THE TWA CT CF FULL AND CECEPTTVE DISCLCS URES CN ANXTETY AND AF FECTTVE SETTSTTWLT‘T’ I ”ms for the Degree of Ph. D. “3‘ {W STATE U. T 3.3: TY :TTICTZN Ul.';i;R:.l&J: SC LulAVfiTER 1975 This is to certify that the thesis entitled THE IMPACT OF FULL AND DECEPTIVE DISCLOSURES ON ANXIETY AND AFFECTIVE SENSITIVITY presented by IRWIN LAWRENCE SCHACHTER has been accepted towards fulfillment of the requirements for PH . D . . PSYCHOLOGY degree in AA :3th ALLL; I V Major professor Date 5‘2k75 0-7639 " named a; T .‘ WAG 8: SONS' 800! mm M.- ABSTRACT THE IMPACT OF FULL AND DECEPTIVE DISCLOSURES ON ANXIETY AND AFFECTIVE SENSITIVTIY By Irwin Lawrence Schachter The adoption of new ethical guidelines for psychological research with human subjects has caused considerable discussion and controversy by advocating limitations on the use of deception in experiments with humans. Yet some researchers insist that deception is necessary for the conduct of certain experiments, especially in the area of social psychology. The present study examines the impact of deceptive and non- deceptive feedback. The subjects consisted of both high and low trait anxious college students who received either deceptive or non-deceptive feedback within an experiment examining the relationship between trait anxiety and affective sensitivity. Subjects were administered the Spielberger State Anxiety Inventorz_followed by his Trait Anxiety Inventory and were then divided into high and low trait anxious groups. Subjects in each group then received either deceptive or non-deceptive feedback based on their trait anxiety scores. Non-deceptive (full disclosure) feedback consisted of a paragraph describing personality character— istics congruent with the subject's trait anxiety score. Deceptive feedback consisted of a paragraph describing personality character- istics which were opposite to those suggested by the subject's trait anxiety score. Following treatment, subjects were administered a second state anxiety test which was followed by the Affective Sensitivity Scale. Dependent variables were (1) change in state anxiety scores from pre-treatment to post-treatment and (2) scores on the affective sensitivity scale. It was hypothesized that (1) deceived subjects will show increased state anxiety scores, relative to non-deceived subjects, (2) non-deceived subjects will score higher on the affective sensitivity scale than deceived subjects, and (3) low trait anxious subjects will score higher on affective sensitivity than high trait anxious subjects. Before debriefing, a post-experimental questionnaire was administered asking subjects whether or not they believed that they had been deceived. A number of deceptive disclosure subjects did not admit to being deceived and a number of full disclosure subjects felt they were deceived. It is difficult to assess the impact of this on the experiment. In this era of post-Watergate, the impact of subjects' beliefs about deception are indeed hard to assess and might warrent separate investigation. Asking subjects about deception might well be a red flag eliciting both rational and irrational responses. Therefore, the author chose to minimize the importance of the post-experimental questionnaire responses and proceed with the statistical analyses. Results showed that significant differences between treatment groups occurred for state anxiety change scores, and those changes were in the direction of the treatment content, regardless of the disclosure condition. When the tendency of state anxiety change scores to regress toward the mean was controlled, low trait anxious subjects had significantly lowered their state anxiety, regardless of disclosure. Results also showed a trend suggesting that low trait anxious females scored higher on the affective sensitivity scale than did high trait anxious females. It was concluded that the relatively "mild" deception used in this experiment was outweighed by the treatment content. In other words, when high trait anxious subjects read feedback based on their high trait anxiety score, they became more anxious; when they read feedback based on a low trait anxiety score, they became less anxious, relative to other groups. Low trait anxious subjects became more anxious when they read high trait anxiety feedback; they became less anxious after reading low trait anxiety feedback, relative to other groups. Thus, in the total population, state anxiety tended to change in the direction of the treatment content. However, when one controlled for regression toward the mean, low trait anxious subjects became less anxious with either high or low trait anxiety feedback. Since significant differences occurred which were not clearly attributable to either deceptive or full disclosure, or treatment content, it was suggested that more attention be paid to subjects' state anxiety and trait anxiety as a general factor in influencing experimental results. Approved: Committee Chairman Date: ‘Thesis Committee: Norman Abeles, Ph.D., Chairman Dozier Thornton, Ph.D. JOhn Wakeley, Ph.D GO rdon Williams , Ph . D. THE IMPACT OF FULL AND DECEPTIVE DISCLOSURES ON ANXIETY AND AFFECTIVE SENSITIVITY By Irwin Lawrence Schachter A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1976‘ DEDICATION To my parents, for their loving support and patience for all these years. To John, who opened my eyes to see reality and To Frank, who taught me to persevere in spite of it. -11- ACKNOWLEDGMENTS I offer my warmest appreciation to Dr. Norman Abeles, my thesis and guidance committee chairman. His steady encouragement, patience, and advice over the years helped me attain what was once only a distant dream. I offer a special thanks to Dr. John Wakeley, Dr. Gordon Williams, and Dr. Dozier Thornton. All three gave generously of their time and knowledge to me as members of this dissertation committee. I am also very grateful to Dale Jaslove for his competent assistance in the preparation and conducting of this experiment, and to Roger Allen for his statistical guidance and personal patience while under significant stress. -iii- TABLE OF CONTENTS Amom EWMENT S O O I O O O O O O O O O O O O I O O O O O O O 0 LIST OF TABLES O O O I O O O O O O O O O O O O O O O O O O 0 LIST OF MPENDICES O C O O O O O O O O O O O O O O O O O O O 0 Chapter I. II. III. IV. VI. INTRODUCTION 0 O O O O O O C O O O O I O O I O O O 0 REVIEW OF THE LITERATURE . . . . . . . . . . . . . Ethical issues in psychological research Deception Experimenter demand characteristics Alternatives to deception Subject-experimenter relationships and ethics Disclosure instructions as an independent variable Instructions and anxiety and performance Anxiety and its relation to empathy METHOD 0 O O O O O O I O O O O O O O O O O O O O O 0 Subjects Instruments Experimental procedure HYPomESEs . O O O O O O O I I O O I O I O I O O 0 RESULTS 0 O I O O O O O O O O O O O O O I O O O 0 O 0 Description of the data Tests of individual hypotheses DISCUSSION . . . . . . . . . . . . . . . . . . . . . B IBLImRAPm . O O O O O I O O O O O O O D I O O O O 0 O O O APPENDICES _iv- Page iii vii 40 51 53 65 73 82 Table 10. 11. 12. LIST OF TABLES Trait anxiety scores defined for this experiment as high trait anxious or low trait anxious for subject sex and university class. . . . . . . . . . . . . . . . . . . . . Means and standard deviations for 31, 82, and AS, for all subjects (n = 224). . . . . . . . . . Means and standard deviations and N's for trait anxiety, $1, $2, and AS, by treatment group (n B 224). o o c o o o s o o s o o o o o o 0 Subjects' post-experimental questionnaire responses within each group (n = 217). . . . . . . Means and N's for state anxiety difference scores for each treatment group (n a 224). . . . . . . Source table--difference scores for all subjects (n = 224). . . . . . . . . . . . . . . 81 means and N's for selected subjects (n a 75). . Means for difference scores, selected subjects (n g 75). O O O O I I O O O O O O O O O O O O O 0 Means and N's for difference scores, males (n a 52). O C C C O C O O O O O O O O O O O O O Source table--difference scores, males (n = 52). Means and N's for difference scores, females (n B 172). o o o o o o o o o o o o o o o o o o o a Source table--difference scores, females (n = 172). ~v—o Page 46 54 54 55 56 56 58 S8 59 60 61 61 13. 14. 15. Means and N's for AS scores, total group (n - 224). C C C C . O C O O O O O C Source table, AS scores, total group (n Means and N's for AS scores, females (n _vi_ 224) 172). 63 63 64 LIST OF APPENDICES Appendix Page A. State Anxiety Questionnaire One . . . . . . . . . . . 84 Trait Anxiety Questionnaire . . . . . . . . . . . . . 85 State Anxiety Questionnaire Two . . . . . . . . . . . 86 B. LADD Debriefing Information . . . . . . . . . . . . . 88 LAFD Debriefing Information . . . . . . . . . . . . . 89 HADD Debriefing Information . . . . . . . . . . . . . 90 HAFD Debriefing Information . . . . . . . . . . . . 91 Middle Trait Anxious Subjects' Debriefing Information 92 -vii- CHAPTER I INTRODUCTION Considerable debate and controversy is associated with the development and publication of the new Ethical Principles in the Conduct of Research with Human Participants (Ad hoc Committee, 1973) . The adoption of these ethical standards imposed new and significant pressure on psychological experimenters to eliminate deception and adopt a full disclosure policy with their subjects. This fosters a firm protection of subject dignity and respect not found in the older standards (A.P.A., 1953, 1963) which were more permissive toward deception, pain and partial disclosure instructions. Debate continues on whether these new standards are necessary or even advisable, especially since these ethical guidelines may have great impact on the nature of the experimental instructions, subject responses, and subject-experimenter relations. Resnick and Schwartz (1973) called for research studying the impact of these newly denveloped standards. In their verbal conditioning study they used bath standard and full disclosure instructions and found that full disclosure instructions based on the new standards were disruptive, fostered defensive subject behavior and produced data incongruent with existing literature. Gergen (1973) hypothesized that subjects were content under the previous ethical principles and he frankly felt that subjects did not care about ethics, research rationale or full disclosure. Sullivan and Deiker (1973) showed that psychologists were significantly stricter about ethical issues than were subjects and wondered if the new standards represented an overreaction to perceived public pressures. Other authors feel differently. Kelman (1965, 1967), Seeman (1969), and Baumrind (1971) argued that deception was unethical and fully informed consent essential whereas Stricker (1967) and Argyris (1968) felt that deception and partial disclosure instructions pro- duced suspicious subjects who resorted to adaptive strategies which led to invalid data. Orne (1962), Schultz (1969) and Carlson (1971) recommended more attention to demand characteristics, a reduction of experimenter control and fuller disclosure in instructions to enhance a subject status of a whole and dignified person who would supply more meaningful and valid information. The present study attempted experimentally to examine ethical issues as they related to disclosure variables, subject state and subject performance. Two groups composed of high or low trait anxious subjects received either full disclosure or deceptive dis- closure feedback. Its impact on subject state and performance was measured by a state anxiety questionnaire and an affective sensitivity scale. These dependent variables were chosen because no studies were found that directly examined the association between subject trait anxiety level and empathic ability and the state anxiety measure also permitted the examination of changes in subject anxiety as a result of disclosure variables. CHAPTER II REVIEW OF THE LITERATURE Ethical Issues in Psychological Research In 1953 the first ethical code developed by empirical methods was formulated and adopted by the American Psychological Association (A.P.A., 1953). Ten years later, after a decade of interest in ethics and the evolution of new issues and the reevaluation of old issues, it was revised (A.P.A., 1963). Then in 1973, following the continued proliferation of researchers and research instruments and the growing public and professional concern for individual privacy, dignity and well—being in the experimental situation, the Association created and adopted a new ethical code. By insisting that the experimenter have the informed consent of his subject before he proceeds with his work, the most recent ethical standards go far toward protecting the right of the individual to ‘make knowledgeable decisions about his experimental participation. However, this right, fully exercised, is often in direct conflict Twith science's basic function, to discover truth and to add to