiological processes, action tendencies, subjective feelings, expressions, and instrumental behaviors”. Motivation and Emotion The motivational approach to emotion has focused on describing the relationship between the affect (emotion) system and the drive (motive) system. According to Tomkins (1962; 1970), motives are signals of bodily need and affects amplify these signals. Whereas drives are primarily concerned with getting objects into or out of the body, affects are more general and, through learning, can become associated with any stimuli. Izard expanded on this approach by arguing that it is not the emotional reactions themselves that are learned, but the cues that trigger these reactions (Izard & Tomkins, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 1966). Izard also went on to embrace an evolutionary point of view, stating that in order to maximize chances for survival, an animal had to have emotionally triggered reactions or behaviors (such as fighting or fleeing) that were themselves initiated by innate releasers, drives, perceptions, or cognitions (Izard, 1972). In this latter expansion, the beginnings of an integrative approach to the study of emotion are seen. One theory that has integrated cognitive, behavioral, and physiological perspectives of emotion is Lang’s motivational theory of emotion. Lang (1995) postulates, similar to Fischer et al. (1990), that emotions can include “functional behaviors, evaluative and expressive language, and physiological events.” Specifically, emotions have been conceptualized here as action dispositions, “states of vigilant readiness that vary widely in reported affect, physiology and behavior” (Lang, 1995). The emotion model accompanying this conceptualization consists of two dimensions : Valence (positive and negative) and arousal (high and low). According to this theory, these two dimensions are activated by two opposing motivational systems in the brain: the appetitive system, which is prototypically expressed by behavioral approach (i.e. consummatory, sexual, nurturant behaviors) and the aversive system, prototypically expressed by escape and avoidance behavior (i.e. protective, withdrawing, defensive behaviors). The Startle Eveblink Reflex One specific defensive response is the startle eyeblink reflex in humans. The startle response serves as a protective mechanism put into use when the organism is faced with aversive stimuli. Therefore, in the context of Lang’s theory of emotion, the startle eyeblink response qualifies as both a functional behavior and a physiological event. The acoustic startle reflex has been studied extensively in animals and the simplicity of its Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3 neural circuit is well-documented (See Davis, 1996). Abrupt auditory input (e.g. a white noise probe) passes through the ventral cochlear nucleus into the reticular nucleus of the caudal pons. From there, efferent connections either pass through spinal neurons to stimulate the whole body startle in animals or innervate the seventh cranial nerve in humans evoking an eyeblink (Davis, 1996). Studies with the startle eyeblink response have consistently shown that variations in the intensity o f the acoustic startle-elicited eyeblink index the brain’s receptivity to information input (Filion, Dawson, & Schell, 1993; Filion, Dawson, & Schell, 1994; Dawson, Schell, Hazlett, Filion, & Nuechterlein, 1995). This is illustrated in the following study by the differences in eyeblink magnitude in response to positive and negative visual stimuli. Thus, a major advantage of studying the startle reflex is that modulation of this reflex depicts a change in a very simple reflex. The startle eyeblink response is modulated when the startle stimulus is preceded by a non-startling stimulus (a prepulse), such as an innocuous tone or a picture in the form of a slide. Early studies using acoustic startle stimuli (i.e. loud bursts of white noise) have shown a modality difference between auditory (i.e. tones) and visual (i.e. pictures in the form of slides) prepulses. With an auditory prepulse, inhibition of the startle eyeblink occurs with an early (300 ms or earlier) presentation of the startle stimulus whereas facilitation occurs later (1000 ms or later). With a visual prepulse, inhibition was observed both early and late. This difference in prepulse effect was explained by the subject’s focus of attention (Graham, 1980). With a visual prepulse, attention was in a different modality (visual vs. auditory) than the startle stimulus. However, Lang and his colleagues found in their emotion studies that an aversive visual prepulse (i.e. a negatively valenced picture) presented with a late (1000 ms or later) acoustic startle stimulus results in augmentation Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 of the startle eyeblink.1 This does not fit with cross-modal research and seemed to disprove the hypothesis that affective modulation is secondary to differences in modality determined attention (Lang, Bradley, & Cuthbert, 1990). Conversely, from fear conditioning studies with animals we know that the startle response can indeed be augmented when presented within a negative foreground. For example, through conditioning, animals learn to associate a non-startling light (conditioned stimulus) with an electric shock (unconditioned stimulus) presented sometime after. The presentation of the light thus induces a state of fear in the animal and when auditory startle probes are presented in this negative foreground, startle augmentation occurs. This well- established phenomenon in the animal literature fits with the findings by Lang and his colleagues. Furthermore, Lang and his colleagues have also found that the startle eyeblink response is inhibited when the startle stimulus is presented within a positive foreground (i.e. a positively valenced picture). A feasible explanation for this phenomenon is that it depends upon the matched or mismatched valence of the prepulse to the reflex (Lang et al., 1990). Consider again the proposition that the function of the startle reflex is protective. Therefore, if the prepulse has a negative valence (engaging defensive behavior), the startle eyeblink response is augmented. If prepulse has a positive valence (engaging appetitive behavior), a smaller response results. Again from animal studies, this emotional modulation of startle neural circuit has been well worked out. (See Figure 2) 1 The same affective modulation of startle was found with visual startle stimuli (flash of bright light). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5 Visual Prcpulse Pathway Perirhinal Cortex Lateral Geniculate Retina Lateral &Baso lateral Nuclei y Central Nucleus VCN NRPC Facial Nerve (Startle Eyeblink) Acoustic Probe Pathvvav Figure 2: Emotional Modulation of Startle Circuit (Davis, 1996) Due to their robust nature, startle modification paradigms are ideal for studying special subject populations (Lang, 1995). Attentional modulation of startle has already played a substantial role in schizophrenia research (Dawson, Hazlett, Filion, Nuechterlein, & Schell, 1993; Dawson et al., 1995), and recently, emotional modulation of startle has been used to study psychopathic and non-psychopathic prisoners (Patrick, Bradley, & Lang, 1993) and anxiety disorders (Cook, Hawk, Davis, & Stevenson, 1991). Through further use of startle modification paradigms, more can be learned about how these and other psychological disorders may be related. Schizotvpv: A Schizophrenia Spectrum Disorder Over the last thirty years much has been learned about cognitive deficits, biological anomalies, and the probable etiology of schizophrenia. One such accomplishment is the conceptualization of the schizophrenia spectrum. The schizophrenia spectrum is meant to serve as a useful conceptual tool to understand the multi-dimensionality of schizophrenia according to the different aspects of symptomatology characterizing the disorder. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Whereas the characteristic symptoms of schizophrenia involve a range of both cognitive and emotional dysfunctions, it has been well established that these varying symptoms can all be broadly conceptualized as positive or negative. Some examples of positive symptoms are delusions, hallucinations, disorganized speech and disorganized or catatonic behavior. Negative symptoms include restrictions in the range and intensity of emotion, thought and speech, and goal-directed behavior. Where there is emotional dysfunction, one may find psychological disorders. In fact, for several disorders listed in the DSM-IV (1994) excessive, diminished, or distorted emotion is a central feature. For example, schizophrenia and nearly every personality disorder along its spectrum share a blunted affect characteristic. Specifically, constricted range of affect is one core feature of schizotypal personality disorder. It has only been fifteen years since this major personality disorder was named in the DSM III (1980) as an Axis II disorder (Raine, Lencz, & Mednick, 1995). The description of the disorder was essentially unchanged in the more recent DSM IV (1994). Schizophrenia researchers are increasingly turning to the study of schizotypal populations, optimistically hoping to clarify issues that in schizophrenia research have often been clouded by the onset of psychotic episodes or possibly contaminated by the use of various antipsychotic medications (Raine & Lencz, 1995). As time passes, more and more researchers grow convinced that the study of schizotypal personality disorder carries significant potential for further understanding the etiology of schizophrenia. Among findings encouraging this anticipated link between schizophrenia and schizotypy is the high prevalence of schizotypal personality disorder among first degree relatives of schizophrenia patients (Ingraham, 1995). This suggests that there may exist a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7 genetic link between schizophrenia and schizotypal personality disorder. It is further hypothesized that this genetic link could result in some common biological predisposition that in schizophrenia sufferers has somehow become activated but has remained dormant in schizotypals (Raine & Lencz, 1995). Schizotypal personality disorder, like schizophrenia, has come to be conceptualized as a dimension with positive and negative aspects delineated by positive and negative symptoms similar in type to those for schizophrenia, but not as severely debilitating (Siever, 1995). Some of the positive symptoms of schizotypal personality disorder are ideas of reference, magical thinking, perceptual aberration, and paranoid ideation. Negative symptoms include excessive social anxiety, a lack of close friends, and constricted affect. A more recent conceptualization of both schizophrenia and schizotypal personality disorder identifies three symptom clusters: Positive symptoms,2 negative symptoms/ and disorganized behavior4 (Arndt, Alliger, & Andreasen, 1991; Raine, Reynolds, Lencz, Scerbo, Triphon, & Kim, 1994). These three-factor models appear to account better, theoretically and empirically, for the patterns of symptoms found in schizophrenia and schizotypy respectively. Moreover, the respective factors appear to share somewhat analogous symptoms (see footnotes). Thus, studying schizotypal populations in terms of these three factors may prove more effective than simply relying on old two-factor conceptualizations. : In schizophrenia, delusions, hallucinations, etc. (Arndt, ct al., 1991); In schizotypy. perceptual aberration, magical thinking, etc. (Raine et al., 1994) 3 In schizophrenia, anhedonia. affective flattening, etc. (Arndt, et al.. 1991); In schizotypy. lack of close friends, blunted affect, etc. (Raine ct al.. 1994) 4 In schizophrenia, thought disorder, bizarre behavior (Arndt, et al.. 1991); In schizotypy. odd behavior, odd speech Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Schizotypal Personality Questionnaire (SPQ) is a 74 item scale using DSM III-R (1987) criteria for the assessment of schizotypal personality disorder. Each item represents an aspect of one of the nine schizotypal traits listed in the DSM-IV. Each of these nine traits loads onto one of three factors: (1) the Cognitive-Perceptual factor, including positive symptoms, (2) the Interpersonal factor, including negative symptoms, and (3) the Disorganized factor, including odd speech and odd behavior. This questionnaire has been proven a valuable tool in assessing schizotypal traits in a non- forensic population. 