UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Incongruous pain display as a source of self-deception Swalm, Delphin Marlene 1992

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

Item Metadata


831-ubc_1992_spring_swalm_delphin_marlen.pdf [ 1.75MB ]
JSON: 831-1.0098896.json
JSON-LD: 831-1.0098896-ld.json
RDF/XML (Pretty): 831-1.0098896-rdf.xml
RDF/JSON: 831-1.0098896-rdf.json
Turtle: 831-1.0098896-turtle.txt
N-Triples: 831-1.0098896-rdf-ntriples.txt
Original Record: 831-1.0098896-source.json
Full Text

Full Text

INCONGRUOUS PAIN DISPLAY AS ASOURCE OF SELF-DECEPTIONbyDELPHIN MARLENE SWALMB.A. (Honours) University of Alberta 1984M.A. University of British Columbia 1987A THESIS SUBMITTED IN PARTIAL FULFIL]EI4ENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Department of Psychology)We accept this thesis as conformingto the required standardsTHE UNIVERSITY OF BRITISH COLUMBIAMarch 1992J Deiphin Marlene Swalm, 1992I+ National Litxaryof Canada Biblio(hèque nationaledu CanadaCanadian Theses Service Service des thèes canadiennesOuawa. CanadaKIAON4The author has granted an irrevocable nonexclusive licence allowing the National Ubraryof Canada to reproduce, loan, thstzibute or sellcopies of his/her thesis by any means and inany form or format, maldng this thesis availableto interested persons.The author retains ownership of the copyrightin his/her thesis. Neither the thesis norsubstantial extracts from it may be printed orotherwise reproduced without his/her permission.L’auteur a accordé une licence irrevocable etnon exdusive pennettant a Ia Bibtiothèquenatiönale du Canada de reproduire, préter,distribuer ou vendre des copies de sa thesede quelque manlére et sous quelque formeque ce soit pour mettre des exemplaires decette these a Ia disposition des personriesintéressées.L’auteur conserve Ia propciété du droit d’auteurqui protege sa these. Ni Ia these ni des extraitssubstantiels de celle-ci ne doivent êtreimprimés ou autrement reproduits sans sonautorisation.CariadISBN 0-315-75399-4In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of______________The University of British ColumbiaVancouver, CanadaDate %4VtcI 3DE-6 (2188)AbstractIn some cases of chronic pain, the complaints seem out ofproportion to pathophysiological findings. Several models ofpain have been forwarded to account for such cases, but no oneexplanation can account for the underlying processes involved inthe genesis of chronic pain in all cases. The present analysisoffers the additional factor of self—deception, defined as acontradiction between one’s words or attitude and behavior. Byattempting to demonstrate subjective pain to observers, a painpatient convinces him- or herself of the displayed painfulnessthrough a process of self—deception. Several psychologicalphenomena are included in the model of self—deception, includingcoping strategies, cognitive dissonance, self-perception,impression management, and attentional and memory biases.To explore the self—deception model of chronic pain, alaboratory analogue study was devised using female studentvolunteers who rated the painfulness of shock—induced stimuliunder conditions designed to foster self—deception. Painfulnesswas measured 1) verbally by means of two visual analogue scaleswhich reflected the pain intensity and affective unpleasantnessand 2) nonverbally by means of quantified facial musclemovements. For each subject, individual pain threshold andtolerance levels were established. She then underwent a pretestcomprising five random shocks from her threshold to tolerancerange. Next, in the manipulation phase the subject was asked todisplay more, less or the same degree of pain while undergoinganother random series of shocks. A final posttest was identicaliito the pretest and provided a measure of the durability of thealtered pain display effect. In the first of two studies, thealtered pain display was nonverbal: subjects exaggerated,diminished or did not change their facial expressiveness whileundergoing the pain stimuli. In the second study, the alteredpain display was verbal: subjects were told that at the end ofthe series they would be required to tell a fellow student (viavideotape) that the shocks hurt more, less or about the same aswhat they had expected. Half of all subjects were further toldthat their deceptive communication would have negativeconsequences for viewers.Misleading fellow students about the pain experienced wasexpected to make the subjects feel badly, motivating them tochange their attitude or beliefs about the pain experienced.They were expected to change their pain reports in keeping withthe deceptive communication. That is, other deception wasexpected to foster self-deception. This effect was expected toendure and it was expected to be greatest for those in thenegative consequences condition.The first study showed that exaggerated facial expressionsof pain appear to be an amplification of normal pain expression.However, changes in facial expression did not bring about changesin verbal report of pain perceived, calling into question thefacial feedback hypothesis.The results of the second study suggested that pain wasaltered only for subjects who prepared to state that their painfelt less painful than expected. This effect reachediiisignificance on the pain intensity visual analogue scale for lowintensity shocks. This effect did not carry over into theposttest phase, nor were negative consequences effective inamplifying the manipulation, leaving the theoretical mechanismunderlying the change in pain unclear. Moreover, the effect didnot vary amongst subjects who scored differently onquestionnaires measuring self—deception as a trait, presentanxiety, or adaptive coping strategies.The self—report measure of self—deception was related tofactors found to predict adaptive coping or good functioning inchronic pain patients, namely a sense of control over pain andthe absence of catastrophizing thoughts.One particular facial movement (brow lowerer) wasconsistently related to the verbal pain reports, attesting to thevalidity of facial expression as a measure of pain.The results are discussed with implications and suggestionsfor future research. One major problem with research involvingsubject deception is that subjects may appear to comply withexperimental instructions to deceive others while avoidingpersonal responsibility through a variety of mechanisms yet to bedetermined.ivTable of Contentslbstract iiList of Tables viiiList of Figures ixAcknowledgements xIntroduction 1Literature Review 4Facial feedback in chronic pain 7Cognitive—behavioral models of chronic pain 8Self—deception 13Self-deception as a coping style or strategy 15Cognitive dissonance 18Self—perception 21Impression management 22Complementary processes 23Attentional and memory biases 24Self—deception in pain: Previous studies 25An Integrated Model of Self-deception in the Genesis ofChronic Pain 27Pain Measurement 32Verbal pain measurement 32Facial expression of pain 34Statement of the Problem and Hypotheses 36Method 40Subjects 40Apparatus and materials 41Procedure 42vFacial Action coding.45Data Analysis: Overview 47Study 1 Results (N=43) 50Study 2 Results (N=60) 59The relationship between the dependent measures 65Discussion 72Discussion of Study 1 Results 72Discussion of Study 2 Results 76Future directions 79Summary 86References 90Appendices 110A. Visual Analogue Scales 111B. FACS Pain-related Action Units (AU5) 112C. Experimental Participation Consent Form:Studies 1 & 2 113D. Protocol pertaining to manipulation phase: Study 1 . 114E. Protocol pertaining to manipulation phase: Study 2 . 115F. Prepared Statements for Study 2 116G. Post-experimental Questionnaire: Study 1 117H. Post-experimental Questionnaire: Study 2 119I. Debriefing: Study 1 120J. Debriefing: Study 2 122K. Study 2. Manipulation phase questionnaire 124L. Study 2: Simple interactive effects analysesfor the Sensory Scale 125M. Study 2: Simple interactive effects analysesfor the Unpleasantness Scale 127viN. Study 2: ANOVAR Results with Self-Deception asa Grouping Variable (Sensory Pain Ratings) 1290. Study 2: ANOVAR Results with Self-Deception asa Grouping Variable (Sensory Pain Ratings) 130P. Study 2: ANOVAR Results with Catastrophizing asa Grouping Variable (Sensory Pain Ratings) 131Q. Study 2: ANOVAR Results with Sense of Control asa Grouping Variable (Sensory Pain Ratings) 132R. Sensory Scale ANOVAR Study 2 Cell Means 133S. Unpleasantness Scale ANOVAR Study 2 Cell Means 134T. Multiple Regression Results Using Self-Deceptionas a Moderator Variable 135viiList of Tables1. Experimental design across time 432. Standardized Canonical Discriminant FunctionCoefficients for the First Function 573. Classification Function Coefficients 574. Classification Results 585. Pain-related AU5 and Surprise AUs with respectivefactor score coefficients 656. Multiple Regression Results for the Sensory Scale Ratings 667. Multiple Regression Results for the Unpleasantness Scale 668. Multiple Regression Results for BIDR Self-Deception 70viiiList of Figures1. Model of the self-deception cycle in chronic pain 292. Predicted Pain Ratings Across Time 393. Condition X Time Interaction for the UnpleasantnessScale (Study 1) 524. Facial Activity for Three Intensity Levels AcrossTime (Study 1) 545. Condition X Time Interaction for Facial Activity (Study 1) 566. Condition X Time X Intensity Interaction for theSensory Scale, Lowest Two Intensity Levels (Study 2) 607. Condition X Time X Intensity Interaction for theUnpleasantness Scale (Study 2) 62ixAcknowledgementsThanks to the many students who participated in thisexperiment. Obviously, without you this study could not havebeen completed.Thanks to my supervisor, Kenneth D. Craig, for guiding methrough my years of research and schooling. You provided anatmosphere that fostered independence through allowing me tolearn at my own pace, while stepping in to direct me when Ifloundered.My committee members also provided valued assistance,especially when the research failed to follow my expectations.Dr. Del Paulhus, your critical thinking skills forced me towardsprecision and clarity in my thinking. Dr. Wolfgang Linden, yourconcrete suggestions about the research project helped todetermine the direction that it eventually took.And finally, my parents and friends deserve credit forproviding caring support whenever I needed. Mom and Dad werealways happy to hear of my progress, slow though it may haveseemed.x1IntroductionPain is a private experience that is difficult to define,explain, and measure. The biomedical disease model suggests toboth lay persons and physicians that pain results from underlyingtrauma or disease, and so patients seek appropriate diagnoses andpain cures from their doctors (White, 1990). In fact, itconstitutes the primary reason for visiting a physician in NorthAmerica (White, 1990). In the words of Proust, “Illness is themost heeded of doctors: to goodness and wisdom we only makepromises; pain we obey” (cited in Auden & Kronenberger, 1966).Unfortunately, health professionals can offer few effectivetreatments for many kinds of pain and a sizeable proportion ofthe population at large (about 11%) remains dogged by chronicpain complaints (Crook et al., 1986).Despite its prevalence, pain is poorly understood,especially when it lasts beyond the normal healing time or whenit strikes individuals who appear to have no underlyingphysiological pathology (Reesor & Craig, 1988). The term chronicpain syndrome has been reserved to classify patients whosecomplaints of pain are out of proportion with objectivepathophysiological damage either in intensity and/or in duration(most definitions specify a minimum duration of six months). Thebiomedical disease model fails to explain how pain can persistbeyond the normal healing time. In such cases, psychosocialvariables help to explain the appearance and maintenance of paincomplaints. For example, cognitions, emotions, and behavioralcontingencies can lead to pain complaints that are2disproportionate to bodily injury (Eisenberg, 1977; Reesor &Craig, 1987).. General biopsychosocial models of pain haveemerged which embrace physiological and psychosocial factors(Engel, 1977) largely as a result of Melzack & Wall’s (1982) gatecontrol theory of pain which demonstrated how psychologicalprocesses might enhance or dampen incoming pain signals.Because many psychological factors have merit in explainingthe etiology of chronic pain, they were included in Flor & Turk’s(1984) comprehensive “diathesis—stress model” which integrates“physical, psychological and social factors. . .to explain thedevelopment of chronic back pain” (p. 215). The diathesis orpredisposition to increased muscle tension in the back may resultfrom genetic influences, past injury and/or past learninghistory. Coupled with a stressful environment, low paintolerance, and/or poor coping responses, repeated episodes ofpain and tension may cause ischemia, release of pain chemicals,muscle spasm, and hence more pain. This constitutes the pain—tension cycle that has long been used to explain the genesis ofchronic pain in psychophysiological terms (Livingstone, 1943;Turk, Meichenbaum & Genest, 1983). Fear of increasing the painby physical activity leads to conditioned fear responses wherebythe patient learns to avoid activity and thereby suffers morefrom muscle tension and pain (Fordyce, 1989; Philips, 1987). Butalthough pain patients may have areas of increased muscletension, higher muscle tension is not correlated with higher painintensity (Arena et al., 1991). Therefore other constructs are3needed in comprehensive models that fully explain the origins ofchronic pain.Other approaches that have been utilized to explain theetiology of the chronic pain syndrome include the medically-basedpsychoanalytic viewpoint; the family systems approach; andpsychological approaches derived from social—learning, behavioraland cognitive-behavioral theories. A model is proposed in thisthesis which focuses upon a self—deception process to account forpersisting reports of pain. Investigations are presented thatexamine self—deception as a mechanism involved in the persistenceof reports of high levels of pain in the absence of adequatenoxious stimulation. As well, controls are included to examinealternatives to self—deceptive processes, which