mankind's body of knowledge (Panel, 1967). The problem facing psychologists .and ethics committees is how to resolve this conflict. -4- Although Freund (1967) seriously questioned whether the law was ready to deal with the complexities and variables of human exper— imentation, the Ad hoc committee (1973) responsible for the new ethical standards had to try to deal effectively with the questions he raised. How can the law judge accurately whether consent was informed, freely given, and valid? When does informed consent interfere with blind or double-blind experiments? Do children, convicts, psychotics, or students really give consent freely, uncoerced and knowledgeably? These issues helped shape the ten principles that, along with relevant discussion and examples, make up the Ethical Principles in the Conduct of Research with Human Participants (A.P.A., 1973). Necessarily, the code is a compromise, an attempt to adjudicate between competing rights and values. What results from it, however, is not only an essential flexibility but also a vulnerability to evasion by researchers. Hence the formation in the scientific community of professional committees commissioned to rule on the ethical propriety of research proposals. These committees are concerned that any laxness on their part will bring public disapproval and arbitrary interference, as Ruebhausen and Brim (1965) had warned. The researcher's fear of community imposed standards has intensified only recently, wrote Smith (1967). In earlier years, he continued, when behavioral research was a smaller area, with few researchers doing few studies with private financial support, the scientist viewed responsibility for ethical actions as entirely his own concern. The scientist was inconspicuous, and few people challenged his methods or power (Smith, 1967). With the expansion of behavioral research came a steady increase in public funding and attention which resulted in growing pressures on the psychological community to face ethical issues. Some of these issues are extremely complex. For example, in research with children, the privacy and sensitivity of both the child and the parents must be carefully considered. Justifying the use of deception with children by citing a post-experimental debriefing is often unacceptable because many of the young subjects simply do not understand the debriefing no matter how carefully presented. Informed consent is certainly not possible with young subjects, but how informed or appropriate is parental or school consent? Who is, ultimately, protecting the children's rights and dignity without conflict of interest if schools or other social agencies act as consenters in loco parentis when running their own experiments (Smith,l967)?v Special ethical problems also occur in psychopharmacological research with human subjects. Beecher (1966) pointed out that in many drug studies most subjects would not have agreed to participate had i they been fully informed of potential drug effects. A decision to withhold an effective drug and offer a placebo or a decision to give an untested drug to patients is not easily made. Also, certain research= designs require that a subject he misinformed or deceived in order to permit the study of the effect of expectations and suggestions on his symptmms. There have even been studies where subjects were not told they had received any drug at all (Katz, 1967). The use of questionnaires raises ethical issues especially related to confidentiality and use of the information gathered. For example, a question asking if you have ever broken the law, or if you are of illegitimate birth, presents an invasion of privacy, an obvious ethical issue. Conrad (1967) formulated criteria for the evaluation of acceptability of items, but warned that the criteria cannot be applied routinely owing to the special circumstances of each situation and the nature of each subject pool. In each of the aforementioned "special" situations involving children as subjects, psychopharmacological research, or questionnaires, the same ethical concerns about subject privacy, the use of deception, informed consent, and experimenter consideration of subject welfare, recur, magnifying the difficulty of--and the need for caution in-- designing experiments and examining the resulting data. In the following sections these ethical concerns are more closely examined, particularly deception and demand characteristics within the research situation. Also discussed are suggestions for dealing with these problems and some of the criticism these suggestions have provoked. Deception Perhaps the thorniest of the ethical issues involves deception. Vinacke (1954) questioned the necessity of using deceit in social- psychological experimentation and wondered if other designs could be used to gather the same information. He asked if the standard post- experimental debriefings in deception experiments were really helpful in restoring the subject's self-respect and dignity and trust in the researcher and the research process. Vinacke suggested research studying these questions and asked the reader to question seriously whether or not the use of deception is even ethical. In a particularly controversial experiment using deception, Milgram (1963) ordered naive subjects to administer what the subjects believed to be electrical shocks as punishment to a "victim" (who was really an experimental accomplice) in the guise of a learning experiment. Although the subject would hear the "victim" scream in pain, the exper- imenter would order the continued administration of what the subject thought was increasingly large electrical shocks. In this obedience study the dependent variable was the maximum "shock" administered by the subject before he refused to cooperate. Extreme states of anxiety and tension such as sweating, stuttering, trembling and uncontrollable seizures starting from nervous laughter were common as subjects obeyed the experimenter and continued to administer stronger "shocks." The severity of stress observed in the subjects was unexpected by the experimenter but no attempt was made to halt the experiment, although Milgram reported that at the experiment's end "an effort was made to reduce tensions that arose as a result of the experiment." Baumrind (1964), in reflecting on Milgram's study, believed that' Milgram.was unconvincing in his casual statement of experimenter effort to reduce tensions, especially given the traumatic nature of the reactions. Baumrind also pointed out that in the act of volunteering the subject assumed implicitly an attitude of obedience and trust and thus the experimental situation was inappropriate to study obedience. Given the real possibility of long term negative after-effects on the subject and the lack of ecological validity of the study, she questioned the propriety of possibly inflicting harm on subjects in an experiment of questionable usefulness. In response to Baumrind, Milgram (1964) argued that it is precisely because the subject was obedient and trusting that the experiment did have ecological validity. He also clarified subject debriefing and follow-up procedures and reported that, although momentary stress was generated, dissipation was quick, no injurious after-effects were reported following debriefing or longer term follow-up, and most subjects felt fortunate to have been in the experiment. But how much stress is permissible in any given experiment? Kaufman (1967) wrote that this question cannot be answered without taking into consideration the potential gain in knowledge provided by the experiment and the effectiveness of post—experimental debriefing. He noted that the subjects' well being was maintained in the Milgram study and that insights into ourselves were achieved that few other studies have offered. Broader issues involving deception are apparent beyond the appropriateness of this particular controversial experiment. Kelman (1967) reviewed different types and degrees of deception and lamented that deception was a common norm in social-psychological experimentation. Seeman (1969) and Stricker (1967) also cited with alarm the large and growing use of deception as an acceptable research design, particularly in personality and social-psychological research. But deception not only occurred within the experiments; in some experimental situations it "encompass[ed] the whole definition of the relationship between the parties involved" (Kelman 1967). Thus the development of mutual trust between experimenter and subject seemed increasingly difficult. For example, in Rosenthal's work with experimenter effects (1963, 1964) his subjects thought they were experimenters. Although the only way to study this important area may have been through deception, it further establishes the importance of studying the whole experimental situation. Kelman (1967) pointed out the unfortunate but typical use of deception as a standard operating procedure rather than as a last resort for gaining significant information about important problems. He cited a few dramatic deception experiments, including studies where (a) male subjects were led to believe they were homosexually aroused by photographs of men (Bramel, 1962, 1963), and (b) a drug that induced temporary interruption of respiration was administered to subjects under the guise that it was connected with a possible therapy for alcoholism (Campbell, Sanderson & Laverty, 1964). Also mentioned were studies in which misleading information was given or sought in what subjects perceived as real life and not experimental situations. Kelman (1967) wondered apprehensively if we "tend to regard [the experimental -10- situation) as...not quite real, [pné] that can be isolated from the rest of life like a play performed on a stage, and to which, therefore, the usual criteria for ethical interpersonal conduct become irrelevant" (p. 5). Yet with the high incidence of deception, particularly in experiments dealing with conformity, cognitive dissonance, balance theory and decision making, there were few studies found that reported information about subjects' suspiciousness of the deceptions used (Stricker, 1967). Since Common sense would require the separation of data of subjects reporting suspicions from the data of unsuspicious subjects, the results of most studies using deception would be questionable. In addition, Orne (1962) pointed out that subjects would be unlikely to report their suspicions if they felt that the experimenter had intended them to be unaware of the deception, whereas Rosenthal (1963) wrote that experimenters would be reluctant to classify or exclude suspicious subjects. Clearly, many researchers have mistakingly taken for granted both the effectiveness of methods incorporating deception and the validity of the ensuing results, further accenting the need for guidelines to evaluate investigations of this sort (Stricker, 1967). Indeed, according to Stricker, experiments that used deception offered very questionable data for scientific use, and others (Orne, 1963, Argyris, 1968, Schultz” 1969) also agreed with him.that the experimenter may have been mistaken in assuming naiveté on the part of the subject. -11- A number of studies supported this point. In an experiment to assess the extent of subject suspiciousness about deceptions, Stricker et al. (1967) appraised the degree of subjects' suspicions on two conformity procedures, (a) a simulated group version of the Asch situation and (b) questionnaires with fictitious norms. Using 190 paid high school volunteer subjects, who were assumed to be more naive than college populations, the experimenters surprisingly found a high rate and generality of suspicion. Many subjects rightly suspected that the purpose of the experiment was to see if their responses would be influenced by other's responses and appropriately mistrusted the questionnaire norms. In a resulted study, Allen (1966) examined the effect of subjects' knowledge of deception in an Asch-type conformity situation in which others' responses, viewed by the subject, were controlled by the experimenter. Successfully deceived subjects were found to respond differently than subjects who reported suspicions in the post- experimental questionnaire, and both groups responded differently than controls who were tested individually without group norms. Rubin and Moore (1971) also conducted an experiment which studied dfittePtion and found differences between suspicious and nonsuspicious subjects. They found suspicious subjects had more self-esteem, more training in psychology and were more resistant to experimental influence. 