55% o f high scores (defined as the top 10% of any given distribution) are likely to meet DSM-IV criteria for schizotypal personality disorder. The remaining 45% of high scores possess 3-5 schizotypal traits at a threshold or sub-threshold level. Most research on schizotypal personality disorder has revolved around the attempt to tie it in to the etiology of schizophrenia. Even now when this research is essentially only in its rudimentary stages there seems to be little doubt among the researchers involved that there is some type of etiological relationship between the two. Another personality disorder that may be etiologically related to schizophrenia, but one for which there is not as much evidence, is antisocial personality disorder. Linking Antisocial Personality to the Schizophrenia Spectrum A study by Heston (1966) was among the first to describe this possible link between antisocial behavior and schizophrenia. In this study, Heston showed that children o f schizophrenia patients tend to display criminal tendencies in addition to suffering from schizophrenia symptoms. A review by Silverton (1988) has additionally suggested that antisocial behavior may be legitimately placed along the schizophrenia spectrum because it Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 serves as a marker for a specific genotype at high risk for schizophrenia. As mentioned above, however, research with schizophrenia patients may be methodologically complicated or even contaminated due to psychotic episodes and/or the administration of antipsychotic drugs. Therefore, it is worthwhile to investigate the relationship between antisociality and schizotypy and attempt to determine where each disorder may fall along the schizophrenia spectrum. Closer examination of the relationship between these two disorders and the factors they may share could prove highly valuable in the investigation of the etiology of schizophrenia. It has been well established that psychopaths strongly exhibit an abnormal level of emotional detachment (Cleckley, 1976). Patrick et al. (1993) found that psychopaths did not exhibit the normal pattern of emotional modulation of startle. Instead of showing facilitation of the startle reflex with negative slides, these subjects exhibited significant inhibition. In fact, their startle patterns did not differ between the positive and negative slides, though both differed from the neutral slides. An important detail of this study is that Hare’s Psychopathy Checklist - R (PCL-R, 1991) was used to define groups high and low in psychopathy. The PCL - R consists of 22 behavioral & personality items, each scored on a 3 point scale from interview and file information. The items are based on Cleckley’s (1976) clinical conception of psychopathy. Each item loads onto one of two factors: Factor 1 is concerned with core personality traits, and inferences about interpersonal processes. Patrick et al. have conceptualized this factor as measuring “emotional detachment.” Factor 2 measures chronic antisocial behavior. It is basically a measure of social deviance and depends on identifying the occurrence of specific Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. behaviors.5 Patrick et al. hypothesized that abnormalities in affective modulation of startle patterns should be related to their emotional deficits rather than their antisocial behavior. Therefore, Patrick et al. divided subjects into groups based on their PCL Factor 2 (antisocial behavior) scores (high and low) and examined only the ‘"high” antisocial group. Patrick et al. then divided these subjects into two subgroups based on their Factor 1 (emotional detachment) scores. Subjects with low emotional detachment (Factor 1) scores showed the normal pattern of emotional modulation (inhibition for positive slides, facilitation for negative slides) whereas subjects with high emotional detachment scores showed no difference between positive and negative slides. Patrick et al. concluded that abnormalities in psychopath’s startle modulation are related specifically to their emotional and interpersonal deficits. These findings have been replicated and extended in a recent study (Levenstein, Patrick, Bradley, & Lang, 1996). In addition to their failure to show the distinctive blink pattern for positive and negative slides, psychopaths also failed to show as much prepulse inhibition at early lead intervals (300 ms) as non-psychopaths (Patrick et al., 1993). Although no directed attentional task was given in this study, this preliminary finding does seem to suggest that psychopaths may also display attentional deficits similar to schizophrenia patients. Thus, it would appear that one link between antisociality and schizotypy is their emotional deficiency, relating both to the negative symptomatology of schizophrenia. These emotional deficits, in turn, may be indexed through the observance of abnormalities in affective modulation patterns. Another shared characteristic might be their failure to 5 Interestingly. DSM IV criteria Antisocial personality' disorder strongly correlates with Factor 2. Factor I correlates highest with Narcissism Personality disorder. Total PCL scores positively correlate with prototypicality ratings of DSM III antisocial, histrionic and narcissistic personality' disorders, negatively Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11 show significant prepulse inhibition at early lead intervals. To date, we are not aware of any studies testing a schizotypal population with the emotional slide paradigm, but Cadenhead, Geyer, and Braff(1993) did find subjects with schizotypal personality disorder to be deficient in (auditory) prepulse inhibition at early lead intervals. In addition, studies with psychosis-prone college students have suggested that these individuals share attentional deficits similar to those of schizophrenia patients (Schell, Dawson, Hazlett, and Filion, 1995). The Role of Electrodermal Activity Another physiological response system widely studied in the field of schizophrenia research is the electrodermal response system. The skin conductance orienting response (SCOR) is a sensitive measure thought to reflect the amount of attention allocated to a particular stimulus and the extent to which that stimulus (presumably a novel stimulus) is cognitively processed (Dawson et al., 1989). An enhanced SCOR to a stimulus, for example, is thought to reflect enhanced attention to that particular stimulus. Interestingly, whereas 90% of normal individuals will show orienting responses to a novel stimulus, 40- 50% of schizophrenia patients do not show SCORs at all. Conversely, a significant minority of schizophrenia patients are hyperresponsive, showing very little or no habituation to the orienting stimulus (Raine, Lencz, & Benishay, 1995). Referring to the schizophrenia spectrum, evidence suggests that schizophrenia nonresponsivity is associated with negative symptoms while schizophrenia hyperresponsivity is associated with positive symptoms (Raine, Lencz, & Benishay, 1995) Both types of responding reflect overall abnormal attentional processes in schizophrenia patients. correlate with avoidant personality disorder (Harpur, Hare, & Hakstian. 1989). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Studies have also noted the prevalence of psychophysiological underarousal of the autonomic nervous system with antisocial personality (Raine, 1988; Raine & Venables, 1992). Only a partial link could be established, however, until the subjects were divided into two groups: those with and without schizoid tendencies. Raine and Venables (1984) were among the first to demonstrate this distinction. In their study, subjects (all antisocial) were divided into groups based on responsivity or nonresponsivity of SCORs. Their results showed that the non-responders could all be characterized by schizoid tendencies and not sensation-seeking while the responders showed the opposite characterization. These results were interpreted as evidence that SC non-responding could represent a common biological factor in antisocial as well as schizophrenia disorders, and may possibly reflect abnormal allocation of attentional resources to incoming stimuli (Raine & Venables, 1992). Like schizophrenia and schizotypal personality, it would seem that antisocial personality too can be conceptualized as a dimension with positive (sensation-seeking, SC responsivity) and negative (schizoid tendencies, SC nonresponsivity) features. Attentional deficits are also well incorporated into the antisocial literature. Many antisocials may have suffered from attention deficit disorder as children and even as adults show problems with sustaining attention (Raine & Venables, 1992). One hypothesis about the link between antisociality and schizophrenia speculates that (a) there is a genetic link between the two disorders and (b) specific environmental stressors can interact with this genotype to produce either schizophrenia or antisocial personality (Raine & Venables, 1992). It has also been found that an unusually high number of people diagnosed with schizotypal Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13 personality disorder in the negative aspect are also diagnosed with antisocial personality disorder (Siever, et al. 1990). The Present Studies This investigation was conducted in two parts. The first study attempted to replicate past Lang et al. emotional modulation studies using college students as subjects. Some paradigm differences were introduced and, in addition to the examination of emotionally modulated startle patterns, SPQ factor scores of these college students were examined. It was hypothesized that students scoring high on the Negative Symptom factor of the SPQ (Factor 2) would show no differences in startle eyeblink response to positive and negative stimuli at the 4500 ms lead interval, signifying a deficit in emotion processing. Subjects scoring low on the Negative Symptom factor of the SPQ would show normal affective modulation patterns at the 4500 ms lead interval; augmented startle eyeblink responses to negative stimuli compared to positive stimuli. In addition, presuming that subjects scoring high on the Negative Symptom factor of the SPQ would have a higher overall SPQ score, it is predicted in accordance with Cadenhead et al. that these high negative symptom subjects will show less prepulse inhibition at the 300 ms lead interval (regardless of slide valence) than subjects scoring low on this factor. The second experiment used the same emotional slide paradigm on a subject sample consisting of volunteers from the metropolitan community of Los Angeles. These subjects also had SPQ scores and, in addition, PCL scores to examine. For this second experiment, in addition to the hypotheses delineated above, it was expected that Patrick et al.’s (1993) findings with groups high in antisociality would be replicated as follows. For the group high in emotional detachment (PCL- factor 1) and in antisociality, there would Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14 be no differences in startle patterns between positive and negative slides at the 4500 ms lead interval, whereas the group high in antisociality but low in emotional detachment would show normal startle modulation. Unlike in Patrick et al., the groups low in antisociality were also examined here. For these groups it was likewise expected that those high in emotional detachment would display abnormalities n affective modulation compared to those low in emotional detachment. Subjects high in emotional detachment and/or schizotypal negative symptoms were expected to show less prepulse inhibition at 300 ms than others in accordance with the findings of Cadenhead et al.