could alter one’sattitudes about pain and/or the experience of pain, includingfacial feedback mechanisms, cognitive dissonance, self—perception, and impression management. As a conceptual model, itis contended that self—deception may encompass components fromall of these competing theories. Moreover, self—deception mightrepresent an adaptive coping mechanism for dealing with pain.4Literature ReviewEmbracing the medical disease model, psychoanalytictheorists conceptualize persistent and unexplained pain as asymptom of disease——not physical, but mental. Pain is construedas signifying psychiatric illness such as “conversion neurosis,depressive equivalent, or hypochondriacal reaction” (Turk & Flor,1984, p. 210). The several editions of the medically—basedDiagnostic and Statistical Manual for Mental Disorders have onlyrecently dropped the term “psychogenic pain disorder” and havereplaced it with the term “somatoform pain disorder” (AmericanPsychiatric Association, 1987). This still differs from thepredominant term in current use by researchers and clinicians——“chronic pain syndrome”--and testifies to the reluctance thatsome authorities have in moving beyond psychoanalytic /psychiatric nomenclature and beliefs about pain. Researchevidence supporting this approach has been characterized asriddled with logical and methodological flaws (Turk & Flor,1984), not the least of which is a heavy reliance uponcorrelational methods that fail to determine causal direction.For example, to treat pain as a symptom of depression is nowconsidered simplistic (Blumer & Heilbronn, 1981; Large, 1986;Romano & Turner, 1985). Depression often accompanies pain(Craig, 1989; Romano & Turner, 1985), but it is reasonable tocontend that depression results from pain, especially prolongedpain that prohibits normal work and leisure activities (Fordyce,1976). Younger patients are more likely to suffer fromdepression than older patients (Keefe, Dunsmore & Burnett, in5press) perhaps because it interferes more with their highexpectations of long—lasting health, work and leisuresatisfaction. Moreover, pain intensity has been shown to predictlevel of depression in one longitudinal study of arthriticpatients (Brown, 1990). Part of the problem in differentiatingdepression from pain might be a measurement issue: manyquestionnaires that purport to measure depression include itemsrelated to reduced physical functioning which taps symptoms ofchronic pain (Wesley et al., 1990). Moreover, definitions ofdepression vary from study to study (Haythornthwaite, Sieber &Kerns, 1991). Anxiety also accompanies chronic pain in manycases (Jamison, Rock & Parris, 1988; Smith, 1986; Wade et al.,1990). It can be safely concluded that psychological distress isoften correlated with pain, but the relationship between thesevariables needs further longitudinal study to determine thedirection of causality.Chronic pain generally has an impact upon all aspects of anafflicted individual’s life, including family relations. In someinstances, family members have been said to promote chronic paincomplaints. For example, family systems theorists who examinedthe sick role in its familial context suggested that a family mayavoid conflict by focussing upon one member’s illness (Flor &Turk, 1984). Illness may take the form of pain if that is themost common physical complaint in the family’s history (Craig,1978, 1983; Turk, Rudy & Flor, 1985). Emotional pain mightreceive a somatic label in chronic pain families (Violon, 1985).Like the psychoanalytic approach to pain, this model is highly6speculative and relies upon correlational and case—studyevidence. Some of the constructs used to explain family dynamicsare difficult to measure or refute (e.g., enmeshment) and aretherefore not easily researched (Turk et al., 1985).Nonetheless, chronic pain patients usually have pain sufferers intheir families, more so than comparable medical patient controls(Violon, 1985). Thus, there is some link between family andchronic pain that deserves further study.The social learning / modelling approach suggests that inthe family and other social settings, we learn vicariously how tobehave when injured and suffering (Craig, 1978; 1983). Thisformulation helps to understand why an individual’s painbehaviors seem to correlate with the family’s history of paincomplaints (Violon, 1985). It also provides an account of howpain behaviors may be patterned within a specific culture throughobservational learning (Craig, 1978; Mechanic, 1961; Escobar etal., 1987).The behavioral approach (Fordyce, 1976) involvescontingencies provided by family members (and others) for painbehaviors. This formulation has been indirectly supported byrepeated research findings that behavioral treatment of painreduces disability and improves physical and emotionalfunctioning (Flor & Turk, 1984; Fordyce, Roberts & Sternbach1985). A central biopsychological assumption here has been thata cycle is established whereby pain leads to heightened musculartension which leads to avoidance of activity and further pain(Ross, Keefe & Gil, 1988; Turk, Meichenbaum & Genest, 1983).7Physical deconditioning occurs as the patient avoids activityformerly associated with pain (Fordyce, 1989; Philips, 1987).Fear of pain, anxiety and depression feed the cycle.Through the use of operant reinforcement, Fordyce (e.g.,1976) has successfully designed treatment programs to reduce“pain behaviors” but not necessarily pain ratings (Kerns et al.,1986; Prkachin & Cameron, 1990). Fordyce, Roberts & Sternbach(1985) defend their position by pointing out that their aim hasbeen to reduce disability not pain. Increased activity is shapedand rewarded thus extinguishing avoidance of activity based onfear of pain.Facialfeedback in chronic painIt is yet to be determined whether pain perception can bealtered by changing its social contingencies (Fordyce, Roberts &Sternbach, 1985). Altering one specific pain behavior—-facialgrimaces——might cause changes in pain per se. The facialfeedback hypothesis (Adelman & Zajonc, 1989) holds that changesin facial expression cause neurocirculatory changes from theskin/muscles of the face to trigger neurochemical alterations inthe brain. That is, facial expression might be a cause ratherthan a result of emotional experience (Adelmann & Zajonc, 1989;Izard, 1990; Kraut, 1982; Laird, 1984). Ongoing emotional tonemay be amplified or diminished by feedback systems from the faceto underlying systems. What seems to follow from this hypothesisis that if someone fakes more or less of an emotion, he/she mayactually come to experience an amplification or diminution ofthat emotion. A second corollary of the facial feedback8hypothesis would be that the system is dynamic, changing over thecourse of time, and therefore fluctuations in emotional tonewould be expected to quickly follow changes in facial movements.If applied to pain, this facial feedback model suggests thata change in facial expression causes a change in the painexperience. While this view is controversial (Buck, 1980; Izard,1990), it can be supported or refuted. One group of studies inthe pain literature supports some of its assumptions (Colby,Lanzetta & Kieck, 1977; Kieck et al., 1976; Lanzetta, CartwrightSmith & Kleck, 1976): male subjects exposed to shock-inducedpain when observed by others exhibited less facial expressivenessand reported less pain while their skin conductance responses- also were attenuated. It is uncertain whether the facialexpression preceded the pain experience, making causal directionunclear. Moreover, there has been some difficulty replicatingthese results (Kieck, personal communication). Even morecontroversy exists in attempts to define facial cues of deceptivecommunication (Cody & O’Hair, 1983), especially when genderdifferences are examined. Part of the problem is methodological:studies abound in which a detailed analysis of facial movementswas not used to define facial features (Matsumoto, 1987).Cognitive—behavioral models of chronic painPerhaps altering verbal behaviors and/or cognitions canalter one’s pain (Bayer, 1985). The cognitive-behavioralapproach to pain includes an “underlying assumption that affectand behavior are largely determined by the way in which theindividual construes the world” (Turk, Meichenbaum & Genest,91983, p. 4). Coping is a primary topic of concern to cognitive-behavioral therapists who wish to determine how coping affectspatients’ adaptation to chronic pain.Coping refers to a process whereby an individual appraises asituation as a stressor and then mobilizes thoughts and/orbehaviors to manage the impact of the stressor (Lazarus &Folkman, 1984). Lazarus and Folkman have divided the constructinto problem—focused and emotion—focused coping strategies. Butit may also include the apparent absence of strategies such asthe use of “denial” to cope with a situation (Tunks & Bellissimo,1988). Coping thoughts and acts vary across time and contexts,thus are best construed as processes rather than as styles ortraits (Lazarus & Folkmari, 1984).Much research has sought the identification of adaptivecoping mechanisms. Some researchers have noted few differencesin coping strategies used across a variety of chronic diseasessuch as hypertension, diabetes mellitus, cancer, and rheumatoidarthritis (Felton, Revenson, & Hinrichsen, 1984). But copingstrategy use interacts with factors within a situation. Forexample, appraisal of the situation and its meaning may determinethe type of cognitive strategy selected and its effectiveness(Lerman, 1987; McCrae, 1984). Or a more generally adaptivedisposition such as “optimism” may color subjects’ responses oncoping questionnaires making some people appear to use a widevariety of coping mechanisms when they really have a generalresponse bias such as an optimistic outlook (Scheier & Carver,1985, 1987).10Strategies used specifically for coping with pain have beenidentified with the Coping Strategies Questionnaire (Rosenstiel &Keefe, 1983)——a 44-item questionnaire with 7 subscales and tworating scale. Three major factors and two single—subscalefactors have emerged from the scale (Lawsen, Reesor, Keefe &Turner, 1990). The first factor refers to active cognitivecoping strategies (e.g., ignoring, coping self—statements,reinterpreting the pain). The second factor refers to a self—evaluation of the patient’s ability to control their pain. Thethird factor refers to passive strategies of coping (e.g.,praying, hoping, diverting attention). The subscales ofbehavioral coping and catastrophizing weighted heavily on thefourth and fifth factors, respectively.Prior research involving medically incongruent chronic lowback pain (Reesor & Craig, 1988) and osteoarthritic knee pain(Keefe et al., 1987a & l987b) indicated that a poor sense ofcontrol over the pain and catastrophizing thoughts were relatedto greater pain. Improvement in perceived control over pain hasbeen shown to predict six—month follow—up measures of pain reportin chronic low back pain patients (Spinhoven & Linssen, 1991).This finding underscores the. usefulness of treatments designed toimprove patients’ sense of control over pain.In a recent review of coping with chronic pain, Turner(1990) concluded that useful techniques emphasize active copingand avoidance of catastrophizing, while promoting a sense ofcontrol over pain. Attention diversion seems ineffective forsuch patients. Furthermore, flexibility of coping strategies11used across different settings or circumstances may be useful(Tunks & Bellissimo, 1988; Turner, 1990). One such circumstancemay be the pain severity (Blew, Patterson & Quested, 1989).Given the finding that a perception of control is useful,perceived coping self-efficacy may be a key ingredient toadaptiveness (Bandura et al., 1985; Dolce, 1987). For example,Devine and Spanos (1990) found that a variety of copingstrategies attenuated pain response, especially if the subjectbecame absorbed in their use as a result of positive self—efficacy expectations.As Turner (1990) noted, strategy usefulness may further varyas a function of time. In a meta—analysis of coping strategiesused over time for a variety of stressors (such as pain, stress,and anxiety), Suls and Fletcher (1985) concluded that 1) eitheravoiding (e.g., distraction) or attending to sensations (e.g.,monitoring sensations) was a good short—term strategy; and 2) thewhole class of attentional strategies (such as focusing on thesource of stress and physiological reactions) was associated withbetter long—term adjustment. Again, the use of any strategy wassuperior to using no systematic strategy.Attempts have been made to match treatment modes withpreferred coping style. Pain was reduced in oral surgerypatients who were given strategies consistent with their style ofcoping (Martelli et al., 1987) as measured by a modified versionof the Ways of Coping Scale (Folkman & Lazarus, 1980). Subjectswho preferred a high level of surgery information responded bestto a problem—focused strategy, while those who preferred a low12level of information responded best to an emotion—focusedstrategy (although this latter result was only marginallysignificant, p < .10). Laboratory subjects trained in copingwith cold pressor pain have also been shown to benefit from theuse of a wide range of coping strategies such as divertingattention, reinterpreting the pain sensation, engaging in fantasyincompatible with the pain context, and avoiding catastrophizingthoughts (Devine & Spanos, 1990; Miller & Bowers, 1986).A recent study demonstrated that treating catastrophizing asan individual difference variable helps to predict differentialtreatment response (Heyneman et al., 1990). Catastrophizerstended to benefit from instructions to alter self-statements(self-instructional training) during a cold-pressor task. Noncatastrophizers tended to benefit most from attention diversionstrategies.To summarize, a variety of psychological approaches havebeen offered as explanations for the persistence of chronic painin the absence of adequate pathophysiological explanation. Noneon its own can fully account for the genesis or maintenance ofchronic pain, but before attempting to integrate these into acomprehensive model, the construct of self—deception will beintroduced. It, too, provides a partial formulation of theprogression towards the chronic pain state (Bayer, 1985).13Self-deceptionThe concept of self—deception arose in the domain ofphilosophy in which self-deception has been formally defined asbelieving a proposition, p, and its opposite, not-p,simultaneously (Bach, 1981, 1985; Demos, 1960). Most definitionssuggest further that self-deception must be motivated’ (Bach,1981, 1985; Davidson, 1986; Paulhus, Fridhandler & Hayes, inpress; Pears, 1986; Sackeim & Gur, 1978); for example, self—deception results in the manipulation of one’s emotional state(Silver, Sabini & Miceli, 1989). The construct of motivation isnecessary to separate self—deception from akrasia2, biasedthinking, wishful thinking, or mere ignorance. While motivated,self—deception is not thought to be intentional (Bach, 1981; Bok,1980; Mele, 1987), which further differentiates it from wishfulthinking. The definition used in the present study is that self-deception is a process that involves a contradiction betweenone’s words or attitude and behavior and is motivated by theprotection of self—esteem.To think and/or behave in two opposing directions at thesame time is a paradox or contradiction which many writers haveattempted to resolve. Some philosophical viewpoints challenge1 For a cognitive functionalist view of motivation, see Weiner(1986) who provides empirical evidence supporting his view thatemotions, many of which arise out of attributions of causality,precede actions. Thus, we typically think, feel, and then act:emotions provide the motivating force for actions.2 Akrasia may be defined as voluntarily and consciously actingagainst one’s better judgement in a particular situation when thepreferred course of action is psychologically weak. An akraticaction might be that which is more salient, or more habitual, orthat which peers are doing. (Rorty, 1980)14the logical possibility of self-deception whereas others blendinto psychological explanations. For example, Bok (1980) hasargued that self—deception is merely a semantic problem and thatbetter words for the phenomenon might be “bias, rationalization,and denial” (p. 924). Or self-deceivers might be “twisting” theevidence in a way that remains rational, but is self—deceptive(Szabados, 1985). Perhaps what appears to be self-deception ismerely behavior being emitted out of habit (Chanowitz & Langer,1985; Rorty, 1986). Self-deception may occur by repetitive“jamming” of our thoughts until we believe that which we wish tobelieve (Bach, 1981). Behaviors that vary across time orsituations may appear self-deceptive but may still be internallyconsistent (Airislie, 1986; Chanowitz & Langer, 1985). Althoughdiscussed by philosophers, these latter explanations becomeacceptable to behavioral psychologists since there seems to beagreement that persons emit habitual, repetitive behaviors whichare under the control of situational stimuli. For example,Fordyce (1989) is a behavioral psychologist who has developed acognitive—behavioral model of chronic pain that emphasizes therole of learning and conditioning in the genesis of pain andsuffering behaviors.15Self-deception as a coping style or strategyCognitive-behavioral views of coping intuitively seemrelated to self—deception. Self—deception may be a componentunderlying some kinds of coping such as the denial/avoidantstrategies (Lazarus, 1979; Power, 1984). Self-deception may beviewed as a health-enhancing coping style or strategy if itallows a person to muster up emotional strength before facing adistressing situation. In some cases, self—deception mightpromote a sense of control, self—efficacy, or optimism aboutresolution of disease symptoms such as pain. It may allow apatient to feel hopeful in the face of a crisis, especially ifthe information about the crisis is ambiguous (Lazarus, 1979) asit so often is with pain problems-. Catastrophizing would seem tobe a special case of self—deception: a process that promotes aworsened state of the self. This is all speculative since theconstructs have yet to be compared systematically, although suchan undertaking is planned (Paulhus, personal communication).Thus, descriptively and at a practical level, self—deceptionmay turn out to be a useful coping mechanism in some cases.Whether the self—deception construct proves useful theoreticallyor to have supportable clinical applications is yet to bedetermined.The present review proposes that pain report may be alteredby self—deception. Social or clinical circumstances may dictatea pattern of self—representation that leads to self—deception.Exaggerating or diminishing one’s display of pain in order todeceive others may also deceive the self. To illustrate, the16patient who believes that his/her pain is diagnostic of seriousdisease might exaggerate the pain report to ensure that medicalprofessionals thoroughly test for all possible causes of the painor otherwise provide the best possible care. Such a patientwants to put his/her “worst fears to rest” and might visit manyspecialists (Blackwell & Gutmann, 1986). As the patient repeatshis/her pain story, a cycle of persistent pain report is set inmotion until he/she is assured that cancer, arthritis, or otherserious diseases have been ruled out. By then, he/she may havebecome fully convinced about the persistence of the painexperienced. The complaints of pain have become habitual.An extreme conflict between the beliefs of patient andmay also entrench a patient’s pain complaints. Avariety of scenarios is plausible. Physicians or others mightsuggest that a patient has a psychiatric disorder--Chronic PainSyndrome (or “somatoform pain disorder” in the DSM—III—R) ——whenthere are no or insufficient organic signs of underlying physicaltrauma to account for the pain complaints. The patient whosenses that this label means that the physician views the pain as“all in the head” might desperately attempt to find a specialistwho can correctly find the cause of very real pain. Moreover,the chronic pain patient often searches for the “correct”diagnosis that will lead to subsequent cure of pain (Kotarba,1983). Thus, complaints of pain might become firmly entrenched:to stop complaining might cause a “loss of face” when one’sphysical and/or mental health is questioned.17The opposite scenario is equally plausible: hiding one’spain might come to decrease it in the long run. For example,blue collar workers and professional athletes fear losing theirjobs if they take sick time or complain of inability to carry outphysically taxing activities (Kotarba, 1983). Others may hidetheir pain from family members in order to decrease their worryand to preserve the family system with roles kept intact(Kotarba, 1983; Tunks & Roy, 1990). Perhaps these people becomeexperts at hiding pain from others and thus from themselves.Either extreme of pain reporting can have deleteriouseffects on a patient’s functioning. Exaggerated pain complaintshand in hand with decreased physical activity and avoidance ofwork activities that ordinarily build self-esteem, fuelling acycle that can lead to chronic pain (Ross, Gil, & Keefe, 1988).The patient’s pain may worsen as a result of weakened muscles andweakened spirits. Increased medication use may lead to abuse ordependency, and subsequent attempts at reducing self—medicatingcan trigger increased pain complaints. At the opposite extreme,the patient who inappropriately avoids accepting the sick rolemay prolong recovery from acute trauma. Work efficiency maysuffer along with increased stress and irritability.If we could pinpoint the underlying psychological andphysiological processes involved in such self—deception, we mightbetter understand some cases in which pain seems out of keepingwith medical findings. Moreover, we might bolster discoveriesthat some pain patients can be treated by having them reducetheir pain complaints. This would be consistent with and support18the Fordyce (1976; 1989) behavioral approach to pain managementwhich has been accused of treating the symptoms but not theunderlying pain (Prkachin & Cameron, 1990). If changing painbehaviors does alter the underlying pain, then patients couldclearly benefit from treatment aimed at reducing displays ofpain. In short, the better we understand pain behavior, thebetter equipped we should be to change it and/or itspsychological concomitants.Other psychological viewpoints of self-deceptive phenomenato be discussed next include cognitive dissonance, self—perception, impression management, and attentional and memory(Bayer, 1985). It is contended that a combination ofthese processes brings about self—deception. These will beincorporated into a model accounting for some kinds of chronicpain complaints.Cognitive dissonanceCognitive dissonance provides an avenue for understandingself-deception. If a person is induced to publicly engage in a“behavior that is inconsistent with their private attitude and ifthey have insufficient justification for the inconsistentbehavior, dissonance will occur” (Penner, 1986, p. 427). Thisfosters a state in which discrepant cognitions (beliefs,attitudes, ideas or thoughts) are smoothed out, made moreconsistent. “Dissonance is like a negative drive; people aremotivated to reduce or eliminate it” (Penner, 1986, p. 427).For example, cognitive dissonance might occur if a personwho believes that he/she is good performs a bad act such as19telling a lie to others (Gilbert & Cooper, 1985; Snyder, 1985).There is a discrepancy or dissonance between these cognitions ofgood- v. bad-me (Festinger, 1957). The person becomes motivatedto reframe the act as less negative or to decrease his/her senseof responsibility for the act. In other words, the person wouldengage in “excuse—making” while filtering out discrepant data,thereby enhancing the self-esteem (Snyder, 1985) and bringingabout self-deception.Older versions of cognitive dissonance theory held thatsimply doing or saying something discrepant with one’s beliefs,with low justification for doing so, would arouse dissonancewhich could lead to attitude change. In the forced compliancemanipulation, subjects who performed a boring task were asked totell the next subject (actually a confederate) that the task wasinteresting. Subjects given a small reward ($1) for engaging inattitude—discrepant behavior came to rate the task as moreinteresting than subjects in a large reward condition (paid $20)or in a control condition.Recent versions of cognitive dissonance theory add that thenecessary triggers of change in attitude are: 1) that thesubject freely chooses to deceive others and so cannot attributeresponsibility to another person or to circumstances beyondhis/her control; and 2) that there are foreseeable, negativeconsequences of this deception (Cooper & Fazio, 1984; Scher &Cooper, 1989; Schlenker, 1982). Whether these are indeednecessary is debatable. They may exacerbate dissonance20(Berkowitz & Devine, 1989) or they may increase the unpleasantthreat to self-esteem (Hilgard, 1949; Steele, 1988).Fazio (1986) further contends that the consistency betweenattitudes and resultant behavior should improve if one’sattitudes are developed through direct experience and/or if theattitude is made salient (due to “priming” or other means offocusing attention on attitudinal issues). From this it followsthat if one develops a negative attitude about pain (e.g., fearof chronic pain and disability), any sensation that seemssomewhat painful might become more salient, especially if oneactually vocalizes one’s complaints to family members or thephysician. In Fazio’s (1986) model, such a salient (andnegative) attitude developed through past pain experiences shouldmake a patient’s behaviors consistent with the attitude: fear ofpain should consistently lead to behaviors indicative ofavoidance of pain (e.g., guarding, physical inactivity).Zimbardo et al (1969) have examined cognitive dissonance andpain tolerance. Subjects agreed to accept further electricshocks for a questionable purpose, after already completing anexperimental protocol. This was expected to arouse dissonanceand thereby alter subjects’ pain tolerance. That is, “agreeingto be shocked is not so irrational if one expects that the‘shocks will not hurt as much this time” (p. 109). Theresearchers found only tentative support for this.Competing theories that offer explanations of forcedcompliance phenomena include self—perception and impressionmanagement (Cooper & Fazio, 1984; Paulhus, 1982).21Self-perceptionBy self—perception is meant more than observing one’s ownpast behaviors such as in the case of self—modelling where pastexperiences with painful stimuli are said to predict futurereactions (Craig, 1980). Instead, self-perception refers toinstances of observing and subsequently making inferences fromone’s own behavior when that behavior is out of character(Bandler, Madaras & Bern, 1968; Bern, 1972). If internal cuesabout one’s internal state are “weak, ambiguous, oruninterpretable”, then the person relies on external cues toinfer the internal state, in the same manner that an outsideobserver makes inferences (Bern, 1972). Add motivation to weightthe observations in one direction and the result is self—deception.In a test of the self-perception model (Bandler, Madaras &Bern, 1968), subjects undergoing shock-induced pain trialsobserved themselves either escaping or not escaping the shocks.Subjects were instructed that they should press a button to endthe shock when the red light came on, constituting the escapecondition. In the no—escape condition signalled by a greenlight, subjects were told that they should not end the shock bypressing the button. The researchers found support for thehypothesis that the perception of painfulness is partiallydetermined by one’s inferences regarding his/her own behavior:Those subjects who saw themselves escape from the shocks ratedthe felt discomfort higher than subjects in the “no—escape”condition, although the shock intensities were the same in both22conditions. It seems that subjects deceived themselves about theamount of pain perceived, since subjects should have rated theshocks as equally intense in the two experimental conditions(Bayer, 1985). The “no—escape” subjects observed themselvestaking more shocks and presumably concluded that the shocks wereless severe, otherwise they would have appeared odd indeed foragreeing to suffer.Impression managementThere is an alternative explanation for