1118th suspicious subjects with scores indicating high authoritarianism (F Scale) yielded to experimental demand characteristics while those -12- with low F scale scores were oppositional to perceived demands. Also alarming is the Silverman, Schulman, & Wiesenthal (1970) examination of the impact of deception and debriefing on subject performance in later experiments. Inexperienced subjects participated in either an experiment using deception or in a non-deceptive exper- iment. All subjects were then given a number of measures to assess differences. Deceived subjects differed from non—deceived subjects on all measures in that they seemed sensitized to possible ulterior motives of experimenters, had increased evaluation apprehensiveness, and tried to present themselves as strong and stable. These data further suggest that the results of deception experiments should be viewed with "suspicion." Both Seeman (1969) and Baumrind (1971) abhorred the use of deception, and argued for principles advocating fully informed consent by subjects and an open, trusting relationship between subjects and experimenters. Seeman cited the Supreme Court's Escobedo and Miranda decisions where confessions were set aside because individual rights were not adequately protected by legal advice. The Court ruled that the government must accord dignity and integrity to its citizens and that an individual had the right to remain silent unless he freely chose to speak. It was this principle of informed consent to an action affecting one's welfare that Seeman strictly applied to.research subjects. Since the outcome of a research project was embedded in the means used, he wrote, and since deception was not ethical, then the resulting data could not be valid. -13- Seeman, in fact, called deception "anti-therapy"--a systematic reversal of the major and valued therapeutic properties of trust, congruence, and reality. Deception experiments fostered distrust. The experimenter was not what he appeared to be, suggesting incongruence. The deceptive experimenter reversed reality, distorting it, and asked subjects to respond to the distortion as if it were real. The subject's adoption of a reality distortion as a basis for action was an exact reversal of therapeutic goals. Moreover, although "debriefing" usually occurs after deception experiments, subjects may still be confused as to when the deception occurred and will have difficulty disentangling the truth from the falsity of the relationship. Seeman concluded that the goal of science should not be knowledge, raw and feelingless, but rather wisdom. "Knowledge for what" must be asked. If, within psychology, our knowledge is attained at the expense of another's unique humanness, maybe the cost is too high. Experimenter Demand Characteristics Many experimenters are not offensively deceptive, but are criticized for other reasons. For example, Orne (1962) argued against treating the subject as a stimulus-response machine or empty automaton. He viewed the psychological experiment as a special type of social inter- action and focused on certain non-experimental variables which he classified as the demand characteristics of the experiment. These in- cluded the subject's tendency to place himself largely under the control of the experimenter and his desire to be a "good subject" who would -14- validate the experimenter's hypothesis and contribute to science. Thus not only the intended experimental variables but rather the totality of cues available which convey the experimental hypothesis to the subject would be important determinants of behavior. The extent that the subject responded to these demand characteristics rather than experimental variables would influence the validity of the results. Orne (1962) reviewed studies that demonstrated this phenomenon and suggested ways to minimize and control it, intending to gain greater replicability and ecological validity. Partial confirmation of Orne's hypothesis of good subjects responding to demand characteristics was offered by Holmes(l967), who investigated the relationship of demand characteristics to subject experience level and found that experienced subjects performed differently than naive subjects. Using a number of experiments over time, he concluded that experienced subjects were more likely to become aware of the reinforcement contingency in a verbal re- inforcement experiment and attempt to be increasingly cooperative, that is, be an "even better" subject. But beyond a certain experience level, subjects stopped trying to determine what the experiments were about while remaining good subjects. It seemed that the "good subjects" got "better" with experience and awareness, at least for a while. Argyris (1968) drew analogies between experimental demand charac- teristics and the theory of organizations. In particular, he pointed out how subjects came to experiments expecting to be tricked and tried -15- to guess the deception or "beat the management." But if the subject liked the experimenter, he might have cooperated, like a good worker in an organization. Argyris further suggested that the adaptive strategies prescribed by organizational theory may also apply in the experimental situation because the relationship between the subject and the researcher is so similar to the relationship between employee and employer. If he was right, then the subject's adaptive strategies might lead to internal psychological states that would alter both his perception of the experiment and his responses to it. Can the experimenter "control" the subject's adaptive strategies as the employer tries to control his employees? Argyris believed the researcher could control the subject's dependence, submissiveness, and short time perspective so that these factors would not result in adaptive strategies for subjects and problems for experimenters. This could be accomplished, he suggested, by reducing experimenter control and providing subjects with greater influence and involvement, thus reducing the subjects' need for strategies which interfere with information gathering. He believed that if researchers showed by thought and action their belief that subjects were trustworthy and responsible, a step would be taken in developing more appropriate research designs and in attracting more meaningful subject data. But increased subject influence and involvement may still lead to a "good subject" who would try to please the researcher by confirming the assumed hypothesis. If some contamination was inevitable, "under -l6- what conditions can the researcher have the greatest awareness of, and control over, the degree of contamination?" asked Argyris. Cook, Bean, Calder, Frey, Krovetz, & Reisman (1970), interested in.Argyris' questions about subject control and influence, designed two experiments to evaluate the impact of deception and demand characteristics on subjects in an attempt to ascertain the relevance or generalizability of three conceptualizations of subject type. .These three types were (a) the "good subject" who tries to validate the experimental hypothesis by fully cooperating with the experimenter towards its success (Orne, 1962), (b) the faithful subject who only follows instructions and does not seek to ascertain the variables and hypothesis (Fillenbaum, 1966), and (c) the negativistic subject who does not want to be controlled and therefore tries to figure out what is expected of him so he can do the opposite (Masling, 1966). The results were confusing, and the authors felt it was naive to think that in some experiments subjects adopt exclusively one subject response style over another. They found it useful to instead ask "which kinds of deception and demand characteristics inTwhich settings with which justification elicit which subject role in.which persons?" In an attempt to answer this question Weber and Cook (1972) conducted a comprehensive review of relevant data concerned with subject roles and demand characteristics. They concluded that most evidence about subject role was confounded, that most subjects were apprehensive about performance evaluation, and performance was biased -17- when subjects were provided information about experimental hypothesis. Schultz (1969) and Carlson (1971) both stressed, as did Argyris (1968), Cook et a1. (1970), and Weber and Cook (1972), the need for researchers to see subjects as full persons capable of cooperative, collaborative and meaningful efforts toward valid data gathering in psychological experiments. Schultz also examined subject roles from a historical perspective and reminded the reader of the structuralist's high regard for their specially trained subjects who offered their introspective reactions as skilled and motivated "recording instruments." But irregular and inconsistent results were generally disregarded--to the growing concern of those who wondered what conclusions could not be proved by eliminating inconsistent data? Debate grew between the structuralists, who looked for general laws of the human personality ' versus the newer and thereby saw individual differences as "noise,' functionalists who sought an understanding of individual differences. Eventually, with the use of undergraduate and graduate students as subjects and the emergence of behaviorism, the trend toward the total use of experimental observation of subject behavior was completed. The structuralist's trained and involved observer became today's observed and uninvolved subject. But Schultz (1969) believed that today's subjects are not really uninvolved objects or empty organisms. He stated that large numbers of "subject pool" subjects expressed strong negative attitudes, such as annoyance, irritation, fear and apprehension about the experimental -l8- situation. These attitudes caused some subjects to try to "beat" the experimenter, others to cooperate and still others to expect to be deceived. Thus the idea that psychologists use naive subjects seemed rather naive. Subjects responded to demand characteristics (Orne, 1962; Page, 1969; Banuazizi & Movahedi, 1975) as well as reacted oppositionally from a sense of dependency and powerlessness rooted in their subordinated relationship with the psychologist/experimenter. The next section discusses some suggested methods to avoid problems inherent in deception and cope with problems pertaining to demand characteristics. Alternative to Deception Schultz had recommended scrapping the obsolete view of man as an organic stimulus—response machine and replacing it with one which included an interaction with and an integration of the subject's experience. One suggested method of accomplishing this might be role playing (Schultz, 1969), in.which the subject's cooperation would be elicited and, hopefully, he would consciously participate in the experimental task rather than be an impersonal'victim". A number of studies utilized role play approaches (Greenberg, 1967; Horowitz and Rothschild, 1970; Willis and Willis, 1970), but Miller (1972) and Freedman (1969) each separately reviewed the evidence and both concluded that the prospects for role play as an alternative to deception were very poor. They cited as serious drawbacks (a) subject inability to role play, (b) the fact that role played behavior is not "real" -19- behavior, and (c) the need for direct comparison with "standard" methodology as proof of appropriateness. Another suggested way of attempting to change the mode of subject participation involved Jourard's (1968) suggestion of mutual self— disclosure between experimenter and subject to facilitate mutual openness, cooperation and trust. Although time-consuming, this technique might enlist the subject's cooperation and reduce the experimenter's need for deception. Research indicated that self- disclosure by one person elicited self-disclosure by another (Jourard and Landsman, 1960; Jourard and Richman, 1963; Rivenbark, 1971), even when intimacy (Ehrlich and Graeven, 1971) and the number of statements disclosed (Chittick and Himelstein, 1967) were varied. Disclosure was shown to effect the nature of the subject's responses in experiments and his view of the experimenter. Experimenters who dis- closed were found to elicit more disclosure from subjects and were rated more trustworthy (Drag, 1968) and more positively (Jourard and Friedman, 1970) than experimenters who did not self-disclose. When the experimenters self-disclosed in a subject-experimenter inter- action before the performance measure, subject performance was different than when subjects had no prior contact with the experimenter. Subjects of self-disclosing experimenters changed their responses to the Edwards Personal Preference schedule (Jourard and Kormann, 1968), gave more revealing responses on the Rotter Incomplete Sentences Blank (Heifitz, 1967) and required fewer trials to criteria on a paired associate learning task (Frey, 1968). -20- Jourard (1969) argued that experimenter self—disclosure would result in more honest subject responses and in increased cooperation and trust between subjects and experimenters. Although the evidence indicates this view may be accurate, the use of experimenter self- disclosure as a technique to elicit a new type of relationship between subject and experimenter in studies not examining disclosure is uncommon. Additional research was suggested by Cozby (1973) to determine under which conditions