(1993) and Levenston, et al. (1996) and smaller SCORs to slides (regardless of valence) were expected from subjects high in emotional detachment and/or schizotypal negative symptoms. Experiment 1 Method Subjects 34 undergraduate students at the University of Southern California were recruited from an introductory psychology class and received course credit for participation in one test session. These students had filled out the Schizotypal Personality Questionnaire (SPQ) distributed in class prior to participation in this experiment, but the experimenter was blind to all scores. Eight subjects were excluded due to missing SPQ scores. Three additional subjects were excluded due to inadequate responsiveness to the startling stimuli, meaning that their 20 percent trimmed average baseline startle measured less than one microvolt (gV). Therefore the total number of subjects was 23 (4 male, 19 female). Of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. these 23 subjects, 8 were Asian/Pacific Islander, 6 were White, 5 were Hispanic, and 4 were Middle Eastern. The median age was 18 years. Design This study used a 2 X 2 X 3 mixed factorial design. The first variable consisted of two subject groups defined by a median split on the SPQ Factor 2 scores (Mdn = 15). This resulted in 12 subjects in the Low Negative Symptom group and 1 1 subjects in the High Negative Symptom group. The second variable consisted of two types of prepulses: (I) positively valenced slides, and (2) negatively valenced slides. These slides were selected from the International Affective Picture System (IAPS; Lang, Ohman, & Vaitl, 1988).6 The third variable consisted of three lead intervals: 300, 800 and 4500 ms following slide onset. Procedure Upon arrival, subjects were asked to sign a consent form giving a brief overview of the testing sessions. The session then began with a hearing screening test, testing the subject’s hearing at 1000, 1500, and 2000 Hz. These frequencies are within the range of the human voice. A brief introduction was then presented on an audiotape followed by the attachment of electrodes for the recording of skin conductance and startle eyeblink. The testing session began with a tone counting task lasting approximately 20 minutes.7 Subjects were then given a total of four emotion questionnaires to fill out.8 Upon completion of these questionnaires, subjects were presented with audiotape instructions 6 For positive pictures, the IAPS identification numbers are: 160. 175. 234. 253. 465. 468, 520. 735. 803. 818; for men. 421 and 429; for women 447 and 449. For negative pictures the identification numbers au-e: 107. 112. 273. 300. 312. 314. 318, 323. 620. 623. 823. and 904. This tone counting task is an attentional modulation paradigm. Due to time and space limitations the findings associated with this part of the experiment will not be discussed here. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16 stating that their task was to watch the slides presented for the entire time that they were on. Subjects were also told that brief loud static noise bursts would be presented occasionally during the slide presentation but that these could be ignored. A total of 24 slides were presented - 12 positive and 12 negative. These 24 slides were organized into three blocks of eight slides each, four positive and four negative in a fixed, mixed order. Two of the slides (one positive and one negative) were unprobed to allow measurement of skin conductance responses. The order of lead intervals in the remaining six slides was varied in three sequences so that each lead interval appeared equally often in each ordinal position for each valence type of slide within a trial block (300 positive, 300 negative, 800 positive, 800 negative, 4500 positive, 4500 negative). There were two slide order sequences. The eight slides within each block were pseudorandomly ordered in pairs. Each pair was switched for the second slide order sequence.9 Startle stimuli were presented during 18 of the inter-trial intervals, 6 occurring during each trial block. These stimuli were presented at random intervals between 10 and 20 seconds after slide offset. The startle-blink magnitudes elicited by these intertrial interval probes (ITI probes) served as baseline measures with which to compare blink magnitudes to the same startle stimuli presented during the slides. Upon completion of the last slide the subjects were dismissed. 8 Due to time and space limitations, the questionnaire scores will not be discussed here. 9 A slide order error was discovered in Slide Order 1 toward the end of data collection. Howev er, these data were compared with data from Slide Order 2 subjects and there were no significant differences between the two distributions. This error is explained in detail in Appendix A. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 17 Experimental Stimuli The startle stimulus consisted of a 104 dB(A) white noise that was 40 ms in duration. The startle noise was generated by a Grason-Stadler 901B noise generator and was gated at a near instantaneous rise/fall time. All of the auditory stimuli were presented binaurally through headphones (Telephonies TDH-49P). The onsets, durations, and intervals between stimuli, except for the time period between trials, were all controlled by a 486 microcomputer with a Metrabyte DAS-16 A/D board and a custom program written in ASYST. Recording and Scoring of Dependent Variables The primary dependent variables were the modification of the startle eyeblink magnitude produced during the probed prepulses at the 300 ms, 800 ms, and 4500 ms lead intervals and the magnitude of the skin conductance orienting response (SCOR) collected during positive and negative non-probed slides. SCORs were measured according to the standard procedures outlined by Dawson, Schell, and Filion (1990). Skin conductance was recorded from the volar surface of the distal phalanges of the first and second fingers of the nonpreferred hand with silver silver- chloride electrodes filled with .05 molar NaCl in a Unibase paste. Double-sided adhesive collars allowed a 10 mm contact area with the skin. A constant 0.5 V was applied across the electrodes, and the skin conductance signal was amplified by a Grass 7P1 preamplifier and a 7DAE driver amplifier. SCORs were computer scored off-line as increases in skin conductance beginning between 1.0 and 4.0 seconds following tone onset, having a minimum response amplitude of .05 gS, and a slope of 300. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Startle eyeblink amplitudes were measured as electromyographic activity (EMG) from two miniature electrodes (4 mm in diameter) placed over the obicularis oculi muscle of the left eye, one electrode centered below the pupil and the other approximately 10 mm lateral to the first. The EMG signal was fed to a Grass 7P3 wide band integrator/preamplifier and a 7DA driver amplifier. Eyeblinks were recorded at full wave rectification, with and without integration at a time constant of 20 ms. The EMG signal was digitized at a rate of 1000 Hz for 300 ms following the presentation of each startle- eliciting loud noise. The startle eyeblink amplitude was then scored off-line with the SPD- EMO Peaks and Averages program, a custom algorithm developed by Dr. Eric Vanman. In this algorithm, the amplitude of each EMG response is scored in microvolts (gV) as the difference between the mean EMG activity in the 200 ms preceding the onset of the startle probe and the peak of the response to the startle probe. The peak of the response is defined as the highest gV average calculated across three EMG samples (because raw EMG was collected at 2000 Hz, the average was calculated across 1.5 ms). Differences were then computed between the mean baseline startle-alone eyeblink amplitude and the intratone eyeblink amplitudes. Because difference scores in absolute gV units are correlated with baseline startle blink amplitude, the difference scores were converted to percentage change units, which partially removes the dependence on baseline. A positive SEM (startle eyeblink modification) score indicated startle facilitation relative to baseline, whereas a negative SEM score indicated startle inhibition relative to baseline. Results The following results are reported in two sections. The first analyses discussed were performed using robust methods in S-Plus statistical software. The second section Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 19 discusses comparisons performed in BMDP statistical software using traditional methods. In this traditional method section, analyses performed with the raw data and with replacement of outliers are included. Data Analysis Using Robust Methods Due to the numerous problems associated with standard hypothesis testing methods in terms of detecting differences between groups or associations between random variables (see Huber, 1981; Hampel, Ronchetti, Rousseeuw, & Stahel, 1986; Wilcox, 1996; 1997; in press), this set of analyses were performed with robust tests using custom written functions for S-Plus Statistical Software.1 0 Robust methods achieve higher power and more accurate probability coverage than standard ANOVA and regression methods with data sets that are affected by skewness, heteroscedasticity, and/or outliers (Wilcox, 1997; in press). Therefore, robust methods are far more likely to detect differences between groups (if they exist) and more accurately depict the magnitude of these differences. In other words, when standard techniques reject null hypotheses, groups probably do differ, but often, standard techniques failing to reject null hypotheses may be missing important group differences (Wilcox, in press). Descriptive statistics. Trimmed means1 1 for both the Low and High Negative Symptom groups’ startle eyeblink and SCOR measures are listed in Tables 1 and 2. 1 0 These functions were written by Dr. Rand Wilcox of the Department of Psychology at the University of Southern California and can be found at the Academic Press Website at http://www.apnet.coni/www/ap/bkupdats.htm. S-PIus is distributed by Math Soft and information can be obtained by c-mailing Math Soft at mktg@statsci.com. 1 1 20 percent trimming was used in all analyses using trimmed means. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 1 - Experiment 1 Trimmed Means and Trimmed Standard Errors for Percent Change in Startle Modulation (pV) Slide Valence/ Probe Position High Negative Symptom n=l 1 (trimmed mean) High Negative Symptom n=l 1 (trimmed standard error) Low Negative Symptom n=12 (trimmed mean) Low Negative Symptom n= 12 (trimmed standard error) Positive 300 -38.86 18.19 -38.35 10.84 Negative 300 -19.36 20.88 -32.82 19.95 Positive 800 -17.02 16.90 0.75 13.48 Negative 800 -13.91 16.25 7.21 22.34 Positive 4500 8.56 17.48 23.51 26.31 Negative 4500 48.36 36.88 28.14 24.47 Table 2 - Experiment Trimmed Means and Trimmed Standard Errors for Skin Conductance Orienting Response Amplitudes (pS) Slide Valence High Negative Symptom n - 1 1 (trimmed mean) High Negative Symptom n=I 1 (trimmed standard error) Low Negative Symptom n=12 (trimmed mean) Low Negative Symptom n=12 (trimmed standard error) Positive 0.26 0.04 0.04 0.04 Negative 0.36 0.18 0.32 0.17 21 Emotional valence and the startle response. The Yuen - Welch test with a percentile t bootstrap for dependent groups was performed within each group at each lead interval to determine whether a significant amount of emotional modulation (difference in response to negative versus positive) was present. This test computes a .95 confidence interval for the difference between two 20% trimmed means (pt 1 - pt2) in dependent groups using a bootstrap size of 599 (Wilcox, 1997). When combining this percentile t boot strap method with trimmed means, in most cases, near optimal probability coverage and power are obtained (Wilcox, in press). Performed on the overall sample in experiment one, there were no significant differences between positive and negative responses at any lead interval. Neither were there significant differences between positive blink responses and negative blink responses for either group at any lead interval at the a = .05 level (that is, all .95 confidence intervals computed contained zero and were therefore not significant). However, although not significant, the High Negative Symptom group did show a rather large positive - negative difference at 4500 ms in the predicted direction (See Figure 3). The .95 confidence interval yielded was equal to (-96.57, 74.42) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. % c h an g e ( M V ) 50-i 45- 40- 35- 30 - 25 - 2 0 - 15 - 10 - 5 - 0 - - I * Positive □ Negative High Low Negative Negative Symptom Symptom Figure 2: SPQ-defined Groups as 4500 ms Lead Interval Prepulse inhibition. The Yuen - Welch test with a percentile t bootstrap for independent groups was performed between the two groups at 300 ms to determine any group differences in the amount of prepulse inhibition exhibited, regardless of valence. No significant difference in prepulse inhibition at 300 ms between the two groups was found; however the High Negative Symptom group did show slightly less prepulse inhibition (Trimmed M = -29.58) than the Low Negative Symptom Group (Trimmed M = -36.52). Skin conductance. There were no significant differences in skin conductance responses to positive and negative slides in either group. However, the High Negative Symptom group (Trimmed M = 0.26 pS) displayed significantly larger SCORs to the positive stimuli than the Low Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23 Negative Symptom group (Trimmed M = 0.035 gS). The .95 confidence interval yielded was equal to (-0.36, -0.13). No significant differences between groups were found for the skin conductance responses to the negative slides nor for overall skin conductance amplitude size (regardless of valence). Data Analyses Using Traditional Methods Descriptive statistics. When re-analyzing these data using traditional methods, no trimming was used so inadequate responsiveness was defined as having a mean baseline startle measuring less than one microvolt (gV). As a result of using this new criterion, in the raw analyses there were 13 subjects in the Low Negative Symptom group (as opposed to 12 in the above analyses). The number of subjects in the High Negative Symptom group remained the same (n=I 1). In the outlier-identified and replaced analyses, one subject from the High Negative Symptom group had to be dropped due to irreplaceable outlier values. Thus, for the outlier-identified and replaced analyses only, the High Negative Symptom group had 10 subjects. Means and standard errors are listed in tables 3 through 6 for both the Low and High Negative Symptom groups’ startle eyeblink and SCOR measures. Tables 3 and 4 contain statistics from the raw analyses, whereas tables 5 and 6 contain statistics from outlier-identified and replaced analyses. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3 - Experiment 1 Means and Standard Errors for Percent Change in Startle Modulation (pV) - Raw Analyses Slide Valence/ Probe Position High Negative Symptom n-11 (mean) High Negative Symptom n=l 1 (standard error) Low Negative Symptom n=13 (mean) Low Negative Symptom n=13 (standard error) Positive 300 -45.81 9.54 -31.04 13.42 Negative 300 -31.47 11.23 -39.76 11.74 Positive 800 6.26 35.87 -4.55 10.46 Negative 800 -8.25 18.59 3.03 14.01 Positive 4500 2.39 16.68 11.60 16.49 Negative 4500 46.59 27.84 25.41 15.40 Table 4 - Experiment 1 Means and Standard Errors for Skin Conductance Orienting Response Amplitudes (pS) - Raw Analyses Slide Valence High Negative Symptom n=l 1 (mean) High Negative Symptom n=l 1 (standard error) Low Negative Symptom n=13 (mean) Low Negative Symptom n= 13 (standard error) Positive 0.34 0.08 0.16 0.07 Negative 0.40 0.09 0.39 0.12 t o •t. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 5 - Experiment 1 Means and Standard Errors for Percent Change in Startle Modulation (pV) - Outlier-Identified and Replaced Analyses _____________ ___________________________________________________________ Slide Valence/ Probe Position High Negative Symptom n=10 (mean) High Negative Symptom n=10 (standard error) Low Negative Symptom n=13 (mean) Low Negative Symptom n=13 (standard error) Positive 300 -48.25 10.15 -28.84 13.26 Negative 300 -31.45 12.38 -38.13 11.52 Positive 800 11.91 39.23 -3.12 9.77 Negative 800 -2.21 19.57 6.36 13.68 Positive 4500 7.64 17.57 13.95 15.88 Negative 4500 56.51 28.82 31.80 16.72 l v > 26 Emotional valence and the startle response A 2 (Group: Low Negative Symptom, High Negative Symptom) X 2 (Valence: Positive, Negative) X 3 (Lead Interval: 300 ms, 800 ms, 4500 ms) analysis of variance (ANOVA) with repeated measures was performed on percent change scores, first on the raw data and then again after accounting for and replacing outliers1 2 . The raw analysis revealed a marginally significant main effect of valence, F (1,22) = 3 .34, p = 0.0810, and a significant main effect of lead interval, F (2, 44) = 10.33, g = 0.0002. For the outlier-identified analysis, this valence effect was significant, F (1,21) = 5.48, g = 0.0292, in addition to a significant main effect of lead interval, F (2, 42) = 13.12, g < 0.0001. Subjects were exhibiting the largest startle responses to the negative stimuli probed at 4500 ms. There was no main effect of group and no significant interactions. T-tests were performed on a within group basis at each lead interval. For the raw data, there was an overall (all subjects tested) significant difference between responses to the positive and negative stimuli probed at 4500 ms, T (32) = -2.06, g = 0.0476. For the High Negative Symptom group, the difference at 4500 ms was marginal, T (19)= -1.94, g = 0.0670, whereas the Low Negative Symptom group was quite far from being significant, T (21) = -0.92, g = 0.3679. There was no significant emotional modulation at the 800 ms lead interval. For the outlier-identified analysis, the positive-negative difference at 4500 ms was marginal for all subjects, T (31) = -1.84, g = 0.0754 and neither the High nor the Low Negative symptom groups displayed significant emotional modulation at 4500 ms. Again, there was no significant emotional modulation at the 800 ms lead interval 1 2 Outliers were defined as 3 standard deviations above the mean and two standard deviations above the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 27 Prepulse inhibition. There were no significant between-group differences in the amount of overall prepulse inhibition (regardless of valence) at 300 ms for either the raw or outlier-identified analyses. Skin conductance. A 2 (Group: Low Negative Symptom, High Negative Symptom) X 2 (Valence: Positive, Negative) analysis of variance (ANOVA) with repeated measures was performed on SCOR data. No outliers were detected, so only one analyses was performed. A significant main effect of valence was found, F (1, 21) = 5.32, p = 0.03, with slightly larger responses to the negative stimuli. There was no main effect of group, but it would appear that the High Negative Symptom subjects were displaying much larger SCORs to positive stimuli than were the Low Negative Symptom subjects. Both groups appeared to respond similarly to negative stimuli. Interpreting the Different Analyses Results from the traditional analyses, in some instances, varied slightly from those of the robust analyses. Whereas the robust analyses did not detect any overall significant valence effect, the raw analyses revealed a marginally significant main effect of valence, and a significant combined group difference between positive and negative stimuli at 4500 ms. Additionally, the raw analyses revealed a marginally significant difference between responses to positive and negative stimuli for the High Negative Symptom group at 4500 ms. The outlier-identified and replaced analyses revealed a significant main effect of valence, but only a marginal combined group difference at 4500 ms. Neither group preceding value. Once identified, outliers were replaced on a within-subject basis with an adjacent value. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28 showed significantly different responses to positive and negative stimuli at any lead interval under the outlier -identified and replaced analyses. The question of which analyses procedure should be trusted and which results should be accepted with the most confidence must now be addressed. In the outlier- identified and replaced procedure outliers are taken care of in a somewhat questionable manner. Once a value is identified as an outlier, by criteria, however reasonable, some might call subjective, it is replaced by an adjacent value. The choosing of this adjacent value is also somewhat subjective. In addition to this already dubious process, the effects on the overall sample are not really taken into account. In changing extreme values to less extreme values, the entire distribution is changed. Perhaps this new distribution is seemingly more representative of the majority of the data, but mathematically it really does more harm than good. Specifically, the standard error assumed under such changes is inaccurate. The outliers have not really been taken care of at all. Power is still relatively low, compared to that using robust methods, and confidence intervals are inaccurate. Moreover, after all these problems, this method relies on the sample mean as the measure of location best suited to represent the data. Why the sample mean is less than desirable to use in this situation is explained in the discussion of the raw traditional analyses. The problem with raw traditional analyses is that they rely on means and do nothing to take outliers into account. The sample mean has a finite sample breakdown point of 1/n. The finite sample breakdown point is the smallest proportion of n observations that can render a measure of location meaningless (Wilcox, 1996). Thus, one single aberrant observation can inflate the sample mean to such an extent that it is no longer the best data representative. The fact that outliers are not at all considered tells us Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 29 simply that we do not know how many “aberrant” data points have affected our mean. As opposed to the outlier-identified and replacement method where we try to fix outliers, this raw method does nothing. At this time in mathematical and quantitative psychology circles, using the percentile t bootstrap with the trimmed mean is the best method known to provide accurate probability coverage and maximize power (Wilcox, 1997; in press). Although there may be some instances where other methods give better results (no method works aU the time), and it may be tempting to adopt those methods to fit our purpose, the trimmed mean and percentile t bootstrap are mathematically sound procedures that will work most of the time. Thus, in applied and experimental work, it seems most sagacious to choose the robust analyses results to rely on. Those are what the discussion section will interpret. Discussion It was hypothesized for this experiment that subjects displaying more negative schizotypal symptoms (those scoring high on SPQ Factor 2) would display abnormal patterns of emotional startle modification. Although there were no significant findings with respect to the startle eyeblink data, the opposite of what was predicted occurred. The High Negative Symptom group displayed positive - negative differences in the normal direction while the Low Negative Symptom group did not. Although this finding seems counterintuitive, looking at the schizotypal traits in the DSM - IV that load