the above findingthat subjects rated the pain as higher in the escape conditionthan in the no—escape condition: The report of reduced painmight have been motivated by a desire to protect the outwardappearance of consistency in the presence of a respectedexperimenter——impression management.Some researchers have postulated that social anxiety mightmediate the forced compliance effect (Tedeschi & Rosenfeld,1981). But this “assertion that an individual is aroused as aconsequence of having freely brought about an aversive eventmakes it very difficult to find many lingering points of contrastbetween impression management and dissonance” (Cooper & Fazio,1984, p. 252). Schlenker (1982) emphasizes the importance ofprotecting one’s self-identity as a motivating force inimpression management which makes impression management seem muchlike self-deception.Questionnaires designed to measure impression managementshould elicit different responses in public versus privatedisclosure settings (Paulhus, 1984). That is, the construct of23impression management represents deception of others much morethan self-deception (which is not expected to vary considerablyacross settings).Complementary processesCognitive dissonance has been differentiated from self—perception (Bern, 1972) in that the former is thought to occurwhen behaviors are clearly discrepant with one’s attitude, whilethe latter occurs when behaviors are more congruent with one’sattitude (Cooper & Fazio, 1984; Fazio, Zanna & Cooper, 1977).Impression management is a similar construct whose boundariesoverlap with various so—called intrapsychic phenomena (Tetlock &Manstead, 1985) including cognitive dissonance and self-- perception. Paulhus (1982) has suggested that a combination ofcognitive dissonance and impression management leads to attitudechange in forced—compliance situations. This underscores thesuggestion made by Tetlock and Manstead (1985) that the areaneeds an all—encompassing theoretical framework which capturesthe range of findings related to cognitive dissonance, self—perception, and impression management. The contention in thecurrent thesis is that one such unifying construct that underliesthese concepts is self—deception.Self—perception, impression management, and cognitivedissonance all protect self—esteem (and thus are motivational)and all imply some contradiction between one’s words or attitudeand behavior. Therefore, they fit the general definition ofself—deception. These cognitive-behavioral perspectives suggestthat the self-deceiver “weights the data” in favor of proposition24p (Mele, 1987). Our beliefs of p and not-p are probabilisticrather than dichotomous; there is really no paradox to resolve(Mele, 1987). The manner in which we process information makesthis possible.Attentional and memory biasesIn situations that are ambiguous and allow for variousinterpretations, we might bias our beliefs in such a way as topromote a positive self—image or to protect self—esteem (Demos,1960; Gilbert & Cooper, 1985; Rorty, 1986; Silver et al., 1989;Szabados, 1985; Weiner, 1986). We may engage in “selectiveattention” that confirms our positive self-image (Gilbert &Cooper, 1985; Rorty, 1986; Snyder, 1985). Memory studies haverevealed that simpler events are recalled better than complexevents (including feeling states) under high arousal (Paulhus &Suedfeld, 1988). What this means for self-deception is thatmemory processes may serve to bias recall, especially inambiguous, complex, and emotionally—arousing situations (Paulhus& Suedfeld, 1988; Schelling, 1986; Wilson, 1985)One such complex situation might occur when uncomfortableand/or distressing physical sensations remind someone of the lasttime he/she suffered from pain. Physical sensations are oftenambiguous and their labelling may be shaped by modelling andlearning in childhood (Craig, 1980). They may be misinterpretedas painful if past experience has led to reinforcement for actingout one’s suffering (Fordyce, 1989) or if alternativeperspectives are unavailable (Turk, Meichenbaum & Genest, 1983).Moreover, the greater one’s distress, the greater the likelihood25of “mislabeling ambiguous signals as indicators of threat orharm” (Fordyce, 1989, P. 58). Chronic patients have been shownto blur the distinctions between painful and nonpainfulsituations, and may interpret a wide variety of experiences,especially affective distress, in pain terms (Clark & Yang, 1983;Pennebaker, 1982; Pennebaker & Skelton, 1978; Yang et al., 1983).Medical professionals fuel the fear of pain by directing patientsto exercise to the point of pain; this leads to a “failure todistinguish hurt from harm” (Fordyce, 1989, p. 58; Philips,1987). And so one’s fears and/or emotional distress cause one tocautiously respond to ambiguous sensations by labelling them aspain, in order to avoid harming oneself. The cognitive—behavioral treatment approach to pain aims to change patients’misperceptions and misconceptions about pain (Turk, Meichenbaum &Genest, 1983) which further attests to the centrality ofpsychological processes involved in altering chronic pain.Self-deception in pain: Previous studiesAt least two previous laboratory studies have examined theeffect of self—deception on pain expression. In the first,Quattrone & Tversky (1984) told subjects that a diagnosis oflongevity was associated with either increased or decreased paintolerance after physical exercise. Their tolerance to coldpressor pain was measured before and after exercising on astationary bicycle. Most subjects tolerated more or less pain,in keeping with their self-diagnosis about their longevity. Thiswas likely a self—deceptive process because most of the subjectsdenied altering their pain tolerance. In the authors’ terms,26self—deception was defined as choosing an action that isdiagnostic of a positive outcome without realizing that they“purposefully selected the action in order to make the diagnosis[of longevity]” (p. 240).In a different study (Jamner & Schwartz, 1986), paintolerance was positively correlated with the degree to whichsubjects generally deceived themselves, as measured by the LieScale of the Edwards Personality Inventory. In addition,subjects who scored high on self—deception required greater shockintensity levels before reaching the same self—reporteddiscomfort and pain levels as subjects who scored low on self—deception. Further, pain sensation threshold was not affected bysel-deception; only affective qualities of pain were. However,further study is necessary before conclusions can be drawn aboutthe role of self-deception in pain tolerance and self—report.For example, the conclusions of the latter study are uncleargiven Paulhus’s (1986) finding that the EPI Lie Scale measuresimpression management rather than self—deception.27An Integrated Model of Self-deception inThe Genesis of Chronic PainGiven the wide array of definitions and theoreticalformulations of self—deception, it is not surprising that no oneaccount of how or why self—deception comes about has beengenerally accepted. No doubt several variables bring about self—deception, including biological, cognitive, and socialpsychological processes (Hyman, 1989). The model of self-deception in chronic pain proposed here (depicted in Figure 1)emphasizes cognitive—behavioral psychological processes andintegrates several of the key positions discussed earlier. It isan elaboration of a model originally formulated by Bayer (1985).If the situation at hand is ambiguous so that the ground isfertile for misinterpretation, then the first phase of the self-deception model can take place. First, the subject engages indeceitful behavior by stating an untruth. For example, unusualphysical sensations are often ambiguous and threatening, so thepatient may complain of pain while unsure whether the sensationsconstitute pain or just fatigue, tightness, stiffness, warmth,etc. Many pain patients have no idea about what their diagnosisis, nor do they understand the underlying processes involved inpain perception (Rowat & Jeans, 1989), and so unusual sensationsbecome feared as signs of impending disease or suffering(Fordyce, 1989). If repeated complaints are seen by the patientto be of questionable accuracy or to be selfish pleas forattention, he/she becomes aroused and uncomfortable with thisnegative view of the self. Cynical or displeased medicalpersonnel may feed this negative view, especially when medical28interventions fail to achieve a cure (Kotarba, 1983) and theattending physician overtly or covertly suggests psychologicalcauses for the chronic condition (Blumer & Heilbronn, 1981).Next, arousal and discomfort arise upon the discovery ofinformation about the self that is discrepant with one’s positiveview of the self (Festinger, 1957; Higgins, 1989). Such out—of-character information would result from engaging in contemptiblebehaviors that society has taught us are socially undesirable orhave aversive consequences to others. This information threatensself—esteem which provides motivation to decrease the aversivearousal (Higgins, 1989). This heightened emotional arousal isassociated with reduced cognitive complexity (Paulhus & Suedfeld,1988; Schelling, 1986; Wilson, 1985) and memory and perceptualbiases which aid in the filtering process whereby the individualfocuses upon self—confirmatory information (Lockard & Paulhus,1988). In the pain example, heightened emotional arousaltriggers a decrease in cognitive complexity so that theevaluative dimension of pain becomes most salient (e.g., “Thispain is awful; it’s really bad”). Fleeting weak and ambiguoussensations are specifically attended to and interpreted asconfirmations of the pain complaint. The lack of a physicaldiagnosis feeds the cycle of increasing emotional distress, pain,and catastrophizing thoughts (Flor & Turk, 1984; Kotarba, 1983).Finally, we have a situation whereby unintentionally biasedinterpretation of the events has served to preserve self—esteem.The pain patient protects his/her self-esteem through a selfdeceptive process: “this pain is very serious”. He/she29continues to seek a definitive diagnosis by visiting more andmore specialists, hopeful that the “right one” will diagnose andtreat the formidable pain (Kotarba, 1983). And thus thecomplaints become chronic, entrenched beliefs that the pain isvery real and indicative of undiagnosed disease.Figure 1.Model of the self-deception cycle in chronic pain.1.) Ambiguous, threatening sensations come to be labelled as“pain”Tell an untruth = contemptible behavior“I have this terrible pain again. Comfort me.”2.) Arousal & emotional discomfortReduced cognitive complexityMemory & perception help one to attend to selfconfirmatory information3.) Biased interpretation of events such that self—esteem ispreserved“I am not a liar; my pain is very real and not in myhead.”= Self-Deception30As will be seen in the next section, affective qualities ofpain are generally more amenable to psychosocial interventionsthan sensory qualities. Thus, under conditions of affectivearousal, it is reasonable to predict greater possible alterationsin the evaluative and/or affective dimensions of pain than in thesensory dimension.To explore this model of self—deception, a laboratoryanalogue of pain with healthy volunteers was used in the presentstudy. While such a population may not provide directgeneralizations to chronic pain patients, it would beinappropriate to ask pain patients to deceive others (andthemselves) about the amount of pain that they were experiencing.Furthermore, precise control over the repeated pain stimuliallows for a direct analysis of effects over time resulting fromthe independent variables of interest. As well, it becomespossible to determine if social conditions produce and maintaineither an extraordinary or a diminished pattern of pain responseif there is already available a baseline assessment of thepatterns of response to controlled stimuli. These conditions arenot available in the clinical environment. Shock stimuli providea means of inducing pain that is similar to chronic pain in atleast one major fashion: both kinds of pain induce distress(Gracely, 1989) which will be quantified using a standard measureof state anxiety--the State-Trait Anxiety Inventory (Spielberger,Gorsuch & Lushene, 1970).Moreover, deliberate dissimulation of pain expression in thelaboratory might provide clues to the diagnosis and underlying31processes of such behavior in pain patients. The issue of howbest to measure pain has been a major topic of interest to painresearchers and so deliberate dissimulation affords insight intothis important measurement issue.32Pain MeasurementThe first step to the study of pain is to define and measureit. The International Association for the Study of Pain (1986)has defined pain as “an unpleasant sensory and emotionalexperience associated with actual or potential tissue damage, ordescribed in terms of such damage” (p. S217). This definitionindicates that at least two major dimensions to pain can bedifferentiated: a sensory/intensity dimension and an affectiveor unpleasantness dimension. Also documented in the literatureis an evaluative dimension whereby the individual attachesmeaning to pain (Melzack, 1984). For example, labour pain isevaluated more positively than pain resulting from accidentalinjuries. Measurement strategies have included primarily verbalquestionnaires and visual analogue scales while nonverbalmeasures have been less frequently utilized, especially inlaboratory studies of pain (Philips, 1983).Verbal pain measurementSensory and affective qualities of pain can be quantifiedusing two sets of ratio—scaled adjectives developed by Gracely,McGrath, and Dubner (1978a). An alternative method of measuringthese two pain dimensions is to use visual analogue scales thatare highly correlated to the two 15-word rating scales (Price etal., 1983).Affective and sensory qualities of pain are independentlyaltered by various treatments and conditions. The Gracelygroup’s ratio scales differentially reflect pharmacological paincontrol agents (Gracely et al., 1978b, 1979; Heft, Gracely,33Dubner, & McGrath, 1980) as well as a variety of laboratorymanipulations (Craig & Patrick, 1985; Hyde, 1986; Lee, 1985;Patrick et al., 1986). Research evidence further suggests thatthe affective dimension of pain is affected by one’s