experimenter self-disclosure would effect experimental results, as well as to determine the nature of the effects themselves. Thus there appears to be no clear and easy way to alter experimental designs to cope with issues involving deception and experimenter demand characteristics, even though the new ethical standards attempts to do just that. The next section discusses these new standards as they pertain to deception, demand characteristics and subject-experimenter relationships. Subject-Experimenter Relationships and Ethical Standards Carlson (1971) complained that researchers did not really study the whole person inpersonality research, and argued that under the current highly impersonal contractual relationships, conscripted subjects were used, abused, and highly controlled, and that the use of existing experimental methods usually permitted the exploration of only one or two experimental variables. The irrelevance of research findings may result from research methods that violate human dignity and permit -21- manipulation, deception, depersonalization, and mistrust, for under these conditions few could ever glean meaningful information about the whole person. Carlson suggested dropping this contractual relationship and replacing it with a collaborative one where subjects and settings would be chosen for their intrinsic relevance to the problem at hand and the basic motive for the subject's research participation would be his intrinsic involvement in exploring his own experience. Carlson saw this model as more demanding and rewarding for both subjects and experimenters alike and also as fostering a deeper understanding of the human experience, personality and development and organization. The ethical concerns of Carlson and others repeatedly pertain to informed consent, full disclosure, deception, subject dignity, privacy and protection. These issues are intimately related to and imbedded within deceptive, contractual, and collaborative experimental research designs. A reading of the new Ethical Principles in the Conduct of Research with Human Participants (Ad hoc Committee, 1973) further adds awareness of the increasing pressures toward the design of less deceptive experiments and toward creating more collaborative relationships with research subjects. For example: Principle 3. Ethical practice requires the investi- gator to inform the participant of all features of the research that reasonably might be expected to influence willingness to participate, and to explain all other aspects of the research about which the participant in- quires. Failure to make full disclosure gives added emphasis to the investigator's responsibility to protect the wel- fare and dignity of the research participant. -22.. Principle 4. Openness and honesgrare essential characteristics of the relationship between investi- gator and research participant. When the methodological requirements of a study necessitate concealment or deception, the investigator is required to ensure the participant's understanding of the reasons for this action and to restore the quality of the relationship with the investigator. (p. 29) These principles clearly state that full disclosure to and openness with the subject is considered essential; they represent a dramatic shift in attitude from the contractual and deceptive experimental designs more permissible under previous ethical standardsQA.P.A., 1953, 1963). Following earlier guidelines, a psychologist is sometimes "justified in exposing research subjects, both children and adults, to physical or emotional stress as part of an investigation" ( A.P.A., 1963, p. 7). The shift is significant but its full impact is as yet unclear. Resnick and Schwartz (1973) pointed out that none of the changes in the new standards were themselves made subject to scientific investigation. Using these most recent A.P.A. ethical standards as their guide, the authors conducted a typical verbal conditioning experiment with two subject groups. The first group participated under a full disclosure ("ethical") instructional condition in which the nature of the experiment and even the verbal reinforcement schedule was explained. The second group underwent the experiment following ("non-ethical") partial disclosure instructions typical of this experiment and not including information pertaining to the -23- experimental variables. Results revealed that the " [partial disclosure] group showed significant positive conditioning, a finding typical of most traditionally conducted verbal conditioning studies, while those subjects in the [full disclosure] group showed a significant negative conditioning..." (p. 367) implying that instructional variables based on ethical considerations can have powerful influences on experimental results. Additionally, the authors reported difficulty in obtaining subjects for the ethical group, associating full disclosure with increased subject suspiciousness and a loss of interest in par- ticipation. The significant negative conditioning of the full disclosure subjects indicated a boomerang effect opposite that of the "good subject" response to the demand characteristics of the experiment. In explanation, it was hypothesized that if demand characteristics were very clear and obvious, the subject would become oppositional and suspicious to avoid being controlled. The authors concluded that the ethical standard used in an experiment could be a powerful independent variable affecting subject behavior and that "care should be taken when introducing methodological demands into a science that they do not weaken the established vigor of that science." This almost obvious statement seemed to acquire dramatic importance when attached to the verbal conditioning experiment of Resnick and Schwartz (1973). This was unfortunate, especially since their ex- perimental design may have been inappropriate or inadequate to assess the new ethical standards. For example: (1) the full disclosure -24- instructional condition.went so far beyond informing participants of features "that reasonably might be expected to influence willingness to participate..." (Principle 3) that subject suspiciousness and mistrust was created rather than reduced, quite contrary to the intent of the standards, (2) by using a verbal conditioning experiment and revealing the reinforcement schedule, the authors created a situation where the subject must consciously choose to be controlled by the already mis- trusted experimenter or to behave oppositionally, a more rational and expected choice under the circumstances, (3) no attempt was made to make the independent instructional variables more experimentally uniform, and (4) the experiment offered little that was of intrinsic value to subjects. The present study attempted to investigate the impact of the new ethical standards while eliminating the inadequacies inherent in the Resnick and Schwartz (1973) study. The relationship between subject anxiety level and capacity for empathy was examined under two different disclosure variables based on ethical considerations. Thus, various levels of disclosure as they relate to subject anxiety and performance will be reviewed, and the relationship between anxiety and empathy will be explored. Others have also felt apprehensive about the new ethical standards. Gergen (1973) stated his belief that further research should be conducted before establishing the principles. He believed that subjects often did not care about the research rationale, and felt that they -25- might even find deception intriguing. He believed these changes would be unimportant to subjects and would only create unnecessary hardships for experimenters. In support of this viewpoint, Sullivan and Deiker (1973) reported that psychologists gave significantly stricter interpretations to ethical issues than did student subjects in eighteen of twenty response categories concerning ethical issues. In the two remaining categories, concerning subjects giving or receiving electric shock, students were more strict, but the differences were not significant. These results might indicate an overconcern on the part of psychologists about ethical issues, and suggest that some manner of contribution by subjects in the formation of ethical principles and decisions may be appropriate. The next section examines how various degrees of disclosure in instructions affect subject performance and also review non-ethically related studies that use instructions as an independent variable. Disclosure Instructions as an Independent Variable In the present study, subject response style is examined as it pertains to differential disclosure instructions based on ethical considerations. The study most clearly related to this is "Ethical standards as an independent variable in psychological researc " (Resnick and Schwartz, 1973), previously reviewed and critiqued. In another related experiment, Wagner and Shaw (1973) measured the effects of the subject's awareness of the nature of a conformity experiment on behavior within that experiment. Some subjects were given standard -26- (partial disclosure) instructions and others given instructions warning them that the observed responses of others might be invalid. Results showed no statistically significant differences in conformity be- tween the two instructional groups, and the patterns of conformity were the same for both groups even though the "suspicious" group showed somewhat less conformity. Other studies have also studied subject response to instructional variables. Wolfer (1967) examined changes in dogmatism scale scores as a function of instructions and found that subjects scoring higher in dogmatism did not decrease their scores at retesting even though they knew the nature of the measure and were proven capable of lowering their D scores at will. In this study, as in Wagner and Shaw's (1973), additional knowledge of the nature of the study did not appreciably alter the experimental results, whereas in Resnick and Schwartz (1973) full disclosure had appreciably affected results. Instructions and Anxiety and Performance A number of researchers examined the responses of high anxious and low anxious subjects to differential instructions. In an attempt to facilitate optimal performance by high anxious (HA) and low anxious (LA) students on a written aptitude test, Heisler and Schill (1973) concluded that positive expectation-eliciting, reassuring instructions helped LA subjects perform significantly better than HA subjects, while HA subjects scored best after hearing negative expectation-eliciting, reassuring instructions. Similarly, Sarason (1956, 1957, 1958, 1959) studied the -27- interaction between HA and LA subjects and differential instructions on various subject performance measures. Sarason (1956, 1957) concluded that high motivational stress instructions were detrimental for HA groups and helpful to LA and middle anxious (MA) groups in serial verbal learning tasks, while under neutral or non—threat instructions, HA subjects performed as well as LA or MA subjects on similar tasks. Further investigation by Sarason (1958) indicated that following reassurance instructions performance improved for high test anxious subjects but worsened for low test anxious subjects. Under standard instructions the performance of low test anxious subjects was facilitated while the performance of high test anxious subjects suffered. Later, Sarason (1959) measured word—association performance as a function of anxiety and instructions and concluded that HA subjects would be the most sensitive to the personal threat implicit in stress instructions and would respond to stress instructions with task-irrelevant responses that interfered with their task performance. This conclusion is supported by an earlier study (Sarason, Mandler, Craighill, 1952) which measured the effect of high versus low ego-involvement instructions on learning with HA and LA subjects. They found the HA subjects receiving high ego-involvement instructions and LA subjects receiving low ego- involvement instructions performed poorly. Conversely, when HA subjects were paired with low ego-involvement instructions and LA subjects were paired with high ego-involvement instructions, performance improved. Winkel and Sarason (1964) further suggested that differential -23- instructions and anxiety level of the (male) experimenter interact with the anxiety level and sex of the subject. In this study instructions seemed to have minor impact on all subject groups with high test anxious experimenters, but in subject groups with low test anxious experimenters, high motivation instructions were more facilitative than neutral or reassurance instructions. Also, femalesperformed best when instructions were given by low test anxious experimenters. The author's concluded, as did Barber and Silver (1968), with a strong argument for careful consideration of experimenter variables in experiments which involve any social interaction between the subject and the