on to this Negative Symptom factor may help clarify these findings. The four traits are: (1) suspiciousness or paranoid ideation, (2) inappropriate or constricted affect, (3) lack of close friends, and (4) excessive social anxiety. Whereas two of these traits (constricted affect and lack of close friends) are suggestive of “withdrawal type behaviors”, the other Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30 two factors (suspiciousness and excessive social anxiety) may be thought of as “fearful” state correlates. A study by Cook et al. (1991) found that startle potentiation was actually enhanced in high fear subjects compared with low fear subjects. This suggests that the extent to which subjects show affective modulation is systematically related to their individual states of fear and anxiety. Thus, the important difference between subjects scoring high on SPQ Factor 2 and those scoring low may be these states of higher fear and anxiety with those scoring low on SPQ Factor 2 characterized by low states of fear and anxiety.1 3 Cook (in press) explains that these low fear subjects showing reduced or even absent affective modification of startle (as in this experiment) may not mean they have deficits in affective processing. Instead, Cook suggests, it may be that low fear subjects simply process affective stimuli through a different mechanism than high fear subjects. An alternative explanation requires that a clear distinction be made between affective state and affective expression. It may be more accurate to describe the affective flattening in schizotypy as a phenomenon of affective expression rather than affective state, as hypothesized by Kring and Neale (1996) for the affective flattening in schizophrenia patients. It should be noted here, that the SPQ is a self-report, verbal questionnaire. Thus, subjects here scoring high on the SPQ’s constricted affect and lack of close friends scales who are also showing emotional modification of startle are displaying a disjunction between their verbal and physiological expressions of emotion. Thus, it may be that these subjects are not in a state of constricted affect, but rather they have problems expressing emotion in conventional ways. 1 3 Post hoc correlations were performed between the four trait scores loading onto SPQ Factor 2 and emotion modification of startle scores. No significant correlations resulted, however this theory definitely merits further investigation. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 31 The finding that the High Negative Symptom group displayed higher SCOR amplitudes to positive stimuli than the Low Negative Symptom group may be a fluke. There is no prior evidence suggesting any reason for valence differences in skin conductance responsivity. However, perhaps this might be partially explained by remembering that autonomic orienting to a stimulus is indicative of the amount of cognitive processing occurring in relation to a stimulus (Dawson Filion, & Schell, 1989). Looking at the trimmed mean amplitudes for each group’s responses to positive and negative stimuli (see Table 2), the Low Negative Symptom group’s positive trimmed mean amplitude is strikingly smaller than that in any other condition. Therefore this finding suggests that the Low Negative Symptom group is not processing the positive stimuli to the same extent that they are processing the negative stimuli. In other words, they are paying more attention to the negative stimuli than to the positive stimuli, whereas the High Negative Symptom group appears to be processing both positive and negative stimuli equally. The implications of this finding are twofold. First, Cuthbert, Bradley, and Lang (1996) have demonstrated that pictures higher in arousal produce larger SCORs. Additionally, subjects showed significantly larger responses to unpleasant stimuli than to pleasant stimuli. This explanation makes sense for the Low Negative Symptom group. For the High Negative Symptom group, it would appear under this interpretation that both positive and negative stimuli were equally arousing. If these subjects are higher in fear and anxiety, it could be that they are simply hypervigilent to all stimuli presented. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 32 Experiment 2 Method Subjects 64 subjects for this analysis were drawn from participants of an ongoing study at the University o f Southern California studying various psychophysiological, neurophysiological, and clinical phenomena associated with schizotypal personality disorder. Subjects were recruited for this study from temporary employment agencies located in the metropolitan Los Angeles area. Prior to participation in this testing session, these subjects had all been assessed for psychopathy with the Psychopathy Checklist (Screening Version, 1995) and completed the Schizotypal Personality Questionnaire. As in Experiment 1, the experimenter was blind to all scores. One subject was omitted from analysis due to missing PCL scores and six subjects were omitted due to inadequate responsiveness to the startle stimuli. This left 57 subjects total (49 male, 8 female). Of these remaining subjects there were 23 White, 19 African-American, 5 Hispanic, 1 Native American, and 9 of unknown ethnicity. The median age was 30 years. Design This study used a 4 X 2 X 3 mixed factorial design. The first variable consisted of 4 subject groups. These groups were defined first by a median split on subjects’ PCL Factor 2 scores measuring antisocial behavior (Mdn = 6)1 4 . These high and low antisocial behavior groups were then divided into subgroups based on their respective Factor 1 medians measuring emotional detachment: For the Low Antisocial group, Factor I Mdn = 1 - 1 Subjects with scores at the median were placed in the Low Antisocial group. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 33 I 1 5 ; For the High Antisocial group, Factor 1 Mdn = 41 6 (see Figure 4). Groups 1-4 were then named as follows: (1) High Antisocial/Low Emotional Detachment (n = 13); (2) High Antisocial/High Emotional Detachment (n = 12); (3) Low Antisocial/Low Emotional Detachment (n = 21); (4) Low Antisocial/High Emotional Detachment (n = 11). The second variable consisted of two types of prepulses: (1) positively valenced slides, and (2) negatively valenced slides. These slides were exactly the same as those used in Experiment 1 . The third variable consisted of three lead intervals: 300, 800 and 4500 ms following slide onset. Subsequent to analyses with these PCL defined groups, parallel analyses were conducted with groups defined by their SPQ Factor 2 scores measuring negative symptoms of schizotypy. Subjects were therefore divided into two groups based on a median split on these Negative Symptom scores (Mdn = 9)1 7 . This resulted in 32 subjects in the Low Negative Symptom group and 25 subjects in the High Negative Symptom group. 1 5 Subjects with scores at the median were placed in the Low Emotional Detachment group. 1 6 Subjects with scores at the median were placed in the Low Emotional Detachment group. 1 Subjects with scores at the median were placed in the Low Negative Symptom group Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ilig h /IIig h 34 CN CO C M CO C O jO|Aeij9g lejoosnuv c 3 fi. Q. * E . c £ Q. L . < u to o C /3 Q . 3 O w o *T3 0 > c < u T ? <J C L < U u_ 3 2 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35 Procedure The protocol for Experiment 2 was exactly the same as that in Experiment 1.1 8 Results As in Experiment 1, the following results are reported in two sections, using first robust methods and then traditional methods. Results Using Robust Methods Descriptive Statistics Trimmed means for the startle eyeblink and SCOR response measures of the PCL defined groups are listed in Tables 7 and 8. Trimmed means for the startle eyeblink and SCOR response measures of the SPQ defined groups are listed in Tables 9 and 10. Analyses were performed separately for the PCL defined groups and the SPQ defined groups, but will be discussed together in the following sections. 1 8 The same slide order error described in Experiment I occurred in Experiment 2. Again, the distributions of scores from subjects in both orders were compared and no differences were found. See footnote 9 and Appendix A. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 6: Experim ent 2 PCL-defined G roup Trimmed M eans anc Trimmed Standard Errors for 'ercent Chan ge in Startle M odulation (gV ) Slide Valence/ Probe Position High-Low 11=13 (trimmed mean) High-Low n=13 (trimmed standard error) High-High N=12 (trimmed mean) High-High N=12 (trimmed standard error) Low-Low N=21 (trimmed mean) Low-Low N=21 (trimmed standard error) Low-High N =11 (trimmed mean) Low-High N =11 (trimmed standard error) Positive 300 -51.41 18.35 -43.04 12.62 -29.02 11.77 -47.79 15.36 Negative 300 -53.51 12.96 -49.74 16.49 -18.13 8.31 -43.24 15.70 Positive 800 -22.30 18.74 -48.44 5.73 -28.21 8.72 10.30 19.56 Negative 800 -22.53 16.79 -35.75 11.72 -9.59 6.16 -5.78 18.22 Positive 4500 -13.45 17.04 -4.54 15.88 -3.32 12.00 8.60 18.61 Negative 4500 46.92 24.71 -8.18 17.31 17.16 12.04 61.44 35.70 Table 7 - Experiment 2 PCL-defined Group Trimmed Means and Trimmed Standard Errors for Skin Conductance Orienting Response Slide Valence High-Low' n=13 (trimmed mean) High-Low n=13 (trimmed standard error) High-High N=12 (trimmed mean) High-High N=12 (trimmed standard error) Low-Low N=21 (trimmed mean) Low-Low N=21 (trimmed standard error) Low-High N=ll (trimmed mean) Low-High N =11 (trimmed standard error) Positive 0.03 0.02 0 0 6 0.06 0.15 0.08 0.01 0 01 Negative 0.05 0.04 0.17 0.09 0.16 0.06 0.02 0.04 O N Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 8 - Experiment 2 SPQ-defined Group Trimmed Means and Trimmed Standard Errors for Percent Change in Startle Modulation (gV) ____________________________________________________________________ Slide Valence/ Probe Position High Negative Symptom N=25 (trimmed mean) High Negative Symptom N=25 (trimmed standard error) Low Negative Symptom N=32 (trimmed mean) Low Negative Symptom N=32 (trimmed standard error) Positive 300 -37.16 8.64 -41.39 10.29 Negative 300 -34.09 9.43 -35.97 11.08 Positive 800 -28.49 7.33 -23.50 9.48 Negative 800 -3.33 12.00 -24.10 6.90 Positive 4500 -5.52 10.35 -0.40 11.96 Negative 4500 37.13 13.40 3.29 10.74 Table 9 - Experiment 2 SPQ-defined Group Trimmed Means and Trimmed Standard Errors for Skin Conductance Orienting Response Slide Valence High Negative Symptom N=25 (trimmed mean) High Negative Symptom N=25 (trimmed standard error) Low Negative Symptom N=32 (trimmed mean) Low Negative S>mptom N=32 (trimmed standard error) Positive 0.02 0.02 0.08 0.05 Negative 0.05 0.03 0.13 0.05 38 Emotional Valence and the Startle Response Unlike in Experiment 1, there was an overall significant difference between responses to positive and negative stimuli with a .95 confidence interval equal to (-44.85, - 1.20). Eyeblink responses to negative stimuli were significantly larger than those to positive stimuli at 4500 ms. There were no significant differences between positive blink responses and negative blink responses for any group at 300 ms or 800 ms at the .05 level. However, the High-Low group did show significant emotional modulation at 4500 ms, yielding a .95 confidence interval equal to (-121.85, - 1.05). The High-High group, however, did not show any emotional modulation (not even trends) at 4500 ms, yielding a .95 confidence interval equal to (-36.38, 35.20) (See Figure 5). Neither the Low-High nor the Low-Low group displayed significant emotional modulation at 4500 ms, although the trimmed means do fall in the expected direction (See Table 3). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39 ® Ftositive o Negative 60 ■ 50 ■ t — I 40 ■ 30 - % change m 20 ■ 10 - -10 ■ High F2 /Low F1 High F2 /High F1 Low F2 /Low F1 Low F2 /High F1 •d e n o te s s ig n ific a n t d iffe re n c e p ^ .0 5 Figure 4: PCL-defined Groups at 4500 ms Lead Interval For the SPQ defined groups, as in the PCL defined group tests, there were no significant differences between positive blink responses and negative blink responses for either the High Negative Symptom or the Low Negative Symptom group at 300 ms or 800 ms at the .05 level. At 4500 ms, however, the High Negative Symptom group did show significant emotional modulation, yielding a .95 confidence interval equal to (-69.93, -6.91), while the Low Negative Symptom group did not, yielding a .95 confidence interval equal to (-32.74, 27.39) (See Figure 6). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40 % change (p V ) 40 35 30 - 25 - 20 15 - 10 - 5 - 0 -5 H -10 High N eg ativ e S y m p to m ■ Positive □ Negative Low N egative Sym ptom •denotes significant difference p i_.05 Figure 5: SPQ-defined Groups at 4500 ms Lead Interval Prepulse Inhibition The same series of tests as above were performed between all four groups at 300 ms to determine any group differences in the amount of prepulse inhibition exhibited, regardless of valence. The High-Low group was significantly different than the Low-Low group, yielding a .95 confidence interval equal to (-68.49, -3.93) (See Figure 7). There were no other significant differences on this variable, even when reexamining the subjects in terms of their SPQ Negative Symptom scores. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 41 (p V ) Ugh High Low Low F2/Low F2/High F2/Low F2/High F1 F1 F1 F1 -10 - -20 ■ -40 • -60 J •denotes significant difference p ;_.05 Figure 6: PCL-defined Groups at 300 ms Lead Interval Skin Conductance No significant differences in skin conductance response patterns were found either within groups in terms of positive - negative differences or between any of the PCL defined groups. Similarly, no significant differences were found within or between the SPQ defined groups Results Using Traditional Methods Descriptive Statistics When re-analyzing these data using traditional methods, no trimming was used so inadequate responsiveness was defined as having a mean baseline startle measuring less than one microvolt (|iV). As a result of using this new criteria, in the raw analyses there were 22 subjects Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 42 in the Low Antisocial/Low Emotional Detachment group. All other groups (including the SPQ defined groups) remained unaffected. For the outlier-identified and replaced analyses, this new criteria added one subject to the High Antisocial/High Emotional Detachment group. All other groups (including the SPQ-defined groups) were unaffected. Means and standard errors are listed in tables 1 1 through 18 for all groups’ startle eyeblink and SCOR measures. Tables 1 1 through 14 contain statistics from the raw analyses, whereas tables 15 and 18 contain statistics from outlier-identified and replaced analyses. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. 10 - Experimen 2 PCL-defined Group Means and Standart Errors for Percent Change in Startle Modulation (gV) - Raw Analyses Slide Valence/ Probe Position High-Low n=13 (mean) High-Low n= 1 3 (standard error) High-High N=12 (mean) High-High N=12 (standard error) Low-Low N=22 (mean) Low-Low N=22 (standard error Low-High N=ll (mean) Low-High N=11 (standard error) Positive 300 -49.23 13.28 -37.65 14.42 -28.88 12.06 -38.30 20.54 Negative 300 -53.10 11.85 -45.76 12.83 -23.15 8.60 -34.32 22.73 Positive 800 -32.00 11.82 -49.26 8.69 -21.63 11.37 9.25 24.87 Negative 800 1.07 29.51 -41.08 11.33 -5.75 11.04 -16.19 14.23 Positive 4500 -8.37 17.24 -10.98 14.60 1.87 13.19 -12.61 12.58 Negative 4500 41.11 27.05 -18.80 11.29 14.05 14.34 37.99 22.04 Table 1 1 - Experiment 2 PCL-defined Group Means and Standard Errors for Skin Conductance Orienting Response Amplitudes (pS) - Slide V alence High-Low n=13 (mean) High-Low' n=13 (standard error) High-High N=12 (mean) High-High N=12 (standard error) Low-Low N=21 (mean) Low'-Low N=21 (standard error Low-High N=11 (mean) Low-High N=ll (standard error) Positive 0.14 0.06 0.25 0.10 0.20 0.05 0.05 0.04 Negative 0.10 0.04 0.31 0.11 0.21 0.06 0.14 0.09 Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 12 - Experiment 2 SPQ-defined Group Means and Standard Errors for Percent Change in Startle Modulation (pV) - Raw Analyses ____________________________________________________________________ Slide Valence/ Probe Position High Negative Symptom N=25 (mean) High Negative Symptom N=25 (standard error) Low Negative Symptom N=32 (mean) Low Negative Symptom N=32 (standard error) Positive 300 -32.10 11,27 -39.58 9.69 Negative 300 -36.41 10.60 -36.09 8.71 Positive 800 -25.43 11.92 -20.74 9.59 Negative 800 1.15 15.94 -23.95 9.11 Positive 4500 -11.05 8.76 -0.63 11.25 Negative 4500 25.33 10.96 12.20 15.32 Table 13 - Experiment 2 SPQ-defined Group Means and Standard Errors for Skin Conductance Orienting Response Amplitudes (pS) Slide Valence High Negative Symptom N=25 (mean) High Negative Symptom N=25 (standard error) Low ' Negative Symptom N=32 (mean) Low Negative Symptom N=32 (standard error) Positive 0.17 0.06 0.18 0.04 Negative 0.19 0.06 0.22 0.05 £ Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 14 - Experiment 2 PCL-defined Group Means and Standard Errors for Percent Change in Startle Modulation (pV) - Outlier- Slide Valence/ Probe Position High-Low n=13 (mean) High-Low n=13 (standard error) High-High N=13 (mean) High-High N=13 (standard error) Low-Low N=21 (mean) Low-Low N=21 (standard error Low-High N= 1 1 (mean) Low-High N=11 (standard error) Positive 300 -49.25 13.28 -38.26 13.24 -34.13 9.47 -38.09 20.53 Negative 300 -53.09 11.86 -49.06 12.26 -23.55 8.83 -42.61 14.91 Positive 800 -32.00 11.82 -49.77 8.00 -20.18 11.79 -2.72 22.52 Negative 800 -14.27 20.70 -37.00 11.07 -4.41 11.65 -14.88 14.03 Positive 4500 -8.35 17.24 -6.11 14,06 2.11 13.71 -10.76 12.56 Negative 4500 41.11 27.04 -17.62 10.30 13.92 14.94 41.84 22.75 Table 15 - Experiment 2 SPQ-defined Group Means and Standard Errors for Percent Change in Startle Modulation (pV) - Outlier- Slide Valence/ Probe Position High Negative Symptom N=25 (mean) High Negative Symptom N=25 (standard error) Low Negative Symptom N=32 (mean) Low Negative Symptom N=32 (standard error) Positive 300 -31.47 11.21 -43.97 7.88 Negative 300 -39.54 7.60 -38.79 8.78 Positive 800 -24.41 11.83 . -25.76 8.59 Negative 800 -4.80 11.90 -23.64 9.11 Positive 4500 -9.39 8.64 0.46 11.34 Negative 4500 28.40 11.24 10.48 15.26 4 ^ L /l 46 Emotional Valence and the Startle Response First, for the PCL-defined groups, a 4 (Group: High Antisocial/Low Emotional Detachment. High Antisocial/High Emotional Detachment, Low Antisocial/Low Emotional Detachment/ Low Antisocial/Low Emotional Detachment) X 2 (Valence: Positive, Negative) X 3 (Lead Interval: 300 ms, 800 ms, 4500 ms) analysis of variance (ANOVA) with repeated measures was performed on the raw percent change scores and also on the outlier-identified percent change scores. The raw analysis revealed a significant main effect of valence, F (1,54) = 5.22, q = 0.0262, and a significant main effect of lead interval, F (2, 108) = 18.16, p < 0.001. The outlier-identified analysis also revealed a main effect of valence, F (1.54) = 5 .02. p = 0 .0292. and a significant main effect of lead interval, F (2, 108) = 24.49, p < 0.001. Next, for the SPQ-defined groups, a 2 (Group: Low Negative Symptom, High Negative Symptom) X 2 (Valence: Positive, Negative) X 3 (Lead Interval: 300 ms. 800 ms, 4500 ms) analysis of variance (ANOVA) with repeated measures was performed on raw and outlier-identified percent change scores. The raw analysis revealed a significant main effect of valence, F (1, 55) = 0.0155, p = 0.0155. and a significant main effect of lead interval, F (2, 110)= 16.79, p < 0.001. For the outlier-identified analysis, there was also a significant main effect of valence, F (1. 55) = 5.55. p = 0.0221, in addition to a significant main effect of lead interval, F (2, 110) = 22.25, p < 0.0001. Subjects were exhibiting the largest startle responses to the negative stimuli probed at 4500 ms. There was no main effect of group and no significant interactions. As in Experiment 1 , t-tests were performed on a within group basis at each lead interval. For the raw data, there was an overall (all subjects tested) significant difference Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 47 between responses to the positive and negative stimuli probed at 4500 ms. T (57) = -2.17. p = 0.0344. For the outlier-identified analysis, the combined group difference at 4500 ms was also significant, T (57) = -2.06. g = 0.0444. There were no significant effects found at any lead interval for any of the PCL-defined groups. For the SPQ-defined groups, both the raw and outlier-identified analyses revealed that the High Negative symptom group was showing significant emotional modulation at 4500 ms: Raw, T (25) - -3.23, g = 0.0035; Outlier-identified, T (25) = -3.29, g = 0.003. The Low Negative symptom group did not display significant emotional modulation at 4500 ms in either analyses. Prepulse inhibition. There were no significant between-group differences in the amount of overall prepulse inhibition (regardless of valence) at 300 ms. for either the raw or outlier- identified analyses. Skin Conductance For the PCL-defined groups, a 4 (Group: High Antisocial/Low Emotional Detachment, High Antisocial/High Emotional Detachment, Low Antisocial/Low Emotional Detachment/ Low Antisocial/Low Emotional Detachment) X 2 (Valence: Positive, Negative) analysis of variance (ANOVA) with repeated measures was performed on SCOR data. No outliers were detected, so only one analyses was performed No significant effects were found. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48 Likewise, for the SPQ-defined groups, a 2 (Group: Low Negative Symptom. High Negative Symptom) X 2 (Valence: Positive, Negative) analysis of variance (ANOVA) with repeated measures was performed on SCOR data. No outliers were detected, so only one analyses was performed. No significant effects were found. Interpreting the Different Analyses As in Experiment 1 , results from the traditional analyses, in some instances, varied slightly from those of the robust analyses. The traditional raw and outlier-identified analyses both failed to reveal the PCL-defined group differences uncovered by the robust analyses, although the SPQ-defined group differences at 4500 ms were significant in all analyses. Once again, for reasons defined in Experiment 1 . the results from the robust analyses will be focused upon in the discussion section. Discussion Whereas Experiment 1 failed, this second experiment succeeded in replicating the overall emotional modification of startle effect with the IAPS picture paradigm. The most obvious explanation for this discrepancy is the difference in sample sizes between the college student sample and the community sample. There were about twice as many subjects in the community sample than in the college sample. Also, this paradigm was slightly different from the traditional paradigm because it included only positive and negative pictures, no neutral pictures. Other studies using the three-picture paradigm usually use response to neutral stimuli as a baseline measure and record changes in response to positive and negative stimuli relative to those baseline measurements. In this experiment, the differences between responses to positive and negative stimuli were considered most relevant. It may be that this method of measuring emotional modification Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 49 of startle is more sensitive to power problems. Another source of difficulty with power was inherent in this study due to the use of median splits in defining groups. Ideally, only extreme scorers on the PCL and SPQ should be compared to each other to maximize power, however a much larger sample would need to be run, initially to generate enough extreme scorers to compare. It is firmly believed that by applying robust techniques of analyses, the most was made of the samples