desire toavoid pain, expectations about whether pain will occur, and howone attends to the pain stimuli (Price, Barrell, & Gracely,1980). That is, affective ratings of noxious heat stimuli weredecreased at low intensity levels if the stimulation wassignalled and hence expected. The warning signal did not altersensory ratings of pain.Patients with a high degree of threat to health or life--cancer patients and chronic pain patients——rated their painhigher on the affective scale than on the sensory scale (Price,Harkins & Baker, 1987). Patients with low threat to health orlife——labour pain patients and experimental pain subjects——ratedtheir sensory pain as higher than their affective pain. Aseparate study confirmed these findings with chronic painpatients: chronic pain patients and those with rheumatoidarthritis were shown to be similar in rating their sensory pain,but the former group rated their affective pain higher than thelatter group (Gaston-Johansson et al., 1985). Hypnotic analgesiaalso has a greater effect on affective than on sensory pain(Price & Barber, 1987).Together these studies suggest that the affective qualitiesof pain are more amenable to psychosocial intervention thansensory qualities. Not surprisingly, for lower-intensity painstimuli there is more potential for reduction of affective34responses than for higher-intensity stimuli both in thelaboratory (Gracely, McGrath, & Dubner, l978b; Price, 1984;Price, Barrell & Gracely, 1980) and in clinical settings (Meizack& Wall, 1982)Facial expression of painBecause verbal reports of pain and other mental processesare easily biased (Craig & Prkachin, 1980; Gracely, 1989; Nisbett& Wilson, 1977), confirmatory nonverbal measures have long beensought in the measurement of pain. One promising avenue is themeasurement of facial activity in response to pain.Slow—action feedback of videotaped reaction patterns affordsthe source data for scientists to study facial movements in asystematic manner. Ekman and Friesen (1978, 1982) developed a“descriptive measurement system” based upon the visible musclemovements of the face. This system, called the Facial ActionCoding System (FACS), is objective and has been shown to be areliable and valid tool in the differentiation of emotions(Ekman, Friesen, & Ellsworth, 1982). With regard to validity,Ekman and Friesen (1982) reported that facial action intensitywas “highly correlated with [facial] EMG readings (Pearson R =.85)” (p. 208). Thus, FACS is highly sensitive in discerningsubtle movements formerly thought to be quantifiable only byphysiological assessment.FACS has proven valuable in the measurement of clinical andexperimental pain (Craig & Patrick, 1985; Craig & Prkachin, 1983;Grunau & Craig, 1987; Grunau, Johnston & Craig, 1990; LeResche &Dworkin, 1988; Patrick, Craig & Prkachin, 1986; Prkachin &35Mercer, 1989; Swaim & Craig, 1991). Facial expression ispositively correlated with self-reported affective qualities ofpain (LeResche & Dworkin, 1988) as well as with sensory intensityqualities of pain (LeResche & Dworkin, 1988; Swalm & Craig,1991). Facially expressed pain was not correlated with self-reports of anxiety and depression (LeResche & Dworkin, 1988);thus facial expression seems to reflect an aspect of pain and notgeneral affective disturbance such as anxiety or depression.In summary, pain is a multifaceted experience that can bemeasured verbally (using visual analogue scales for sensory andaffective qualities) and nonverbally (using facial activity).The affective dimension of pain seems more amenable topsychosocial interventions than the sensory dimension, more so inclinical pain subjects than in experimental subjects.36Statement of the Problem and HypothesesThe present study examined the impact that “telling a lie”had on subsequent verbal and nonverbal pain expression. Theself-deception model predicts that such behavior should elicitpersistently altered pain report, especially if the inducement tolie were poor; the effect should be greatest for subjects whofeel responsible for negative consequences of voluntarilyagreeing to lie.The forced compliance paradigm (Festinger, 1957) was used tocreate a situation conducive to self—deception. Studentvolunteers received shock—induced pain stimuli which they ratedfor felt intensity and affective discomfort. An initialascending series of shocks was administered to determine eachsubject’s pain threshold and tolerance levels. Then five shocksfrom the threshold to tolerance range were given in a randomorder, constituting the pretest baseline phase. Next, subjectswere randomly assigned and exposed to the experimental conditionsand they underwent another series of five shocks. Themanipulation was to deceive others about their pain experienceeither via a nonverbal route in Study 1 (facial expression) orvia a verbal route in Study 2. A control group in each studyreceived no instruction to lie. Half were told further thattheir deceptive behavior would cause negative consequences forothers who would view their videotape.Misleading fellow students about the pain experienced shouldinduce subjects to feel badly; this should motivate them tochange their attitude about the pain experienced. Thus, their37subsequent pain experience was expected to change in thedirection of the lie told if self-deception resulted from other-deception. The durability of this effect was examined in aposttest phase comprising five more shocks.Figure 2 outlines the different results predicted by thedifferent models. The facial feedback hypothesis suggests thatchanges in facial expression during the manipulation phase wouldcause temporary changes in pain ratings. The self—perception andimpression management models suggest that changes in pain ratingswould occur during the manipulation phase (other deception) withself—deception taking place over time to induce changes after themanipulation. The model of forced compliance suggests that theaddition of negative consequences would amplify this effect.Hypotheses:1. “Telling a lie” as a result of instructions toexaggerate or minimize one’s pain will lead to increased /decreased pain expression (verbal and nonverbal) in themanipulation phase.2. This increase or decrease will persist through theposttest phase.3. The effect will be most marked for subjects told thattheir behavior would have negative consequences for others.4. Because the affective component of pain seems generallymore reflective of psychological manipulations than sensoryfeatures of pain, the affective pain report will reflect thegreatest experimental effects.385. Facial expression will change significantly whensubjects are told to exaggerate or diminish the expression ofpain.6. Altered facial expression will persist during theposttest phase.7. The self-deception effect will be greater in subjectswho score high on questionnaires measuring self—deceptive style,state anxiety, and adaptive coping (absence of catastrophizingand presence of a sense of control).8. Self—deception will correlate positively with avoidantor denial coping styles and with a sense of control over pain.9. Self-deception will be negatively correlated withcatastrophiz ing.BSelf-PerceptionorCImpressionManagementDCognitiveDissonance:ForcedComplianceFigure 2Predicted Pain Ratings Across TimePretest Manipulation Posttest39AFacialFeedbackHypothesis0)a:C0S.’////‘S /LegendExaggerate— —— DiminishControlCU-Ca:Ca-‘S5%5%’..Negative ConsequencesNo Consequences,. No Consequences— Negative ConsequencesTIME40MethodThis section outlines the methodology developed to addressthe hypotheses. The effects of altering one’s expression of painduring a series of painful electric shock stimuli were examinedwithin a forced compliance paradigm. Experimental subjects wereinduced to exaggerate or diminish their pain expression, with orwithout negative consequences to others. A control group wasexposed to shock trials without instructions to alter their painexpression, also with or without negative consequences.Study 1 subjects exaggerated or diminished their facialdisplay of pain, purportedly to be judged by data coders intraining. The negative consequence condition suggested thattrainee coders who failed to reliably distinguish suchalterations in pain expression would not be hired.study 2 subjects exaggerated or diminished their verbaldisplay of pain, purportedly to be observed by subjects in asubsequent study. The negative consequence condition furthersuggested that such viewing would cause worse pain for theviewers.SublectsFemale UBC student volunteers from the departmental subjectpool received course credit or pay ($6.00) for theirparticipation in the study. 96 subjects participated in thefirst study, 60 subjects in the second study.41Apparatus and MaterialsThe pain stimulus consisted of electric shocks presented inaccordance with standardized procedures (Tursky, 1974) and asused in previous studies in this laboratory (Craig & Weiss, 1971;Craig & Prkachin, 1978; Patrick, Craig & Prkachin, 1986; Swalm &Craig, 1991). A Farrall Instruments, Inc. Mark-300 AversiveConditioner was adapted for use as the shock generator. Thepower supply was changed from a variable voltage source to avariable resistance source by adding a 5000 Ohm resistor, whichmakes the unit operate as a constant current source and minimizesthe effect that individual variations in skin resistance mighthave on the overall level of current delivered. Maximum currentdelivered was 14 milliamperes. The volar surface of the non—dominant forearm was rubbed with Redux paste before attaching theshock electrode. Standard videotape equipment was contained inthe experimental room, hidden from view by a one—way mirror.Visual analogue scales (VAS) were used to measure theaffective and sensory qualities of the pain as outlined by Priceet al. (1983). Subjects completed the State-Trait AnxietyInventory (state version, Spielberger, Gorsuch & Lushene, 1970),the Balanced Inventory of Desirable Responding (BIDR, Paulhus,1984), and the Coping Strategies Questionnaire (CSQ, Rosenstiel &Keefe, 1983) after the first (ascending) series of pain stimuli.After the final series of stimuli, subjects completed a postexperimental questionnaire designed to elicit suspicionsregarding experimental hypotheses as well as feelings of42responsibility and discomfort regarding their dishonest behavior.(See the Appendices for the questionnaires.)ProcedureThe entire procedure required less than sixty minutes. Thesubject was escorted to a laboratory room and seated at a table.The general nature of the study, procedure, expected duration,the delivery of electric shocks, use of videotape equipment, andrate of pay were explained. The shocks were described asbeginning at very low levels and increasing in small incrementsin the first series. The subject was informed that she was tosay “stop” when she wished to accept no further increases andshocks to be presented in three subsequent series of 5 shockseach were selected from this range. She was informed of theright to withdraw at any time or to refuse to answer anyquestions without loss of remuneration; confidentiality wasassured. A consent form outlining this information was signed.Subjects were initially told that we were gathering dataabout pain expression from healthy people in order to compare itto pain patients, such as back pain or cancer pain patients. Thehypotheses were not explained until a full debriefing at the endof the experimental session.A stack of pages with 10 cm. sensory and affective VAS5 wasshown to the subject who was instructed to listen to tape—recorded instructions regarding their use. The distinction madebetween the sensory and unpleasantness aspects of pain wasderived from Price et al. (1983) who drew an analogy between painand sound by suggesting that sensory aspects of pain are like the43loudness of a sound whereas unpleasantness depends on intensityas well as other factors. The sensory scale was anchored at oneend by “the most intense pain sensation imaginable” and theunpleasantness scale was anchored by “the most unpleasant feelingimaginable”.An overview of the experimental design is depicted in Table1. All subjects were subjected to an ascending series of stimuliusing the psychophysical method of limits to determine tolerancelevels. Then experimental subjects underwent three series offive shocks each: 1) pretest, 2) diminish (D) or exaggerate (E),and 3) posttest series. A control group received three likeseries. These will now be described in detail.Table 1. Experimental design across timeASCENDING PRETEST MANIPULATION POSTTESTSERIESExaggerate (E)*or• *Diminish (D)or*Control (C)* Half of the subjects were told that their behavior would havenegative consequences to subsequent viewers.To establish the subject’s pain threshold and tolerancelevels, pain stimuli were administered in increasing intensities44(by 0.5 mA steps) until the subject requested to stop. Aftereach stimulus the subject rated the felt affective and sensorypain. The first level at which the subject indicated that thestimulus felt painful constituted the pain threshold; the laststimulus tolerated was the pain tolerance level.After the ascending series, the subject completed the STAIstate, the CSQ, and the BIDR. Meanwhile, the experimentercalculated the pain sensitivity range (PSR, Wolff, 1978)-—therange from threshold to tolerance——and three levels spacedequally between threshold and tolerance. The current intensitiescorresponding to these five levels were given in a randomizedorder in the next three series.The pretest series involved the administration of the fiveshock stimuli. Ratings on the visual analogue scales wereobtained as before.The manipulation phase repeated the pretest procedure. Themanipulation in Study 1 comprised asking the subject to volunteerto fool viewers into believing that she had substantially lesspain, or substantially more pain by altering her facialexpression (Study 1; see Appendix D for instructions read tosubjects). The manipulation in Study 2 comprised making a false,prepared statement following the series (Study 2; see AppendicesE & F for protocol instructions and a listing of the statements,including the negative consequence add—on). That is, subjectschose a type—written statement from a box before the manipulationseries shocks, they read the statement out loud as a