experimenter and suggested using caution when examining results of experiments involving differential instructions not accounting for experimenter variables. A related study by Johnson (1973) concerned subject-experimenter interaction. Subject performance on a simple motor task was evaluated utilizing three subject conditions and three experimenter conditions. The experimenters were placed in one of three conditions of expectancy induced by the principal investigator: (a) where the hypothesis was simply told to the experimenter, (b) where the principal investigator conveyed great concern about the outcome, and (c) where the experimenter was told to actively manipulate subject responses. Subject evaluation apprehensiveness about evaluation was manipulated through the use of differentially arousing instructions which fostered high, medium or low evaluation apprehensiveness and ego-involvement. No experimenter -29- expectancy effects were found for groups having experimenters who were simply told the experimental hypothesis, but significant effects were found in groups with experimenters for whom the principal investigator role-played great concern about the outcome. In this experimenter group, the greater the subject's evaluation apprehensiveness, the better the subject's performance on the motor task. In the last condition, where the experimenter tried to actively manipulate subject responses, subject performance was opposite the intention of the experimenter manipulation in all subject conditions. Johnson postulated that the strong negative subject reaction fo the experimenter's manipulation may have been due to the subjects' reacting against the very conspicuous cues emitted by the experimenter. This interpretation is similar to that made by Resnick and Schwartz (1973) in their full disclosure verbal conditioning experiment where subjects knew the obvious experimental variables and reacted antagonistically to them. In a study examining the impact of different ego-involvement instructions on motivation and performance in a verbal learning skill, Ferguson (1971) reported that instructions had weak effects on performance even when they were shown to be ego-involving. She stated that'sometimes the instructions affected motivation, other times performance, and that the impact was not predictable. No measures of experimenter variables, such as anxiety or expectancy, were taken in this study. Two studies were found relating anxiety, instructions, and free verbalization or verbal efficienty. Gynther (1957) gave stress -30- instructions to one group of HA and LA subjects and non-stress in- structions to another group and then asked them for verbal responses to eleven questions. Analysis of communicational efficiency revealed no interaction between anxiety and instructions, although LA subjects had significantly higher communicational efficiency scores than HA subjects, and stressed subjects had lower communicational efficiency scores than non-stressed subjects. Sarason and Ganzer (1962) manipulated anxiety (HA and LA) and instructions (threat and non-threat) in a free verbalization situation in which subjects talked about themselves. Results indicated that HA subjects voiced significantly more negative self-references than LA subjects and that under the threat instructions HA subjects showed significant increases in negative self-references as a function of reinforcement. Again, HA subjects under stress faired poorly. Also related to stress instructions and negative self-references was a study by Bluxom (1968) who reported that subjects under anger-arousing instructions gave more negative self-references on personality questionnaires than subjects under neutral instructions. In summary, in studies where subjects knew the nature of the experiment and the experimental manipulations were obvious, subjects resisted succumbing to the demand characteristics of the experiment (Resnick and Schwartz, 1973; Johnson, 1973). In another experiment, where subjects knew the nature of the experiment but were left in some doubt or uncertainty, results similar to more standard experiments in the area were obtained (Wagner and Shaw, 1973). In studies seeking the -31- interaction of instructional variables with high anxious and low anxious subjects, results were generally consistent. Stress or threat or positive expectation-eliciting instructions were detrimental for high anxious subjects and facilitative for low and middle anxious subjects. Non-stress, non-threat, negative expectation-eliciting instructions were facilitative for high anxious subjects but detrimental for low and middle anxious subjects. It seemed that HA subjects performed better if instructions lowered anxiety, whereas LA subjects performed better if instructions raised anxiety. Thus both the "obviousness" of the experimental contingencies and the interaction of subject anxiety and instructions seemed to be important factors influencing experimental results but were often ignored by researchers. The present study examined subject anxiety and experimental conditions by controlling the interaction of two conditions of subject trait anxiety (HA or LA) and two conditions of experimenter feedback (full disclosure or deceptive disclosure) while the subject’s state anxiety and his performance on a measure of empathy served as dependent variables. State anxiety was chosen as one dependent variable because no study found directly measured subject anxiety level changes as a result of deception. Empathy was chosen as the second dependent variable because no study found directly measured the subject's capacity for empathy as it relates to his trait anxiety level. The next section examines the available literature focusing on the relationship between anxiety and empathy. -32- Anxiety and its Relation to Empathy No studies were found that related subject anxiety levels to empathic ability, although some studies correlated empathy with variables associated to anxiety. These studies as a group tended to show that high anxiety is associated with lower empathic ability. The work most closely relating to empathy with correlates of anxiety is Bergin and Solomon's (1963) study of test correlates of empathy using the Truax Accurate Empathy Scale, the Minnesota Multiphasic Personality Inventory, and the Edwards Personal Preference Schedule, with advanced graduate students in clinical and counseling psychology. 0n the MMPI the authors found Depression (D) and Psychasthenia (Pt) negatively and significantly correlated with accurate empathy. These scales, especially when considered with the scores on other scales, indicated personality disturbance and anxiety in subjects with poor empathy. Vesprani (1969) attempted a cross-validational study of Bergin and Solomon's (1963) work using untrained female college students as quasi-therapists. He also found that neurotic sympto- matology, as indicated by high scores on D and Pt, contraindicated high levels of accurate empathy. But his EPPS variables did not conform with the EPPS results of Solomon and Bergin, and in explanation, Vesprani suggested that a superficial or variable relationship with this ‘measure to accurate empathy may exist. In a related study, Bandura (1956) found high anxious therapists were rated less competent than low anxious therapists by their colleagues -33- and that awareness of anxiety did not improve therapeutic outcome. Competence was defined as the ability to facilitate improvement in the adjustment of patients as judged by supervisors and anxiety was judged by colleague ratings. If it is appropriate to assume that therapeutic competence and empathy are positively related, then this study indicates low therapist anxiety is related to high empathic ability and high therapist anxiety is related to low empathic ability. Other authors also studied empathy and its relationship to other variables seemingly related to anxiety. Dymond (1948), using the Thematic Apperception Test (TAT), the Wechsler-Bellevue Adult Intelligence Test, and interviews, conducted a preliminary examination of the relationship between empathy and insight. She stated that "the ability to describe the thoughts and feelings of others (empathy) [wasl accompanied by a better understanding of the relationship one has with others (insight)" and that those who had low insight into their own interpersonal relations as they revealed them in interviews were also low in empathy. Dymond (1949) examined the subject's ability to predict how others would rate themselves and how they would rate others on six traits and used this technique as a preliminary measure of empathy, defined "as the ability to transpose oneself into the thinking, feeling, and acting of another." Using this measure along with Rorschach, TAT, and other tests, Dymond (1950) studied the relationship between personality .and empathy. Results of this research indicated that insight was -34- positively related to empathy and those high in empathic ability were outgoing, warm and emotional people interested in others. Early family relations were sufficiently satisfying so that they found emotional investment in others rewarding. They were secure and their emotionality was both richly enjoyed and well controlled. Conversely, those low in empathic ability were found to be rigid, introverted, mistrusting, subject to uncontrolled emotionality, and either "lone wolves" or self-centered and demanding. Their family relations were unsatisfying and disturbed, causing them to be unable to invest themselves in others. This inward orientation interfered with their integration with the outside world and prevented them from empathizing with others. Using four projective tests, ten objective tests, and three interviews, Norman (1953) examined the interrelationships among a number of factors, including "realistic perception of others," which was defined similarly to empathy. His findings strongly corroborated Dymond's (1948) work and her(l950) description of those low in empathy. Cline (1955) investigated traits of good and poor judges of personality using a stress interview and sound film techniques. Good judges were defined as having the ability to perceive others in a relatively accurate and undistorted fashion, and this closely parallels Dymond's (1948) definition of empathy. WOrds on adjective self-check lists which characterized the best judges were sympathetic and affec- tionate, while adjectives of poor judges included dissatisfied, irritable, awkward, praising, and hurried. Using the MMPI and other -35- measures, Cline found high empathy for men related to superior intellectual functioning, an absence of hypochondriacal self-interest, a lack of bizarre and paranoid cognitive trends, and freedom from prejudice, bias and authoritarian attitudes. For women, social skills, artistic discrimination, and less compulsiveness in ideational and behavioral processes were all significantly related to good judging ability. Taft (1955) reviewed the literature on judging ability, also defined similarly to empathy, and summarized that high intelligence, insight into one's status with respect to peers on specific traits, good emotional adjustment and integration, and good social skills, were positively correlated with good judging ability. Negative correlates included social dependence and the MMPI psychasthenia score, both indicative of subjective distress. In examining psychological variables of judges and their inter- personal perceptions, Chance and Meaders (1960) found good judges were active and outgoing in social relationships but not markedly dependent upon them, higher in need affiliation and dominance on the EPPS, and had valued strong and close relationships and preferred sociAl to solitary activities. Poor judges were found to be demanding and self-centered or not preferring strong emotional ties with others, again corroborating Dymond (1950). Other studies further indicate high empathy as related to good emotional adjustment and flexibility. Lesh (1970) studied Zen meditation -36- and empathy development in counselors and found that those counselors more able to allow themselves to experience many different situations non-judgementally and able to regress in the service of the ego were Also more empathic, as defined by scores on the Affective Sensitivity Scale, a measure of empathy(l(agan, Krathwohl, & Farquhar, 1965) . Using the MMPI, Leary's Interpersonal Check List (ICL), and the TAT, Goldstein (1961) concluded that empathy was inversely related to repression, tending to confirm that a distrust of one's own feelings and impulses interfered with empathy. Empathy was defined as the average validity of each subject's prediction of the self-ratings of three other members of his group on the ICL. Suinn and Hill (1964) found high anxiety, as measured by the Taylor Manifest Anxiety Scale and the Sarason General Anxiety Questionnaire, was significantly associated with lowered self- acceptance and lowered acceptance of others, as measured