on hand. The Psychopathy Checklist and Affective Modulation of Startle Patrick et al. (1993) first examined their incarcerated subjects in three groups based on total PCL scores using cutoffs recommended by Hare (1991): (1) Non psychopaths, subjects with total PCL scores < 20. (2) Mixed Psychopaths, subjects with total PCL scores between 20 and 30, and (3) True Psychopaths, subjects with PCL scores > 30. This examination resulted in only marginally significant group differences, although when examined separately, the non-psychopaths significantly showed the greatest amount of affective modulation, followed by the mixed group (non-significant, but showing trends), followed by the true psychopath group who did not even show trends. Upon reasoning that it was most likely the emotional deficits suffered by psychopaths that would affect the emotional modulation patterns, Patrick et al. decided to reexamine their subjects based on the PCL factor scores. It was then that strong group effects were found. Patrick et al. (1993) found that incarcerated subjects scoring high on the antisociality factor of the PCL (Factor 2) and high on the emotional detachment factor of the PCL (Factor 1) did not show affective modulation of startle; that is they showed no differences in response to positive and negative stimuli. They also found that subjects scoring high on the antisociality factor of the PCL (Factor 2) but low on the emotional Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50 detachment factor of the PCL (Factor 1) did show affective modulation. Patrick et al. noted that these highly antisocial groups primarily consisted of true and mixed psychopaths. The present study consisted of non-incarcerated community volunteers Thus, as one might expea, overall PCL scores for this sample were lower than those in Patrick et al.’s (1993) study. Nevertheless, the same group differences were found for those high in antisociality even though there was only one true psychopath in this whole group. This finding suggests that a small subgroup of the general population qualifying as mixed or partial psychopaths may to a great extent share emotional deficits more commonly seen in true psychopaths. Of more importance to the present investigation is Raine and Venables' (1992) earlier finding that these medium scorers on the PCL are very often characterized by higher scores on DSM III Borderline and Schizotypal Personality Disorder measures. Thus, it may be this group of mixed or partial psychopaths who would be most valuable to study in the context of schizophrenia research. Failure to find significant emotional modulation in the low antisocial groups (primarily non-psychopaths) was disconcerting. Nevertheless, valence differences are clearly visible in these group's affective modulation scores suggesting that perhaps there simply was not enough statistical power as a result of sample size. An alternative explanation may be that these low antisocial groups who were also lower in overall PCL scores may be slightly less emotionally responsive than the mixed psychopath group In other words, perhaps in terms of emotional responsivity, persons scoring at either extreme on the PCL are not capable of showing optimal modulation of startle patterns. On the other hand, it may simply be that the PCL factors were not a relevant way to split up a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 51 group of non-psychopaths. The presence of other psychological disturbances (specifically negative schizotypal symptoms) may have held more relevance for this low antisocial group. The SPO Negative Symptom Factor and Affective Modulation of Startle As in Experiment 1, and again contrary to what was predicted, subjects scoring high on the SPQ negative symptom factor displayed significant levels of emotional modulation of startle while subjects scoring low on this scale did not. This now significant replication of the trend uncovered in Experiment 1 requires that this finding be seriously contemplated. Once again, this finding might perhaps be attributed to the two "fearful state" traits loading onto this factor (suspiciousness and excessive social anxiety). Subjects scoring high on this Negative Symptom factor may have higher states of fear and anxiety than those scoring low making them more emotionally responsive to the type of stimuli presented in this study.1 9 On the other and, there is a clear disjunction here between verbal and physiological emotional expression, suggesting that schizotypals may be "feeling" emotion, just not really understanding what they feel or how to express themselves. It may follow that those scoring low on this Negative Symptom scale, thus indicating “normal" emotional expression, who do not show the physiological expression of emotion (as measured through the startle reflex), may also be showing this strange rift between verbal and physiological emotional expression. With respect to schizophrenia etiology and emotion research in pathological groups, this finding may prove to be highly significant. 1 9 As in Experiment 1. post hoc correlations performed between the traits loading onto SPQ Factor 2 and Startle modification scores were not significant. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52 Prepulse Inhibition The only significant group differences found in the amounts of prepulse inhibition at 300 ms were between the group of subjects high in antisociality/low in emotional detachment (High-Low) and the group low in both antisociality and emotional detachment (Low-Low). Of the four PCL defined groups, the High-Low groups showed the most prepulse inhibition at 300 ms whereas the Low-Low group showed the least. The other two groups fell in between. This finding is contrary to what was predicted. Because no directed-attention task was specified for the subjects in this study, these prepulse inhibition data are difficult to interpret. However, one explanation may be that the Low-Low group was less interested or less vigilant to the picture slides being shown. As alluded to earlier, it may be that subjects scoring at either extreme on the PCL suffer from some deficit preventing them from showing optimal startle patterns. Whereas later probed trials may reveal the emotional aspect of this deficit, early probed trials may be more apt to reveal the cognitive aspect of this deficit. Skin Conductance There were no group differences found in skin conductance amplitudes In fact, all subjects in this second experiment seemed rather non-responsive in terms of skin conductance. It seems unusual that this sample should have such a high rate of reduced orientating. The usual rate of nonresponsivity in a sample is about 25 percent. This rate appears to remain constant over several age groups (Venables & Mitchell. 1996). however a higher rate of electrodermal nonresponding is often found among schizophrenia patients (about 40-50 percent, Raine & Lencz, 1995). It is also worth noting some major differences between Experiment 1 and Experiment 2 that may be relevant to electrodermal Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 53 responding. Whereas the sample in Experiment 1 was mostly comprised of females around the age of 18 years, the sample in Experiment 2 was mostly comprised of males around the age of 30 years. A study by Venables and Mitchell (1996) found that in a group of subjects aged 20 to 29 years, females displayed larger skin conductance responses than males. Although in older groups of subjects this pattern appears to be reversed, their finding for this specific age group seems relevant here. These subjects in Experiment 2 are not the typical college students used in most studies, nor are they severely debilitated psychiatric patients. These are people not steadily employed at the time of this study but seemingly functioning adequately in society However, other demographic factors should be considered in conjunction with the data discussed here such as, age, gender, and the presence or absence of drug use. They may have a higher occurrence of psychopathology, as well as a broader range of psychological problems or disorders, compared to the college student populations usually studied. Subsequent analyses of these data should examine demographic data, structured clinical interview diagnosis, as well as PCL and SPQ scores with different physiological response patterns to affective stimuli to get a more in-depth picture of how these stimuli are processed. Conclusions Both psychopathy and schizotypy are psychological conditions demarcated by emotional dysfunction, specifically, a deficit in the experience of emotion. The present study attempted to clarify the nature of these deficits bv comparing their emotional modification of startle patterns. The results indicate that the emotional deficits suffered by these persons may manifest themselves in very different ways, but also may be somewhat Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 54 similar at their core. Whereas it has been well known for some time that psychopaths' verbal and actual experience of emotion differ, and whereas Patrick et al. (1993) and Levenston et al. (1996) extended this clinically observed knowledge to the physiological experience of emotion through the measurement of startle eyebiink modification, the present study appears to have revealed a similar (albeit reversed) pattern of disjunction between verbal and physiological emotional expression in negative schizotypals. As mentioned in the introduction, there appears to be evidence that both schizotypal persons and those suffering from antisocial behavior disorder may be classified as groups at high risk for schizophrenia. To theorize further, perhaps schizotvpv and antisocial personality disorder reflect different stages of dormancy for a genetic predisposition to schizophrenia. It may be that some latent factors in schizotypals have been activated in antisocials. Thus, in many ways, those suffering from antisocial personality disorder do not function as well in day-to-day life as many schizotypals appear to be able to do. At any rate, if it proves nothing else, the present study proves that more research with both schizotypals and antisocials is merited in association with schizophrenia research. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 55 References American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington DC: Author. American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed. Revised). Washington DC: Author. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington DC: Author. Arndt, S., Alliger, R. J., & Andreasen, N. C. (1991). The distinction of positive and negative symptoms: The failure of two-dimensional model. British Journal of Psychiatry. 158. 317-322. Cadenhead, K. S., Geyer, M. A., & BrafF, D. L. (1993). Impaired startle prepulse inhibition and habituation in patients with schizotypal personality disorder. American Journal of Psychiatry. 150. 1862-1867. Cleckley, H. (1976). The Mask of Sanity (5th ed.). Saint Louis, MO : Mosby. Cook, E. W„ III, Hawk, L. W„ Jr., Davis, T. L„ & Stevenson, V. E. (1991). Affective individual differences and startle reflex modulation. Journal of Abnormal Psychology. 100. 5-13. Cook, E. W., III. (in press). Affective individual differences, psychopathology, and startle reflex modulation. In M. E. Dawson, A. Schell, & A. Bohmelt (Eds ). Startle Modification: Implications For Neuroscience. Cognitive Science and Clinical Science. Cambridge University Press. Cuthbert, B. N., Bradley, M. M., & Lang, P. J. (1996). Probing picture perception. Activation and emotion. Psychophysiology. 33. 103- 111. Davis, M. (1996). Fear-potentiated startle in the study of animal and human emotion. In R. D. Kavanaugh, B. Zimmerberg, & S. Fein (Eds.) Emotion: Interdisciplinary Perspectives, (pp. 61-89). Mahwah, NJ: Lawrence Erlbaum. Dawson, M. E., Filion, D. L., & Schell, A. M. (1989). Is elicitation of the autonomic orienting response associated with allocation of processing resources? Psychophysiology. 