rehearsal,then sat through the shock trials. They stated their prepared45statement on cue when the experimenter returned to the subjectroom and asked, “How was the pain that time?” Control subjectswere not asked to alter their pain display nor to make a falsestatement about the pain. Study 2 subjects completed aquestionnaire after the manipulation phase which measured whetherthey felt free to choose their experimental condition, theirdegree of felt discomfort at lying about their pain, and theirfelt responsibility for any negative consequences that mightoccur to viewers (see Appendix J for this questionnaire).A final posttest series proceeded in the same manner as thepretest, to establish whether the subjects’ pain ratings andfacial activity changed as a result of the earlier manipulation.Subjects were told that this was just a repeat of the normalseries of shocks. Then subjects completed a post—experimentalquestionnaire which elicited their hypotheses about the nature ofthe study as well as any perceived changes in pain experienceover time, her sense of responsibility for negative consequences,and her discomfort due to misleading others. A full debriefingwas carried out and the subject was offered the opportunity todiscuss any queries or concerns that she had regarding thedeception.Facial Action CodingFacial expressions were scored by a data coder who hadpassed proficiency tests as a FACS coder (Ekinan & Friesen,1978a). She was blind to the experimental hypotheses and tocurrent intensities delivered. In accordance with standardscoring procedures, a 3—second videotape segment capturing a46neutral expression of the subject served as a comparison againstwhich other segments were scored using FACS (Ekman & Friesen,1978b). Facial movements during the threshold, medium, andtolerance level shocks were scored. The interval scored includedthe 0.5-second period preceding the shock, the 0.5—second shockitself, and the 2—second period following the shock. Thisinterval captures the apex of the facial movements and ends priorto the onset of most extraneous movements (in accordance withprocedures followed by Patrick, Craig & Prkachin, 1986). Thus,facial expression data that were analyzed consisted of thethreshold, medium, and tolerance level facial action units scoredin the pretest, manipulation phase, and posttest (9 segments persubject X a subset of 43 out of the 96 subjects in Study 1, X 60subjects in Study 2).47Data Analysis: OverviewThe mean current level at tolerance was 8.5 mA (±4.0), alevel consistent with that found in previous studies usingcomparable shock delivery systems (Tursky & O’Connell, 1972,reported a tolerance level of 9.0 mA in males; Swaim & Craig,1991, reported an average tolerance level of 8.8 mA in males andfemales).Preliminary analyses of the Study 1 results weredisappointing. Time and Intensity showed up as significanteffects, but the between—group experimental variables of interestdid not appear significant. One obvious reason forwarded by theauthor was that many of the Study 1 subjects appeared not to takethe manipulation seriously. For example, five of the Exaggeratesubjects were noticeably laughing during the manipulation phase.Before attempting to repeat the study using more cooperativesubjects, it was decided to analyze the data for those subjectswho did appear to comply with the instructions. The fiveExaggerate group women who were obviously laughing were removedfrom the sample. In addition, four other subjects were removedon the basis of their post—experimental questionnaire responses:One Diminish and three Exaggerate group subjects reported thatthe manipulation caused them to feel more or less pain,respectively, which was counter to the desired manipulationeffect. Finally, two research assistants and the experimenterpreviewed the videotapes to select subjects who did appear tocomply with the manipulation to alter their facial expression.Each of the three judges independently attempted to guess whether48each subject belonged to the Exaggerate, Diminish, or Controlcondition by viewing the pretest and manipulation phases of thevideotape. Retained for further analyses were those subjects whowere correctly classified into the appropriate experimentalcondition by two out of three judges. Since this left more thantwice as many control subjects as experimental subjects, controlgroup subjects were retained only if correctly classified by allthree judges. This left 43 subjects (20 control subjects, 13exaggerate, and 10 diminish subjects). All 60 subjects fromStudy 2 were retained for analyses.A trained FACS coder identified the facial action units(AU5) displayed by the subjects during the low, medium and highintensity shocks of the pretest, manipulation and posttestseries. To maximize power by reducing the number of dependentvariables, AUs previously associated with pain were weighted andsummed according to the procedure described below. The resultantfacial activity scores were subsequently used in the repeatedmeasures ANOVA5 to be described in the Results sections.The process of deriving a weighted sum for the facialactivity scores will be elaborated next. The most frequentlyoccurring AUs, which were found to be pain—related in previousstudies, comprised AUs 1, 2, 4, 6, 7, 12, 14, 25, and 45. Theseoccurred at least an average of .019 times across all subjectsacross all conditions in Study 2 (their descriptions andweightings are in Appendix B). Optimum weights were derived in aprevious study with a similar population undergoing shock-inducedpain trials (Swalin & Craig, 1991). That study employed principal49components factor analysis to derive factor score coefficientssubsequently used to weight the individual AUs which were thensummed. The same weights were used in this study with therationale that cross-validating the weighting scheme would beuseful in extending and replicating the earlier results.*AUs 1, 2 and 14 received negative weights. The rest werepositive. In addition, all of the AU5 except AU5 25, 26 and 27were coded for five levels of intensity. Those AUs were weightedby 0.4, 0.8, 1.2, 1.6, or 2.0 for the five intensity levels Athrough E, respectively. AUs 25, 26, and 27 represent differentlevels of lips parting and were recoded as different intensitiesof AU 25 as per Prkachin & Mercer (1989) (assigned intensitylevels were weighted by 0.4, 0.8, and 1.2).To summarize, pain—related AUs were weighted and then summedto derive a facial activity score for each experimental phase foreach subject. The weights reflected how greatly the individualAU5 were associated with a pain factor (in a previous factoranalysis) as well as how intense each AU appeared to the coder.The facial activity scores and rating scale scores (bothsensory and unpleasantness) were separately analyzed in mixeddesign ANOVA5.*The alternative approach is to derive a weighting scheme thatmaximally discriminates among groups and then to use thoseweights for the dependent measure on the same sample. This wouldhave built in an obvious bias towards finding significant resultssince the sample from which the weights were derived is the samesample on which the dependent variable is measured.50Study 1 Results (N=43)The standard alpha level ( < .05) was divided into two forthe two self—report measures (Sensory and Unpleasantness Scaleratings) to reduce the experiment—wise error rate for theserelated measures. Facial Activity was considered separately,therefore, the alpha level for that measure was equal to 0.05.Each of these three dependent measures was analyzed by means offour—way repeated measures analysis of variance. The between—group factors were Condition (Exaggerate, Diminish, or Control),and Consequences (Negative, None). The repeated measures wereTime (Pretest, Manipulation, Posttest) and Shock Intensity (5levels for the rating scales; low, medium, and high for facialactivity). Simple interactive effects analyses were conductedwhere appropriate; these results are tabled in Appendix L. Therewere no significant group differences for current tolerated asmeasured by an ANOVA (p > .10).Both self-report measures yielded highly significant maineffects for Time and Intensity. The Sensory Scale Time effectproduced an (2,74) = 21.59, p < .0001, and the Intensity effectproduced an (4,148) = 57.83, p < .0001. Newman-Keuls tests(cz=.0l) suggested that ratings decreased significantly over time,with all levels of time significantly different from each other(cell means were: Pretest = 63.28, Manipulation = 56.67,Posttest = 38.28). The intensity variable showed an increase inratings with increased shock intensity. For adjacent intensitylevels, all but the fourth and fifth levels were significantlydifferent as examined by Newman—Keuls tests (a=.O1; cell means51were I = 38.44, 12 = 47.55, 13 = 63.35, 14 = 71.73, and 15 =74.84)The Unpleasantness Scale Time effect gave an (2,74) =12.13, p < .0001, and the Intensity effect gave an (4,148) =54.96, p < .0001. Newman-Keuls tests (a=.01) suggested thatratings decreased significantly over time, with all levels oftime significantly different from each other (cell means were:Pretest = 58.68, Manipulation = 52.65, Posttest = 48.49). Theintensity variable showed an increase in ratings with increasedshock intensity. For adjacent intensity levels, all but thefourth and fifth levels were significantly different as examinedby Newman-Keuls tests (a=.01; cell means were I = 27.1, 12 =42,51, 13 = 55.39, 14 = 68.03, and 15 = 73.34).The Unpleasantness Scale also yielded a marginallysignificant interaction involving Condition by Time, (4,74) =2.53, p < .05. This may be seen graphically in Figure 3. Simpleinteractive effects analyses suggested that Condition was not asignificant effect with any level of Time held constant.Instead, Time was a significant effect within the Exaggeratecondition ( (2,74) 14.11, p < .001) with pretest ratingssignificantly higher than the manipulation and posttest ratingsas analyzed by Newman-Keuls tests. Neither of the otherconditions yielded significant Time effects: the Diminishcondition (2,74) = 1.42 (p > .20) and the Control condition(2,74) = 1.43 (p > .20).52Figure 3Condition X Time Interaction forthe Unpleasantness Scale (Study 1)CoCoCoCoCoCoC)C— -..60504030Diminish.-—..-.-.. ControlExaggeratePretest Manipulation PosttestTime53The Facial Activity scores yielded significant main effectsfor Time and Intensity, as well as an interaction effectinvolving these two factors. The Time effect produced an(2,74) = 4.74, p < .015. Newman-Keuls tests suggested that thiswas due to a significant decrease in facial activity frommanipulation to posttest (a=.05; cell means were Pretest = .264,Manipulation = .340, and Posttest = .198).The Intensity effect gave an F (2,74) = 6.80, p < .005.Newman-Keuls tests indicated that facial activity during thelowest intensity level was significantly less than at the highestlevel (a=.0l; cell means were Low intensity = .212, Mediumintensity = .275 and High intensity = . 314).The interaction between Time and Intensity produced an(4,178) = 3.68, p < .01. The graph of this result (depicted inFigure 4) illustrates that from the pretest to the manipulationphase, facial activity at the low and medium intensitiesincreased but stayed the same for the highest intensity level.Simple interactive effects analyses suggested that at the pretesttime only there was a significant intensity effect. Newman-Keulstests revealed that there were significant differences betweenall levels of intensity at the pretest time except for thecontrast between medium and high (cell means were: low=.l45,medium=.287, high=.360). When intensity was held constant, thesimple effects analyses revealed that at low intensity only therewas a significant time effect. Newman—Keuls tests showed thatthe pretest trials were not different from the posttest (at lowintensity), but all other comparisons were significantly54CUC.) -(UL1Figure 4Facial Activity for Three IntensityLevels Across Time (Study 1)f AC’,a,-qSHigh S4%Medium4%S4%4.4%SS4.4%0.3-0.20.1LowI IPretest Manipulation PosttestTime55different (cell means were: pretest=.l45, manipulation=.328,posttest=.l63).In contrast to the self—report measures, the Condition byTime interaction for facial activity was not significant (p >.44) although the Exaggerate subjects displayed more facialactivity during the manipulation phase as can be seen on Figure5. The cell means for the complete analysis ranged from .0612 to.621 while the standard deviations ranged from .0384 to .9266,which clearly indicates a severe problem with individualdifferences and hence poor power.A discriminant function analysis provided more detailedanalysis of the differences between the groups during themanipulation phase. The most frequently occurring AUS (1, 2, 4,6, 7, 12, 14, 20, 25, 43, and 45) were entered as variablespredicting group membership (Exaggerate, Diminish, and Control).The first canonical discriminant function accounted for 95% ofthe between—group variance and was very useful in discriminatingthe groups (Chi2 = 31.45, p < .026). The standardized canonicaldiscriminant function coefficients for the first function(presented in the table below) illustrate the relativecontribution of each variable to group discrimination.56>C)CuC)CuFigure 5Condition X Time Interaction forFacial Activity (Study 1)Exaggerate0.50.4 -0.3 -0.2 -0.1 - I IPretest Manipulation PosttestTime57Table 2. Standardized Canonical Discriminant FunctionCoefficients for the First FunctionAU1 .500AU 4 .694AU 6 .104AU 7 .142AU 14 .227AU 20 .404AU 25 .028AU43—.251AU 45 .445The table of classification function coefficients belowshows how much the AUs are related to each of the conditions. Ingeneral, most of the AU5 were weighted more heavily for theExaggerate condition and least heavily for the Diminishcondition, with the Control group falling in between. Theexceptions to this pattern were AUs 7 and 25.Table 3. Classification Function CoefficientsAU Exaggerate Diminish Control1 3.166 0.445 0.9064 2.684 —0.081 0.3356 0.626 0.110 0.3767 0.424 0.346 0.10914 2.981 0.345 1.39320 1.683 0.108 0.20225 0.312 0.045 0.46043 —0.815 0.706 —0.04445 0.896 0.418 0.559The overall rate of correctly classified cases based on thediscriminant function was 67%. Successful classification was58best in the two experimental groups (at 80% or