by the Phillips Self-Other Questionnaire. Low anxiety permitted the usual self- acceptance-acceptance of others positive correlation to exist, which implied that high anxiety would interfere with empathy and low anxiety would facilitate it. Ormont (1960) hypothesized that adjustment, as measured by a sixty-eight item Q—Sort of items primarily from the WatsonAMunzer Adjustment Scale, was positively related to the ability to differentiate between others, as measured by a differentiation score derived from four Q-Sorts done by each subject on his co-members. The hypothesis of a positive correlation between adjustment and the ability to see -37- differences between others was confirmed. Similarly, Norman and Ainsworth (1954) found positive correlations between subjects' accurate perception of reality, high empathic ability and good emotional adjustment, whereas all three were negatively related to projection. Terms were defined operationally and this study further supports a positive relationship between adjustment and empathy, as does Jackson and Carr's (1955) work which controlled for projection and found schizophrenics less empathic than normals. In summary, the research generally tended toward agreement that those who were well adjusted, insightful, and socially invested in others were high in empathic ability whereas those who were more rigid, mistrusting, and exhibiting signs of conflict and distress were low in empathy. If well-adjusted people are less anxious than those poorly adjusted, than subjects low in anxiety would be more likely to exhibit high empathy whereas subjects with high anxiety would tend to score poorly on empathy measures. In conclusion, there is reason to believe that low anxiety is associated with higher empathy and high anxiety with lowered empathy, and that the ethical guidelines used within an experiment to test a re- lationship may significantly affect the results. This experiment examines the relationship between trait anxiety and affective sensitivity under two different ethical procedures, (1) a full disclosure treatment condition and (2) a deceptive disclosure treatment condition. It is hypothesized that the deceptive and full disclosure conditions will have an impact -33- on the relationship between trait anxiety and affective sensitivity, showing that ethical considerations do have unintended but serious consequences on experimental results. -39- CHAPTER III METHOD Subjects The total sample consisted of 224 male and female Michigan State University undergraduate introductory psychology students. Some participants were rewarded by extra credit toward their course grade for their voluntary participation in this study. They were recruited by the experimenter during brief visits to psychology classes. Instruments The State-Trait Anxiety Inventory: The State-Trait Anxiety Inventory, or STAI (Spielberger, Gorsuch, and Lushene, 1970), consists of two separate self-report scales: trait anxiety (A-Trait), and state anxiety (A-State). The A-Trait scale asks subjects to respond as they feel generally, and A-State asks subjects how they feel at a particular moment in time. Each scale consists of twenty statements and four response- intensity choices. -40- "State anxiety (A-State) is conceptualized as a transitory emotional state or condition of the human organism that is characterized by subjective, consciously perceived feelings of tension and apprehension, and heightened autonomic nervous system activity. A-States vary in intensity and fluctuate over time. "Trait anxiety (A-Trait) refers to the relatively stable individual differences in anxiety proneness, that is, to differences in the tendency to respond to situations perceived as threatening with elevations in A-State intensity" (Spielberger et al., 1970, p. 3). State anxiety refers to a process or reaction which is taking place now at a given level of intensity, wrote Spielberger (1966), whereas A—Trait indicates a latent disposition for a reaction of a certain type to occur if it is triggered by appropriate stressful stimuli. Levitt (1967) called the STAI the most carefully developed instrument reviewed in his chapter on experimental measures of anxiety, from both theoretical and methodological standpoints. Other studies attest to the transitory nature of state anxiety and emphasize the stability of trait anxiety over time. Allen (1970) gave a battery of test anxiety scales and administered two physiological measures to subjects under three different sets of demand characteristics to study the effect of situational aspects on anxiety scores. Results indicated that A-Trait was not influenced by instructional variables, whereas A-State fluctuated appropriately -41- as demand characteristics provoked increased test anxiety. Newmark (1972) also found impressive the stability of A-Trait and the transitory nature of A-State through repeated administration of both measures to a large group of subjects over a ten month period. Bartsch (1972) used experimental treatments to alter subject anxiety level and found A—State sensitive to these fluctuations and A-Trait stable, even when subjects were told to respond to the A-State scale as they feel "generally" and to the A—Trait as they feel "presently." Affective Sensitivity Scale: The Affective Sensitivity Scale (AS) is designed to measure what might be termed, generically, empathy. Affective sensitivity is conceptualized as "the ability to detect and describe the immediate affective state of another, or in terms of communication theory, the ability to receive and decode affective communication" (Kagan, Krathwohl, 8 Farquhar, 1965). In other words, it is one's "sensitivity to what is going on in another person at a particular moment in time" (Schneider, 1975). Affective sensitivity was measured through the use of a 16mm sound film (form D) containing 30 brief scenes of interpersonal encounters. After each scene the subject answers two or more written multiple choice questions designed specifically for that n scene. These questions ask the subject what the persons are "really Saying to themselves at a particular moment. The film sequences -42- and the multiple choice items together are called the Affective Sensitivity Scale and is designed to differentiate between those sensitive and those not sensitive to the affective state of another (Danish & Kagan, 1971). The scale's instructions begins as follows: You will be viewing short scenes of actual encounters between two or more individuals. You are to identify what feelings the people have toward themselves and toward the person they are working with. You will be instructed just prior to each scene which individual(s) to focus on. Although in any one scene a client or interviewer may exhibit a variety of feelings, for the purpose of this instrument you are to concentrate on identifying hi§_last feeliggs in the scene. In some scenes, you will also be asked about the nature of the last response. After you view each scene, you are to read the items and ask yourself the following questions, depending on the specific instructions for the scene: If the people involved were to view this same scene, and if they were completely open and honest with themselves, (i.e., if they could identify their real feelings) which of these responses would they use to describe their feelings? After you decide which response accurately describes what the people are actually feeling whether about themselves or the others they are with, indicate your choice on the answer sheet. On the following pages are multiple choice items usuilly consisting of five responses each. Most scenes have two items, but a few have one or three items. The instructions continue with sample items and additional information about chasing a response, time limits, etc. -43- For the previous form of the AS, internal consistency reliability coefficients range between .58 and .77, with most coefficients in the .703. Test-retest reliability was .75 (Campbell, 1967; Kagan, Krathwohl, Goldberg, Campbell, Schauble, Greenberg, Danish, Resnikoff, Bowes, & Bondy, 1967) while several further studies (Danish & Kagan, 1971; Campbell, Kagan and Krathwohl, 1971) have yielded relatively impressive figures on concurrent and construct validity checks. In one concurrent validity study, cor- relations were calculated between therapist's ranking of group members sensitivity to feelings expressed by others and member's AS scores, and this relationship was significant (p ‘»01). Average correlations across three studies of concurrent validity was .26 with a range from .42 to .17. The average correlation obtained across studies of scale scores and judgment of affective sensitivity was .38, with a range from .64 to -.10. In a construct validity study conducted to determine if scale scores would increase during a counselor education program, statistically significant growth had occurred that was not attributed to any practice effect. Form D of the AS was developed using the same metholology as the previous form, and therefore the original data is believed appropriate for application to this form (Personal communication with the author, John Schneider). For form.D, the average standard deviation is 8.5 and the standard error of measurement is 3.65. Additionally, the normative data (N - 240) indicates that physicians (N - 115) score -44- higher on health care items, counselors (N - 11) on psychotherapy inter— actions, and high empathizers (N = 2) score higher in general. The author further states that in two separate studies subject's AS score significantly increased after IPR training. In a third similar study, differences were not significant but the subject's scores in those areas that IPR concentrates on did in fact increase. Experimental Procedure Subjects, in groups of 15 to 25, entered a university classroom having movable desks and fluorescent lighting administered Spielberger's A-State and then the A-Trait. To insure confidentiality and anonymity, code numbers were used and explained by the experimenter to the subjects. While the trait anxiety questionnairewas being scored by sides, the experimenter offered more information about the experiment to the subjects. The experimenter said: "The experiment is looking at the relationship between certain personality traits and your capacity to be sensitive to another person's feelings. For example, an experimenter might be interested in whether people exhibiting high levels of warmth are more or less sensitive to another's feelings than people exhibiting low levels of warmth. Or another question an experimenter may choose to examine is whether people with a high tendency to feel guilty will be more or less sensitive than people with a low tendency to feel guilty. You will soon receive a short personality description based on your questionnaire results taken earlier. This general description is based on particular traits measured by the questionnaire and I will tell you specifically what these traits are toward the end of the experiment. After you read this description of yourself you will take one more brief questionnaire similar to the first one you took. This is to see if some of your feelings have -45- changed or stayed the same since you have been here. Then we will start the measure of sensitivity that you were promised." The experimenter distributed the IBM answer forms for the AS. After subjects filled in appropriate identifying information, they received the AS question booklet. The experimenter read the instructions to the subjects as they read along with him and they then reviewed a sample question together. By this time, the aides had completed scoring the trait anxiety questionnaire and had divided subjects into groups of low or high trait anxiety by choosing to use cut off scores which, on the average, eliminated the middle 24% of subjects in each group, according to Spielberger's normative data (Spielberger et al., 1970). Thus, high trait anxious subjects were in the 62nd percentile or greater, and low trait anxious subjects were in the 38th percentile or lower. Subjects scoring in the middle 24% (middle trait anxious) received neutral feedback, remained to complete the dependent variables, and were not evaluated statistically. Table 1 Trait anxiety scores defined for this experiment as high trait or low trait anxious for subject's sex and university class. Freshman Sophomore, Junior,,Senior ‘Male Female Male Female High Trait 40 or more 40 or more 40 or more 39 or more Anxious Low Trait 35 or less 35 or less 33 or less 35 or less Anxious -46- The high trait anxious group and the low trait anxious group were each then randomly divided into a deceptive disclosure group and a full disclosure group, yielding four cells in total. These cells were: high trait anxious, deceptive disclosures (HADD), high trait anxious, full disclosure (HAFD), low trait anxious, deceptive disclosure (LADD), and low trait anxious, full disclosure (LAFD). All subjects heard the same verbal instructions