26. 560-572. Dawson, M. E., Hazlett, E. A., Filion, D. L., Nuechterlein, K. H., & Schell, A. M. (1993). Attention and schizophrenia: Impaired modulation of the startle reflex. Journal of Abnormal Psychology. 102. 633-641. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 56 Dawson, M. E., Schell, A. M., & Filion, D. L. (1990). The electrodermal response system. In J. T. Cacioppo & L. G. Tassinary (Eds.), Principles of psychophysiology (pp. 295-324). Cambridge: Cambridge University Press. Dawson, M. E., Schell, A. M., Hazlett, E. A., Filion, D. L., & Nuechterlein, K. H. (1995). Attention, startle eyeblink modification, and psychosis-proneness. In A. Raine, T. Lencz, & S. A. Mednick (Eds.), Schizotypal Personality (pp. 250-271). Cambridge: Cambridge University Press. Filion, D.L., Dawson, M.E., & Schell, A.M. (1993) Modification of the acoustic startle-reflex eyeblink: A tool for investigating early and late attentional processes. Biological Psychiatry. 35. 185-200. Filion, D.L., Dawson, M.E., & Schell, A.M. (1994). Probing the orienting response with startle modification and secondary reaction time. Psychophysiology. 31. 68-78. Fischer, K. W., Shaver, P. R., and Carochan, P. (1990). How emotions develop and how they organize development. Cognition and Emotion. 4* 81 -127. Graham, F. K. (1980). Control of reflex blink excitability. In R. F. Thompson, L. H. Hicks, & V. B. Shryrkov (Eds.), Neural Mechanisms of Goal-directed Behavior and Learning (pp. 511-517). New York: Academic Press. Hampel, F. R., Ronchetti, E. M., Rousseuw, P. J., & Stahel, W. A. (1986). Robust Statistics: The Approach Based on Influence Functions. New York: Wiley. Hare, R. D. (1991). The Hare Psychopathy Checklist - Revised. Toronto: Multi- Health Systems. Hare, R. D., Cox, D. N., & Hart, S. D. (1995). Manual for the Psychopathy Checklist: Screening Version. Toronto. Multi-Health Systems. Harpur, T. J., Hare, R. D., & Hakistan, A. R. (1989). Two-factor conceptualization of psychopathy: Construct validity and assessment implications. Psychological Assessment: A Journal of Consulting and Clinical Psychology. 1 . 6-17. Heston, L. L. (1966). Psychiatric disorders in foster-home reared children of schizophrenic mothers. British Journal of Psychiatry. 112. 819-825. Huber, P. (1981). Robust Statistics. New York: Wiley. Ingraham, L. J. (1995). Family-genetic research and schizotypal personality. In A. Raine, T. Lencz, & S. A. Mednick (Eds.), Schizotypal Personality (pp. 19-42). Cambridge: Cambridge University Press. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 57 Izard, C. E. (1972). Patterns of emotions: A new analysis of anxiety and depression. New York: Academic Press. Izard, C. E. & Tomkins, S. S. (1966). Affect and behavior: Anxiety as a negative affect. In C. D. Spielberger (Ed.), Anxiety and behavior. New York: Academic Press. Kring, A. M. & Neale, J. M. (1996). Do schizophrenic patients sow a disjunctive relationship among expressive, experiential, and psychophysiological components of emotion? Journal of Abnormal Psychology. 105. 249-257. Lang, P. J. (1995). The emotion probe: Studies of motivation and attention. American Psychologist. 50. 372-385. Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1990). Emotion, attention, and the startle reflex. Psychological Review. 97. 377-395. Lang, P. J., Ohman, A., & Vaitl, D. (1988). The international affective picture system. Gainsville, FL: Center for Research in Psychophysiology, University of Florida. Levenston, G. K., Patrick, C. J., Bradley, M. M., & Lang, P. J. (1996, October). Psychopathy and startle modulation during affective picture processing: A replication and extension. Poster session presented at the annual meeting of the Society for Psychophysiological Research, Vancouver, British Columbia. Patrick, C. J., Bradley, M. M., & Lang, P. J. (1993). Emotion in the criminal psychopath: Startle reflex modulation. Journal of abnormal psychology. 102. 82-92. Patrick, C. J. (1994). Emotion and psychopathy: startling new insights. Psychophysiology. 31. 319-330. Raine, A. (1988). Antisocial behavior and social psychophysiology. In H. L. Wagner (Ed.), Social psychophysiology and emotion: Theory and clinical applications. (pp. 231-250). New York: John Wiley and Sons. Raine, A. (1991). The schizotypal personality questionnaire (SPQ): A scale for the assessment of schizotypal personality based on DSM-III-R criteria. Schizophrenia Bulletin. 17. 555-564. Raine, A. & Lencz, T. (1995). Conceptual and theoretical issues in schizotypal personality research. In A. Raine, T. Lencz, & S. A. Mednick ( Eds.), Schizotypal Personality (pp. 3-15). New York: Cambridge University Press. Raine, A., Lencz, T., & Benishay, D. S. (1995). Schizotypal personality and skin conductance orienting. In A. Raine, T. Lencz, & S. A. Mednick ( Eds ), Schizotypal Personality (pp. 219-249). New York: Cambridge University Press. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 58 Raine, A., Lencz T., & Mednick, S. A. (Eds.). (1995). Preface. In A. Raine, T. Lencz & S. A. Mednick ( Eds.), Schizotypal Personality (pp. xi - xiii). New York: Cambridge University Press. Raine, A., Reynolds, C., Lencz T., Scerbo, A., Triphon, N., & Kim, D. (1994). Cognitive-perceptual, interpersonal, and disorganized features of schizotypal personality. Schizophrenia Bulletin. 20. 191-201. Raine, A. and P. H. Venables (1984). Electrodermal nonresponding, antisocial behavior, and schizoid tendencies in adolescents. Psychophysiology. 21: 424-433. Raine, A. & Venables, P. H. (1992). Antisocial behaviour: Evolution, genetics, neuropsychology, and psychophysiology. In A. Gale & M. W. Eysenck (Eds ), Handbook of individual differences: Biological perspectives, (pp. 287-321). New York: John Wiley and Sons. Schell, A. M., Dawson, M. E., Hazlett, E. A., & Filion, D. L. (1995). Attentional modulation of startle in psychosis prone college students. Psychophysiology, 32, 266-273. Siever, L. J. (1995). Brain Structure/function and the dopamine system in schizotypal personality disorder. In A. Raine, T. Lencz & S. A. Mednick ( Eds.), Schizotypal Personality (pp. 272-286). New York: Cambridge University Press. Siever, L. J., Keefe, R., Bernstein, D. P., Coccaro, E. F„ Klar, H. M., Zemishlany, Z., Peterson, A. E., Davidson, M., Mahon, T., Horvath, T., & Mohs, R. (1990). Eye tracking impairment in clinically identified patients with schizotypal personality disorder. American Journal of Psychiatry. 147. 740-745. Silverton, L. (1988). The genetic relationship between antisocial behavior and schizophrenia. A review of the literature. In W. Buikhuisen (Eds.), Explaining Crime. Leiden: Brill. Tomkins, S. S. (1962). Affect imagery consciousness. The positive affects (Vol. 1). New York: Springer. Tomkins, S. S. (1970). Affect as the primary motivational system. In M. Arnold (Ed ), Feelings and emotions. New York: Academic Press. Venables, P. H., & Mitchell, D. A. (1996). The effects of age, sex, and time of testing on skin conductance activity. Biological Psychology, 43(2), 87-101. Wilcox, R. R. (1996). Statistics for the Social Sciences. San Diego, CA: Academic Press. Wilcox, R. R. (1997). Introduction to Robust Estimation and Hypothesis Testing. San Diego, CA: Academic Press. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 59 Wilcox, R. R. (in press). How many discoveries have been lost by using standard statistical methods and ignoring modem techniques? American Psychologist. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60 APPENDIX A: Slide Order Error in Slide Order 1 Toward the end of data collection, a slide order error was discovered in slide order 1 . Two slides (one positive, one negative) were interchanged in the slide tray. The result of this error was that for every subject run under Slide Order 1, one trial block would have four positive probed slides and only two negative probed slides. The remaining unprobed slides were both negative. The probe position affected depended on which of the three lead interval sequences was assigned. All three probe positions were equally affected. Thus for those affected subjects, instead of having only three positive and three negative eyeblink scores at each lead interval, one probe position would have four positive scores and two negative scores. For all data analyses, averages were computed for each lead interval and valence type across all four blocks. Therefore, for affected subjects the average was taken over four positive and two negative eyeblink scores for the affected probe position compared to only three of each for slide order 2 subjects. The distributions of these averaged eyeblink scores at each lead interval were compared and no significant differences were found. It was therefore concluded that the effect of the slide order error was minimal and did not have any significant bearing on the overall results of either experiment. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IMAGE EVALUATION TEST TARGET (Q A -3 ) 1 . 0 l.l 1 . 2 5 “ 13 13 s “ m t |fi 112.0 1 . 4 1 . 8 150mm A P P L IE D A IIW1GE . Inc j s s s , 1 6 5 3 E a s t M a i n S t r e e t - ■ R o c h e s t e r , N Y 1 4 6 0 9 U S A — P h o n e : 7 1 6 / 4 8 2 - 0 3 0 0 - = = ~ -= = F a x : 7 1 6 / 2 8 8 - 5 9 8 9 O 1 9 9 3 , A p p lied Im a g e , Inc.. All R ig h ts R e s e r v e d Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Hyperactive symptoms, cognitive functioning, and drinking habits
PDF
Affection and conflict in family relationships
PDF
Invariance to changes in contrast polarity in object and face recognition
PDF
Considering the impact of culture on the therapeutic relationship: A look at Latino school-age children in a special education setting
PDF
Effects of personal resource sufficiency on perceived difficulty and desirability of earthquake preparedness
PDF
Depression as a risk factor for dementia
PDF
Insights into the nature of phonological and surface dyslexia: Evidence from a novel word learning task
PDF
Depression and suicidality in Latino adolescents: A study of acculturation and gender role beliefs
PDF
Head injury and dementia: A co-twin control study of Swedish twins
PDF
History of depression, antidepressant treatment, and other psychiatric illness as risk factors for Alzheimer's disease in a twin sample
PDF
A typology of maritally violent men and correlates of violence
PDF
Depression in offspring following severe prenatal stress
PDF
Hedonic aspects of conditioned taste aversion in rats and humans
PDF
Evidence for spontaneous situational inferences
PDF
Expectancies for alternative behaviors predict drinking problems: A test of a cognitive reformulation of the matching law
PDF
Articulated thoughts about intentions to commit anti-gay hate crimes
PDF
Hispanic culture, acculturation, and distress among caregivers of dementia patients
PDF
Asymmetries in the bidirectional associative strengths between events in cue competition for causes and effects
PDF
Explanations, the availability heuristic, and biases of subjective likelihood in economics
PDF
Applying the theory of reasoned action to condom use: The effect of immediate consequences on intention-behavior consistency
Asset Metadata
Creator
Mejia, Veronica Yvette
(author)
Core Title
An examination of affective modulation, psychopathy, and negative schizotypy in college and community samples
School
Graduate School
Degree
Master of Arts
Degree Program
Psychology
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
OAI-PMH Harvest,Psychology, clinical,psychology, cognitive,psychology, physiological
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
Dawson, Michael E. (
committee chair
), Schell, Anne M. (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c16-20204
Unique identifier
UC11337853
Identifier
1390004.pdf (filename),usctheses-c16-20204 (legacy record id)
Legacy Identifier
1390004.pdf
Dmrecord
20204
Document Type
Thesis
Rights
Mejia, Veronica Yvette
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
psychology, cognitive
psychology, physiological