better) and worstin the Control group (at 50%) as can be seen from the tablebelow.Table 4. Classification ResultsActual N Predicted Group MembershipGroupExaggerate Diminish ControlExaggerate 13 11 (84.6%) 1 (7.7%) 1 (7.7%)Diminish 10 0 (0%) 8 (80.0%) 2 (20.0%)Control 20 2 (10.0%) 8 (40.0%) 10 (50.0%)Ratings of the subjects’ sense of responsibility for anynegative consequences that might occur to trainees who view theirvideotape were examined for the relationship to the consequencescondition. The subjects in the negative consequences condition(N=48) rated their felt responsibility slightly higher thansubjects in the no consequences condition (N=48): the means andstandard deviations for the two groups were 2.58 (±1.53) and 2.07(±1.17), respectively. This resulted in a significant differencein the predicted direction as measured by the test (t = 2.15,df = 94, p < .05), but the means do not appear meaningfullydifferent.In summary, subjects in the Exaggerate condition tended torate their pain as less during the manipulation and posttesttrials than during the pretest trials. This effect almostreached significance when subjects who did not comply with the59manipulation were removed from the analyses. Nor did otheranalyses involving the experimental condition reach strictsignificance levels. However, on the Unpleasantness Scale therewas evidence that within the Exaggerate condition, pretestratings were higher than manipulation and posttest ratings.The repeated measures of Time and Intensity reachedsignificance across all dependent measures, suggesting that thesubjects reliably displayed more pain with higher intensityshocks but their pain expression tended to decrease over time.Study 2 Results (N=60)As with Study 1, the standard alpha level ( < .05) wasdivided into two for the two self-report measures (Sensory andUnpleasantness Scale ratings) to reduce the experiment—wise errorrate for these related measures. Facial Activity was consideredseparately with an alpha level equal to 0.05. There were nosignificant group differences for current tolerated as measuredby an ANOVA (p > .10).The Sensory Scale ANOVAR yielded two significant maineffects and one marginally significant interaction effect. Thisis graphically presented in Figure 6 for the two lowest intensitylevels. (Appendix R lists the cell means for all effects.)Time was the first main effect, (2,108) = 14.64, p <.0001. Newman-Keuls analyses suggested that the pretest ratingof Sensory pain was significantly higher than either themanipulation phase or the posttest phase ratings (p < .01), butthe manipulation and posttest trials were not significantlydifferent over time.Figure6C(50E0C>.10Cs,C0U)ConditionXTimeXIntensityInteractionfortheSensoryScale,LowestTwoIntensityLevels(Study2)DIntensity1Intensity2pretestmanipulationposttestpretestmanipulationposttestpretestmanipulationposttestExaggerateDlmlhlshControl61The second main effect was Intensity, F (4,216) = 101.48, p< .0001. Newman-Keuls tests indicated that all pair-wisecontrasts were significantly different (at p < .01) exceptbetween the fourth and fifth highest intensity levels.The Condition by Time by Intensity interaction almost metthe stepped-down alpha level of significance, with an (16,432)= 1.82, p < .027. Simple simple interactive effects analysessuggested that during the manipulation phase, for the two lowestintensity levels there was a highly significant condition effect.At the lowest intensity level, the Diminish condition subjectsrated their pain significantly less than subjects in the Controlcondition (p < .01). At the second lowest intensity level, theDiminish subjects rated their pain as significantly less thaneither the Control or the Exaggerate subjects (p < .01). (Simpleeffects analyses are tabled in Appendix L.)The Unpleasantness Scale results were very similar to thosereported for the Sensory Scale, but the three—way interactioneffect was clearly significant for this self—report measure fordifferent reasons than reported for the Sensory Scale. Theseresults are depicted in Figure 7 with cell means listed inAppendix S.The main effect for Time [ (2,108) = 23.34, p < .0001] wasaccounted for by a significant drop in ratings from pretest tomanipulation phase (Newman-Keuls tests, p < .01), which levelledout towards the posttest phase.Figure7Ce0004-40DConditionXTimeXIntensityInteractionfortheUnpleasantnessScale(Study2)intensity1Intensity2intensity3Intensity4C]Intensity5pretestmanipulationposttestpretestmanipulationposttestpretestmanipulationposttestExaggerateDiminishControl63The main effect for Intensity [ (4,216) = 102.29, p <.0001] was accounted for by significant differences between alllevels of intensity (Newman—Keuls tests, p < .01) except betweenthe highest and second-highest intensities.The Condition by Time by Intensity interaction met thestepped—down alpha level of significance, with an (16,432) =2.35, p < .005. (Simple effects analyses are tabled in AppendixM.) Simple simple interactive effects analyses suggested thatduring the manipulation phase, for the second lowest intensitylevel there was a marginally significant condition effect. TheDiminish subjects rated their pain as less than either theControl or the Exaggerate subjects, but this failed to reach thestepped—down alpha level of significance (p > .01). Probably thesignificant three-way interaction effect was due to highlysignificant effects within the Control condition for Intensity byTime. Simple simple interactive effects were found to be highlysignificant for the top three intensity levels (p < .01).Newman—Keuls tests for Time within each of these Intensity levelsonly suggested significant drops in ratings from pretest tomanipulation.Facial Activity results were similar to the self-reportresults in that there were the two significant main effects forTime and Intensity, but there were no significant interactions.The main effect for Time [ (2,108) = 14.64, p < .001] wasfollowed up by Newman-Keuls tests which again suggested adecrement in response after the pretest, but only the contrastbetween pretest and posttest reached significance (p < .01).64The main effect for Intensity [ (2,108) = 14.09, p < .001]was accounted for by significant differences (p < .01) betweenthe low and medium and between the low and high levels, but notbetween the medium and high levels.To determine whether the facial activity in this study wassimilar in quality to that of the previous study from which theweighting scheme was derived (Swalm & Craig, 1991), a PrincipleComponents Analysis was conducted on the pain—related facialaction units for Study 2 subjects. In contrast to previousresults, this time a two-factor solution rather than a three—factor solution provided the most comparable results. This two-factor solution accounted for 47% of the variance. The firstfactor seemed more of an expression of surprise with eyebrowsraised, while the second factor appeared more like an expressionof pain with furrowed brows, narrowed eyes, and grimacing mouth.The previous solution was the reverse in that the first factorwas called pain—related and the second was surprise. AU 23 (liptight) did not occur commonly enough in the present sample to beretained in the analysis. The factor score coefficientspresented in the table below reveal the relationship between theindividual AUs and the respective factors for both the previousstudy and the present one. The two columns correlate moderatelyfor pain-related AU5 (r=.78) and highly for surprise AUs (r=.96).65Table 5Pain—related AU5 and respective factor score coefficientsAU and descriptor 1987 Study 1991 Study1 inner brow raise -.065 -.0342 outer brow raise -.075 —.0384 brow lowerer .337 .1346 cheek raise .464 .43712 lip corner pull .332 .45014 dimpler —.029 —.02823 lip tight -.041 N/A25 lips part .206 .35545 blink .188 .027Surprise AUs and respective factor score coefficientsAU and descriptor 1987 Study 1991 StudyI inner brow raise .463 .5052 outer brow raise .464 .5044 brow lowerer .029 .0106 cheek raise —.075 .04412 lip corner pull —.043 .03914 dimpler .016 .00323 lip tight -.080 N/A25 lips part —.010 .04845 blink .206 .052The relationships between the dependent measures. Todetermine which of the AU5 were most related to the self—reportmeasures, the pain—related AU5 were entered into two stepwisemultiple regression equations with the self—report measures asthe criterion values. Inclusion and exclusion criteria were setloosely at p < .300. AU5 which predicted the Sensory report were4 (brow lowerer) and 45 (eye blink), accounting for 14% of thevariance in the Sensory report, as can be seen in the tablebelow. The beta weights reflect the magnitude of importance of66the relationship between the variables (each AU) and thecriterion (Sensory Scale rating).Table 6. Multiple Regression Results for the SensoryScale RatingsVariable Multiple R R2 R2 Change Beta (Eqn)1. AU 4 .35 .12 .12 .35 8.07**2. AU 45 .37 .14 .02 —.13 4.63*** p < .01.* p < .02.AU5 which predicted the Unpleasantness Scale ratings were AU5 4(brow lOwerer), 1 (inner brow raise), 2 (outer brow raise) and 6(cheek raiser), accounting for 27% of the variance (see tablebelow). AUs 4 and 2 received the highest beta weights whichsuggests that their relationship to the Unpleasantness Scale wasgreatest.Table 7. Multiple Regression Results for the UnpleasantnessScaleVariable Multiple R R2 R2 Change Beta (Eqn)*1. AU 4 .43 .19 .19 .43 13.412. AU 1 .45 .21 .02 —.14 7.413. AU 2 .50 .25 .05 .82 3.624. AU 6 .52 .27 .02 —.15 1.46*p < .01.67To further examine whether the experimental manipulation wasmost successful with subjects scoring high on the BIDRquestionnaire measure of self—deception, ANOVAR5 were conductedusing high v. low levels of self-reported self—deception. Thecomplete analyses followed this format: 3 (experimentalconditions) X 2 (high v. low questionnaire measure of self—deception) X 3 (time) X 5 (shock intensity), comprising abetween-within design as previously outlined. The BIDR self-deception scores were divided at the median to provide the secondgrouping factor. The consequences condition was dropped as agrouping factor to simplify subsequent analyses (it had not beensignificant in prior analyses). Different extremes of self-deception scores were used to form high v. low groups, but norelationship was uncovered with self—deception across theexperimental conditions. (The results are listed in Appendix N.)To further examine the question of whether self—deceptionmoderated the experimental effects, stepwise multiple regressionanalyses were conducted using the subset of data that had shownsignificant effects reported earlier. This was, namely, onsensory ratings for the two lowest shock intensity levels, theANOVARs suggested that Diminish subjects rated their pain as lessthan Control &/or Experimental subjects. Thus, for theregression analyses, the predictor variables comprised group,self-deception, and group by self-deception while the criterionvariables were sensory ratings at pretest for the first analysisand at manipulation for the second analysis. There was a weakrelationship (.05 < p < .30) between the variables such that the68Diminish subjects who scored high on self-deception tended torate the shocks as less intense than either Exaggerate or Controlsubjects at both times examined. (The results of these analysesare tabled in Appendix T.)Similar median—split analyses were conducted using anxietyas a grouping factor. Anxiety ranged from 21.5 to 67 with themedian at 39 (about the 71st percentile). Anxiety was notrelated to the self-deception manipulation, but it did interactsomewhat with shock intensity (p < .05). Highly anxious subjectsrated all but the highest intensity level as less painful thanless anxious subjects. Anxiety was negatively correlated withshock level tolerated (r=-.20, p=.125). (Results are tabled inAppendix 0.)Analyses were conducted on the Coping StrategiesQuestionnaire items——Catastrophizing and Sense of Control——entered as grouping variables. Again, no relationship wasilluminated between these variables and the experimentalconditions. (See Appendices P & Q for tabled results.)To better understand what correlates with self—reportedself—deceptive behavior, several measures were entered into amultiple regression analysis using the Balanced Inventory ofDesirable Responding measure of self—deception as the criterionvariable. Predictor variables were the subjects’ subscale scoresfrom the Coping Strategies Questionnaire*, self—reported anxiety,*Factor scores have been derived for this scale using chronicpain populations. But because the present study used the CSQ fora novel population, the entire list of subscales was used ratherthan factor scores. Moreover, the 3—factor solution previouslyfound is not much simpler than the 8-subscale format.69and post—experimental questionnaire items. Inclusion andexclusion criteria were set to p < .10. Two predictor variablesentered the regression equation to account for 26% of thevariance in self-deception. Catastrophizing was negativelypredictive of self—deception while a sense of control over painwas positively related to self-deception. See the summary tableat the bottom of the table below for details.70Table 8.Correlations Between Questionnaire MeasuresCC1 CC2 CC3 CC4 CC5 CC6 BC El E2 SDCC1CC2 36CC3 02 —10CC4 38 47 —22CC5 24 17 31 16CC6 38 23 —26 67 26BC 39 20 —05 32 28 34El 25 24 —39 47 —12 57 19E2 22 25 —29 38 —08 40 34 61SD 01 —02 —47 08 —16 21 01 37 24From the Coping Strategies Questionnaire, are the cognitivecoping strategies as follows: CC1 = diverting attention, CC2 =reinterpreting the pain sensations,CC3 = catastrophizing, CC4 =ignoring sensations, CC5 = praying or hoping, and CC6 = copingself-statements. Also from the CSQ are the behavioral copingstrategy of increased behavioral activities = BC and theeffectiveness ratings of a sense of control over pain = El andability to decrease pain = E2.SD = self—deception as measured on the Balanced Inventory ofDesirable Responding.Multiple Regression Results for BIDR Self-DeceptionVariable Multiple R R2 R2 Change Beta (Eqn)*1. Catast .47 .22 .22 —.47 15.902. Control .51 .26 .04 .22 9.79* both values were significant, p < .001.In summary, Study 2 demonstrated a self—deception effect inthe Diminish condition at lower intensities suggesting that thesesubjects rated