and explanations, but the written personality data differed between experimental groups. The FD subjects received personality data sheets congruent with their true results on the trait anxiety questionnaire. The LADD subjects received personality data sheets with information indicating high anxiety. This was opposite their true results on the trait anxiety questionnaire, and the reversal constituted the deception. Similarly, all HADD subjects received data sheets indicative of low anxiety. The experimenter was unaware of the subjects' treatment group status. The personality descriptions for both the HAFD subjects and the LADD subjects read: You are a person who is usually discounted and unhappy and often suffers from periods of feeling blue and down for no apparent reason. You rarely feel genuinely self—confident and frequently cannot forget your problems, causing tension, and your head is often disorganized and distracted. You are usually not very content. The personality descriptions for both LAFD subjects and the HADD subjects read: ~47— You are a person who is usually content and happy and rarely suffers from periods of feeling blue and down. You usually feel genuinely self-confident, take disappointments well and generally handle your difficulties and responsibilities appropriately and competently. You can easily put your problems aside when you wish to and relax for a while. Your head typically feels clear and alert and you are usually quite content. Subjects in the middle 24% of the trait anxiety scores read statements asking them to remain and complete the experiment and get their results on the AS in spite of the fact that their questionnaire results were inconclusive for the purposes of the experiment. Full or deceptive disclosure based on the trait anxiety questionnaire was the only experimental treatment. Full disclosure subjects received accurate feedback based on their trait anxiety questionnaire score, and, conversely, deceptive disclosure subjects read feedback opposite of what their score indicated. After subjects read the feedback (treatment) they took the second state anxiety questionnaire (82) and then the AS. The first dependent variable was change in state anxiety score, which was defined as the score on the second state anxiety questionnaire mdnus the score on the first state anxiety questionnaire (82-81) for each subject. 82 was administered to subjects immediately after the experimental treatment. The second dependent measure was the subject's score on the Affective Sensitivity Scale. The AS was administered to subjects directly following 82. -48- The AS was scored whfle the subjects filled out the post- experimental questionnaire, and then the experimenter began the debriefing process. Subjects then received an explaination of their treatment group, information about their deceptive or non- deceptive status, their AS score,and AS normative scores which they could use to assess their own performance. This procedure studied the impact of two levels of disclosure and necessarily involved the use of deception. The deception was mild and brief, with one half of the subjects deceived and one half not deceived. Debriefing immediately followed the experiment, and deceived subjects were told of the deception and of their true results on the trait anxiety questionnaire. All subjects were asked not to speak of the experiment to others until its final completion. All subject groups were run over four days. The researcher's telephone number was offered for any subjects having concerns or questions that needed a more personal or private response. This experiment was significantly different from the Resnick rand Schwartz (1973) study in that in the present study (1) the full disclosure condition was less revealing, thus less suspicious behavior on the part of subjects was anticipated, (2) the experiment's dependent variables were measures that.more naturally elicited the subjects' self-exploration and cOOperation rather than oppositional tendencies, and (3) the experimental procedure was uniform for each -49- subject in each experimental group. Thus, the present research examined the ethical variable of deception in a framework which attempted to correct the problems inherent in the Resnick and Schwartz (1973) study. It offered the subject significant intrinsic motivation to participate and cooperate, fewer control issues, and less complicated instructions. It was expected that significant information would be obtained pertaining to the impact of disclosure variables on subject state anxiety and performance levels. -50- CHAPTER IV HYPOTHESES The experimental hypotheses were: 1. The deceptive disclosure subjects, regardless of trait anxiety group, will increase in state anxiety relative to the full disclosure subjects. That is, 81 - 81 for deceptive disclosure subjects will be greater than $2 - 81 for full disclosure subjects. (A positive difference score indicates an increase in state anxiety over time, and a negative difference score indicates a decrease in state anxiety over time.) 2. Full disclosure subjects will score higher on the AS than deceptive disclosure subjects, regardless of trait anxiety group. 3. Low trait anxious subjects will score higher on the AS than high trait anxious subjects, regardless of disclosure condition. No interaction was expected between main effects. Statistical hypotheses are as follows: 1. Difference Score (82 - 81) Ho: DD = FD H1: DD4> FD -51- 2. Affective Sensitivity Scale Ho: DD = FD Hl: FD >' DD Affective Sensitivity Scale Ho: LA = HA H1: LA 7 HA -52- CHAPTER V RESULTS Description_gf the data The data was analyzed as a fixed effects 2 x 2 completely randomized factoral design. There was a LAFD, LADD, HAFD, and an HADD group, yielding four cells in total. The first dependent measure was change in state anxiety score, defined as 82 - 81. A positive difference indicated an increase in state anxiety over time and a negative difference score indicated a decrease. The second dependent measure was the subject's score on the AS. The means and standard deviations for all subjects together (Table 2) and also for subjects divided into treatment cells (Table 3) for trait anxiety, 81, 82, and the AS, are as follows: -53- Table-2 Means and standard deviations for SI, 82, and AS, for all subjects (n = 224). Mean SD S1 37.317 9.090 82 36.477 9.705 AS 66.495 9.101 Table _1_3_ Means and standard deviations and N's for trait anxiety, 81, $2, and AS, by treatment group (n = 224). Trait Anxiety LA HA Mean SD __N Mean SD _N_ TA 30.294 3.37 51 48.189 6.72 58 DD 81 31.294 5.28 51 42.655 9.37 58 82 32.176 10.49 51 39.776 7.43 58 AS 67.117 8.09 51 64.448 11.14 58 Disclosure TA 31.134 2.94 52 47.424 6.50 63 FD 81 35.096 7.78 52 40.889 8.46 63 52 31.654 8.43 52 40.905 8.71 63 AS 67.557 6.96 52 67.000 9.26 63 The post-experimental questionnaire contained two questions. The first was: "At the start of this experiment I felt the experimenter was open and honest with me. T or F." Of 217 responses, 209 indicated trust and 8 indicated mistrust of the experimenter. 7 subjects did not respond to the questionnaire. The second question was: "Chose the statement that best describes your feelings at the time you received and read the personality data sheet about yourself: a) I felt the description -54- was inaccurate even though the experimenter was trying to be honest and scientific. b) I felt the description was fairly accurate and I was pleased to read it. c) I felt the description was inaccurate and that the experimenter was trying to somehow trick me. d) I felt the description was fairly accurate but I was not very pleased to read it anyway. e) Other." Subject's responses to this question, divided by group, are as follows: Table 4 Subject's post—experimental questionnaire responses within each group (n = 217). Trait Anxiety LA HA 1.2.2.2.; 1.11-9.12. DD 10 3 25 6 6 10 25 12 2 7 Disclosure FD l 41 l 0 7 l8 0 7 19 17 Tests 2£_individua1 hypotheses Hypothesis one had the state anxiety difference scores for each cell approximately normal and it met the F max test for homogeneity of variance. The data satisfied the other assumptions for analysis of variance. -55- Tab1e_5 Means and N's for state anxiety difference scores for each treatment group (n = 224). Trait Anxiety LA HA Total Mean ._N Mean _N_ Mean N DD 0.882 51 -2.879 58 -l.1l9 109 Disclosure FD -3.442 52 0.016 63 -l.548 115 Total —l.103 103 -l.372 121 Table 6 Source table--difference scores for all subjects (n = 224). Source Sum of Squares DF Means quared F CK D (Disclosure) 28.3997 1 28.3997 0.0496 NS D at LA 481. l 481. 6.94 p <.0.01 D at HA 256. l 256. 3.69 p 4.0.10 A (Anxiety) 1.2801 1 1.2801 0.0185 NS A at DD 387. l 387. 5.58 p < 0.025 A at FD 341. l 341. 4.92 p.< 0.05 D x A 724.4431 1 724.4431 10.447 p <.0.0014 Error(Within) 15,255.1797 220 69.3417 Total 16,009.3026 223 As can be seen from Table 6, main effects for disclosure and trait anxiety were not significant. There was a significant interaction, however, between disclosure and trait anxiety (F = 10.447, p <.0.0014) and therefore simple main effects were calculated. Simple main effects indicated that LADD subjects had significantly higher state anxiety change scores (82 - 31) relative to HADD subjects (F = 5.58, p (10.025). LAFD subjects were significantly lower in state anxiety -56- change scores relative to HAFD subjects (F = 4.92, p < 0.05). HADD subjects had lower state anxiety change scores relative to HAFD subjects, but this difference only approached significance (F = 3.69, p 4 0.10). LADD subjects had significantly higher state anxiety change scores relative to LAFD subjects (F a 6.94, p L-0.01). In other words, according to the state anxiety change scores, HADD subjects became relatively less anxious and LADD subjects became relatively more anxious. Also, according to state anxiety change scores, LAFD subjects became relative less anxious and HAFD subjects became relatively more anxious. Thus, in each of the four groups, the state anxiety scores changed in the direction of the treatment disclosure content, relative to other groups. That is, if subjects read feedback based on high trait anxiety, they responded with an increase in state anxiety, and if subjects read feedback based on low trait anxiety, they responded with a decrease in state anxiety, regardless of whether the disclosure was full or deceptive. This suggests that it is insufficient to examine only the treatment's deceptive or non-deceptive status; the treatment also offers a directionality of content which must be considered. Since the subjects' 81 scores had a range from 21 to 70 and a ‘mean of 37.317, an attempt to control for the possible tendency of 82 scores to regress toward the mean was made. Therefore, subjects ‘whose 81 score was from 34 to 40 only were selected and analyzed. -5 7- These specially selected subjects were divided into the four cells and were found to be equivalent groups. Table 7 S1 means and N's for selected subjects (n = 75). Trait Anxiegy» LA HA Mean __N Mean _N_ DD 36.27 11 36.42 19 Disclosure FD 36.56 23 36.77 22 An analysis of variance was performed on the selected group of 75 males and females examining the difference scores. There was a main effect for trait anxiety (F = 22.02, p <~0.001), but neither disclosure nor the interaction was significant. This means that, regardless of disclosure condition, low trait anxious subjects had significantly lower state anxiety change scores on the 82, when compared to high trait anxious subjects. Table _8_ Means for difference scores, selected subjects (n = 75). Trait Anxiegy LA _N_ HA __N_ Total __b_l_ DD -5.818 11 0.211 19 -2.26 30 Disclosure FD -5.l30 23 1.364 22 -3.28 45 Total -5.353 34 0.829 41 -58- Thus, for the selected group subjects, an examination of the state anxiety difference scores showed significant main effects for trait anxiety. Regardless of the disclosure condition, low trait anxious subjects had significantly lower state anxiety levels after the treatment, when compared to high trait anxious subjects. This differed from the full group (n = 224) results where the difference score for each cell tended to change in the direction of the treatment Statement . A post-hoc examination of possible sex differences was also 'made for each hypothesis using the same statistical procedures. For :males in the total population, with respect to the difference scores, there was a significant main effect for anxiety (F - 8.646, p¢L0.005). Disclosure was not significant but the interaction of trait anxiety and disclosure showed a trend toward significance (F = 2.747, p 4.0.10). Therefore, simple main effects were calculated. Table.2 Means and N's for difference scores, males (n = 52). Trait Anxiety LA HA Total Mean _N_ Mean _N_ Mean _N_ DD -2.500 14 -0.692 13 -l.62 27 Disclosure FD -4.867 15 1.300 10 -2.40 25 Total -3.724 29 0.170 23 -59- Table 10 Source table--difference scores, males (n = 52). Source Sum of Squares 2§_ Means Squared F o( D (Disclosure) 2.679 1 2.679 0.122 NS D at LA 132. l 132. 6.02 p«< 0.05 D at HA 22. 1 22. 1.01 NS A (Anxiety) 189.878 1 189.878 8.646 pl< 0.005 A at DD 22. l 22. 1.01 NS A at FD 228. 1 228. 10.38 p 4 0.005 D x A 60.316 1 60.316 2.747 p ((0.10 Error(within) 1,054.097 48 21.960 Total 1,311.995 51 Simple main effects indicated that although both LAFD and LADD male subjects showed lower state anxiety scores, LAFD subject's scores were significantly lower than LADD subjects (F = 6.02, p4.0.05). Also, LAFD subjects showed significantly lower state anxiety relative to HAFD subjects (F = 10.38, p<;0.005). There was no significant difference between HAFD and HADD nor between LADD and HADD. Thus, for each comparison where there was a significant difference, state anxiety changes were in the direction of the treatment content, relative to other cells. With respect to the difference scores for the females, main effects for disclosure and anxiety were not significant, but there was a significant interaction between disclosure and anxiety (F - 8.816, p 4.0.004) and simple main effects were calculated. -50- Table‘ll Means and N's for difference scores, females (n = 172). Trait Anxiety LA HA Total Mean __N_ Mean ‘ __N Mean __N DD 2.162 37 -3.511 45 -0.95 82 Disclosure FD -2.865 37 -0.226 53 -l.3l 90 Total -0.35 74 -1.73 98 Table l3. Source table--difference scores, females (n = 172). Source Sum of Squares DF Means quared F c¥ D (Disclosure) 3.982 1 3.982 0.048 NS D at LA 468. l 468. 5.68 p{ 0.05 D at HA 263. 1 263. 3.19 p~’— 0.10 A (Anxiety) 79.109 1 79.109 0.96 NS A at DD 653. 1 653. 7.93 pi4 0.01 A at FD 153. l 153. 1.86 NS D x A 726.107 1 726.107 8.816 p.; 0.004 Error(within) 13,837.586 168 82.367 Total 14,648.363 171 Simple main effects indicated that LADD female subjects had significantly higher state anxiety change scores relative to LAFD subjects (F n 5.68, p<:0.05) and HADD subjects had lower state anxiety change scores relative to HAFD subjects, but this difference only approached significance (F = 3.19, p<;0.10). Also, LADD subjects had significantly higher state anxiety change scores relative to HADD subjects (F = 7.93, p IVHMHNOS O (3 OS LHHIVHHGOH 0 <3 08 HDflN AREA 000 (SECOND) SELF-EVALUATION QUESTIONNAIRE Name Student Number DIRECTIONS: A number of statements which people have used to describe themselves are given below. Read each statement and then blacken in the appropriate circle to the right of the statement to indicate how you GENERALLY FEEL. There are no right or wrong answers. Do not spend too much time on any one statement but give the answer which seems to describe how you GENERALLY FEEL. I feel pleasant --------------------------------------------- I tire quickly ---------------------------------------------- I feel like crying ------------------------------------------ I wish I COUId be as happy as others seem to be ------------- I am.1osing out on things because I can't make up my mind soon enough ................................................. I feel rested ............................................... I am "calm, cool, and collected" ---------------------------- I feel that difficulties are piling up so that I cannot overcome them --------------- - ............................... I worry too much over something that really doesn't matter-- I am.happy -------------------------------------------------- I am.inclined to take things hard --------------------------- I 186k self-confidence -------------------------------------- I feel secure ----------------------------------------------- I try to avoid facing a crisis or difficulty ---------------- I feel blue------------------------------------I ............ I am content ................................................ Unimportant thoughts run through my mind bothering me ------- II take disappointments so keenly that I can't put them out ‘of ny'mdnd -------------------------------------------------- I am a steady person ---------------------------------------- I get in a state of tension or turmoil as I think over my recent concerns and interests ............................... -85.. HEAHN ISONTV SEHIIEHOS O NELJO O SXVMHV ISONHV O O (THIRD) SELF-EVALUATION QUESTIONNAIRE Name Student Number Directions: A number of statements which people have used to describe themselves are given below. Read each statement and then blacken in the appropriate circle to the right of the statement to indicate how you FEEL RIGHT NOW, AT THIS MOMENT. There are no right or wrong answers. Do not spend too much time on any one statement but give the arsver which seems to describe your present feelings best. I feel calm ................................................. I feel secure ............................................... I am tense -------------------------------------------------- I am regretful .............................................. I feel at ease ---------------------------------------------- I feel upset ----- - ------------------------------------------- I am presently worrying over possible misfortunes ----------- I feel rested ----------------------------------------------- I feel anxious ---------------------------------------------- I feel comfortable ------------------------------------------ I feel self-confident --------------------------------------- I feel nervous ---------------------------------------------- I am jittery ................................................ I feel "high strung" ---------------------------------------- I am relaxed ------------------------------------------------ I feel content ---------------------------------------------- I am worried ------------------------------------------------ I feel over-excited and "rattled" --------------------------- I feel joyful ----------------------------------------------- I feel pleasant -------- - ------------------------------------ -86- 11V IV ION O IVHHHNOS O 0 OS ATHIVHHOON O O 08 HORN AHEA O APPENDIX B -37- dd,1a DEBRIEFING INFORMATION Name: Affective Sensitivity Scale: Top Middle Bottom third of normative group You responded to three questionnaires that measured anxiety. The first questionnaire measured trait anxiety. Trait anxiety is a relatively stable predisposition toward anxiety proneness, or, in other words, your general tendency to react with anxiety to situations perceived as threatening. Your true score was in the lower range of your normative group. That is, you are a rather low trait anxious person and do not tend to be anxiety prone. The description you read earlier was deceptive. It was opposite your true results. This deception, although mild, was necessary in order to conduct this eXperiment and we are now correcting it with you. Your feedback, based on your true test results, should have been: You are a person who is usually content and happy and rarely suffers from periods of feeling blue and down. You usually feel genuinely self-confident, take disappointments well and generally handle your difficulties and responsibilities appropriately and competently. You can easily put your problems aside when you wish to and relax for a while. Your head typically feels clear and alert and you are usually quite content. The second and third questionnaire measured state anxiety. State anxiety is defined as your anxiety level at a particular moment under particular circumstances. It is transitoryeand fluctuates over time. The Affective Sensitivity Scale measures your sensitivity to another's feelings. Your score, relative to your normative group, is shown above. The experimenter will now more fully explain the experiment to you and answer questions that you may have. Again, thank you for your cOOperation and participation. Win Schachter Ph.D. Candidate Clinical Psychology -33- fd.la DEBRIEFING INFORMATION Name: Affective Sensitivity Scale: Top Middle Bottom third of your normative group You responded to three questionnaires that measure anxiety. The first measured trait anxiety. Trait anxiety is a relatively stable predisposition toward anxiety proneness, or, in other words, your general tendency to react with anxiety in situations perceived as threatening. Your score was in the lower range of your normative group. That is, you are a rather low trait anxious person and do not tend to be anxiety prone. The description you read earlier was based on your low trait anxiety score. The second and third questionnaires measured state anxiety. State anxiety is defined as your anxiety at a particular moment under particular circumstances. It is transitory and fluctuates over time. The Affective Sensitivity Scale measures your sensitivity to another's feelings. Your score, relative to your peers, is shown above. The experimenter will now more fully explain the experiment and answer questions that you may have. Again, thank you very much for your participation and cooperation. Win Schachter Ph. D. Candidate Clinical PsYchology -39- dd,ha DE BRIEFIm INFORMATION Name: Affective Sensitivity Scale: Top Middle Bottom third of normative group You responded to three questionnaires that measured anxiety. The first questionnaire measured trait anxiety. Trait anxiety is a relatively stable predisposition toward anxiety proneness, or, in other words, your general tendency to react with anxiety to situations perceived as threatening. Your true score was in the upper range of your normative group. That is, you are a rather high trait anxious person and do tend to be anxiety prone. The description you read earlier was deceptive. It was apposite your true results. This deception, although mild, was necessary in order to conduct this experiment and we are now correcting it with you. Your feedback, based on your true test results, should have been: You are a person who is usually discontented and unhappy and often suffers from periods of feeling blue and down for no apparent reason. You rarely feel genuinely self—confident and frequently take disappointments badly and feel your difficulties and responsibilities too much to handle. You frequently cannot forget your problems, causing tension, and your head is often disorganized and distracted. You are usually not very content. The second and third questionnaire measured state anxiety. State anxiety is defined as your anxiety level at a particular moment under particular circumstances. It is transitory and fluctuates over time. The Affective Sensitivity Scale measures your sensitivity to another's feelings. Your score, relative to your peers, is shown above. The experimenter will now more fully explain the experiment to you and answer questions that you may have. Again, thank you for your c00peration and participation. Win Schachter Ph.D. Candidate Clinical Psychology -90- fd,ha DEBRIEFING INFORMATION Name: Affective Sensitivity Scale: Top Middle Bottom third of your normative group You responded to three questionnaires that measure anxiety. The first questionnaire measured trait anxiety. Trait anxiety is a relatively stable predisposition toward anxiety proneness, or, in other words, your general tendency to react with anxiety to situations perceived as threatening. Your score was in the upper range of your normative group. That is, you are a rather high trait anxious person and tend to be anxiety prone. The description you read earlier was based on your high trait anxiety score. The second and third questionnaires measured state anxiety. State anxiety is defined as your anxiety at a particular moment under particular circumstances. It is transitory and fluctuates over time. The Affective Sensitivity Scale measures your sensitivity to another's feelings. Your score, relative to your peers, is shown above. The experimenter will now more fully explain the experiment and answer questions that you may have. Again, thank you for your participation and cooperation. Win Schachter Ph.D. Candidate Clinical Psychology -91.. DE BRIEFING INFORMAT ION Name: Affective Sensitivity Scale: Top Middle Bottom third of normative group You responded to three questionnaires measuring anxiety. The first measured trait anxiety. Trait anxiety is a relatively stable predisposition toward anxiety proneness, or, in other words, your general tendency to react with anxiety in situations perceived as threatening. Your score was in the middle of your norm group. That is, your tendency to react with anxiety is. about average for your age and level. Only subjects in either the high or low ranges were used as full subjects in this experiment. The second and third questionnaires measured state anxiety. State anxiety is defined as your anxiety at a particular moment under particular circumstances. It is transitory and fluctuates over time. The Affective Sensitivity Scale measures your sensitivity to another's feelings. Your score, relative to your normative group, is shown above. The experimenter will now more fully explain the experiment and answer questions that you may have. Again, thank you very much for your participation and cooperation. Win Schachter Ph.D. Candidate Clinical Psychology -92- "ImammalES