their pain as less intense during the manipulationphase. Again, the repeated measures of Time and Intensity71suggested that higher levels of shocks received higher painratings and greater facial expression, while pain expressiondiminished over time. Moreover, pain ratings were related tofacial expression, especially ratings on the Unpleasantnessscale. Finally, self—report measures of coping and self—deception suggested that self-deception is highly related to theabsence of catastrophizing and presence of a sense of controlover pain. But these self—report measures did not interact withthe experimental manipulation.72DiscussionThe model of self-deception presented in the literaturereview suggested that telling an untruth about one’s paincondition should arouse and motivate subjects to believe the liethrough a process of self-deception. In keeping with this view,subjects of the two studies were asked to deceive others aboutthe pain that they experienced. It was predicted that this wouldmotivate them to perceive the pain in the direction of thedeception in order to preserve their self—esteem. Forcedcompliance literature suggests that deceiving others leads toattitude change; this manipulation was extended to the presentstudy examining change in pain perception. Feeling responsiblefor negative consequences occurring as a result of the deceptionwas expected to amplify the forced compliance manipulationeffect.Discussion of Study 1 ResultsIn the first study, the experimental manipulation was tohave female subjects exaggerate or diminish their facialexpression, purportedly to test the skills of data coders intraining. The negative consequence condition was to suggest thattrainees who failed to reliably distinguish alterations in paindisplay would not be hired.As mentioned before, a trained FACS coder coded the facialactivity for Study 1 subjects who were judged to comply with theexperimental manipulation. Subjects who were judged as showingmore or less pain in their faces were reliably distinguished fromone another on the discriminant function analysis with Exaggerate73subjects generally showing more of the pain-related AUs thanDiminish subjects, and Control subjects falling somewhere inbetween. The Control group subjects were difficult to separatefrom the Diminish group subjects. Similar findings have beenreported with chronic back pain patients whose exaggeratedexpressions appeared like an exaggeration of the genuine painexpression (Craig, Hyde & Patrick, 1991). But these researchersfound some residual evidence of pain in the diminished (“masked”)expressions which was not evident in the present study.Moreover, Poole and Craig (1991) found that judges could reliablydiscriminate between pain subjects told to diminish their painexpression versus subjects not told to diminish their expression.On the other hand, Kopel & Arkowitz (1974) reported thatobservers had difficulty separating subjects’ “calm” poses fromcontrol group poses during painful shock.One possible reason for not observing residual facial signsof pain in the present study may be that the laboratory subjectstolerated much less pain and felt less pain than the subjects inthe Craig et al. (1991) study who were chronic back pain patientsvideotaped during painful physiotherapy exercises. Moreover, incontrast to back pain patients, the laboratory subjects possiblyperceived a greater sense of control over the pain stimuli sinceeach had determined the maximum stimulus intensity to betolerated through-out the study. Such a sense of control mighthave dampened the overall facial expressiveness as well as thefelt pain. Moreover, the stakes for getting caught lying were74minimal, possibly minimizing the likelihood of leakage orevidence of deceit (Ekman, 1985).Counter to the hypotheses, subjects in the Exaggeratecondition tended to rate their pain as less rather than moreduring the exaggerate and posttest trials than during the pretesttrials. This effect almost reached significance on the ANOVAeven when subjects who did not comply with the manipulation wereremoved from the analyses. R. Kieck (personal communication)suggested that this failure to replicate the earlier studies(Colby et al., 1977; Kieck et al., 1976; Lanzetta et al., 1976)was not unusual. One reason might be that cognitively complexinstructions such as those used in Study 1 served to distractsubjects from the pain, thereby resulting in decreased painreport. The author’s observations in support of the distractionhypothesis were that subjects in the Exaggerate conditiondemonstrated discomfort and self—conscious laughter, suggestingthat they were keenly aware of their faces and less aware of thepainful stimuli. Many seemed unable to take the manipulationseriously. But this alone did not account for the distraction assuggested by the failure to demonstrate increased pain report inExaggerate subjects who did appear to comply with themanipulation. Kleck also reported that their studies involvingmales may not generalize to female volunteers. In addition, henoted that expressive behavior typically plays a relatively smallrole in the variance of pain ratings. For example, using meta—analyis to examine the conclusions of facial feedback studies,75Matsumoto (1987) reported that expressive behavior plays a smallrole in the variance of self—reported mood (12%).As a manipulation check, a difference in facial expressionwas readily apparent on the discriminant function analysis butdid not result in a group difference on the ANOVAR. One reasonis as follows: the ANOVAR facial activity score was a weightedsum of the AU5 using weights derived in a separate sample ofsubjects (Swalm & Craig, 1991). In other words, reliability wasreduced by the attempt to cross-validate the weighting scheme.The negative consequences manipulation was not involved inany significant effects, nor did subjects exposed to the negativerate their responsibility for bringing about negativeconsequences much higher than no—consequences subjects. Thus,the author concluded that the nonverbal manipulation and thenegative consequences manipulation were ineffective in alteringpain. In short, Study 1 did not find any evidence in support ofself-deception or of the Facial Feedback Hypothesis. Furthereffects of self-deception were sought in the follow-up study,Study 2.Compared to Study 1, Study 2 followed more directly fromforced compliance literature in that subjects verbally declaredan attitude about their pain experience that was discrepant withthe truth. Facial expression became a strictly dependent measurerather than the independent variable. Recall that themanipulation was to have subjects state that their pain wasworse, less, or the same as what they had expected. The negativeconsequence was to tell subjects that participants of a later76study would watch the videotape before undergoing a similarseries of shocks; such viewing was said to cause worse pain forviewers. The instructions were simplified to prevent what Kieck(personal communication) called “cognitive overload” anddistraction.Discussion of Study 2 ResultsAs in Study 1, the repeated measures of Time and Intensitywere reflected by all of the dependent measures. Therelationship between the dependent measures was examined andrevealed that AU 4 (brow lowerer) was consistently related toboth the Sensory and Unpleasantness Scales, accounting for mostof the variance in the stepwise multiple regression equations(12% and 19% of the variance, respectively). Thus, if limited toobserving one facial movement only, researchers and practitionersmight be advised to observe pain patients’ foreheads for evidenceof pain.The overall factorial structure of facial activity duringthe shock-induced pain stimuli resembled that of the earlierstudy (Swaim & Craig, 1991). Together these studies suggest thatthe stimuli induce a combination of pain and surprisecharacterized mainly by lowering or raisin characterized mainrespectively. The painful facial expression also typically has asquinting movement called the cheek raise which emphasizes so—called crows feet.If forced to choose one measure of adaptive coping, theresults of Study 2 suggest that self-deóeption (as measured byPaulhus’s Balanced Inventory of Desirable Responding, 1984) is77closely related to the absence of catastrophizing and thepresence of a sense of control over pain (as measured by theCoping Strategies Questionnaire, Rosenstiel & Keefe, 1983), twofacets of coping which are thought to be useful in coping withchronic pain (Keefe et al., 1987a & 1987b; Reesor & Craig, 1988;Spinhoven & Linssen, 1991; Turner, 1990). These self—reportmeasures did not interact with the experimental manipulation.Thus, we are still uncertain about what variables mightpredispose an individual towards becoming self—deceived.It was surprising to discover that subjects scoring high ona measure of anxiety tended to rate the shocks as less painfulthan subjects scoring low on anxiety. This relationship was-possibly confounded by the finding that highly anxious subjectsopted to receive less painful shocks when determining their paintolerance levels.The results of Study 2 further indicated that low intensitylevels of self-reported pain can be reduced by preparing to tellothers that one’s pain is less. Thus, there was support for themain hypothesis that altering one’s pain display leads to achange in one’s pain experience. This held for the verbaldisplay of pain but not for the nonverbal display, at the time ofthe manipulation but not afterwards. The effect was greatestwith the Sensory Scale, marginal with the Unpleasantness Scale,but not significant with Facial Activity.Since negative consequences did not amplify the manipulationeffect, and because it was limited in time to the manipulationphase only, one can only conclude that it was due to either self—78deception and/or self—perception, cognitive dissonance, orimpression management phenomena. If self—deception were takingplace, the altered pain report should have endured over timesince the nature of self—deception implies that it results in adeeper cognitive shift than self—perception or impressionmanagement. Thus the forced compliance effect was demonstratedin this study, but negative consequences did not amplify theeffect.One caveat is in order: the results and generalizationsmust be limited to females since only female subjectsparticipated in the study. However, pain perception is generallynot all that different between males and females (Davis, 1981;Lander & Fowler—Kerry, 1989) although when other factors such ascultural background and age are considered, gender may mediatealterations in pain report (Koopman & Eisenthal, 1984). Part ofthe difficulty in sorting out gender differences from culturalstereotypes might be due to the observation that female painpatients are treated differently than males by their physicians(Lack, 1982). For example, female pain patients may be older andmay have suffered longer before a referral is made to a painunit; they may be more likely to receive minor tranquilizers andantidepressants than narcotic analgesics (Lack, 1982).79Future directionsPart of the difficulty in demonstrating that other-deceptivebehavior leads to self-deceptive behavior might be that subjectsresisted the instruction to simulate greater or lesser pain.Previous studies (Goebel, 1983; Heaton et al., 1978) havereported that a surprising number of student subjects (10-20%)reported during the posttest that they had been unwilling orunable to fake malingering as requested by the experimenter.This is a problem recently noted in malingering research(Bernard, 1990). It may have been a greater problem in thepresent study because the design necessitated making subjectsuncomfortable with their deceptive behavior. Thisdiscomfort may have been circumvented by subjects who appeared togo along with instructions just enough to satisfy theexperimenter’s requests but not enough to really fool themselves.In other words, perhaps the experimenter was deceived, but mostof the subjects were not. In light of this, future studiesinvolving subjects deceiving others must include posttestmeasures of the degree to which they felt able to carry out theinstructions to fake. As mentioned by Paulhus (personalcommunication), forced compliance studies require much effort inwording instructions so as to foster a sense of discomfort insubjects but not so much discomfort that subjects elect towithdraw from the study.The second study demonstrated that pain report can bereduced but not increased by means of the forced compliancemanipulation. Several previous studies only demonstrated80reductions in pain report (Bandler, Madaras & Bern, 1986; Totman,1975; Zimbardo et al., 1966) but some have demonstrated thatsubjects report pain with minimal stimulation if they are led toexpect to feel pain (Bayer et al., 1991). The inability todemonstrate increases in pain report in the present study makesdefinitive theoretical statements elusive. It was reasoned thatif pain can be reduced by self-deception, it should be possibleto increase it, too. Mitigating against attempts to increasepain report might be the artifactual effect that over time, painexpression decreased or habituated according to all dependentmeasures. Thus, attempts to increase pain expression workedagainst this time effect.The finding that sensory pain changed more than affectivepain was unexpected. Usually psychosocial manipulations alterthe affective component of pain more than the sensory pain.However, power analyses suggested that for the Condition by Timeinteraction in Study 2, there was much more power (and hencegreater likelihood of finding significant results) associatedwith the Sensory Scale than for the Unpleasantness Scale (.666and .351, respectively). Poor power associated with FacialActivity might account for the dearth of findings for thatdependent measure (power = .149). Thus, it is impossible to knowwhether the manipulation was not evident across all of thedependent measures because of differences in the reliability orindividual variations of the measures or because of theoreticaldifferences between the measures.81One theoretical reason why the self—deception manipulationshowed up on measures of self—report but not on facial activitymight be that this was a “hard case” of self—deception (inPaulhus’s terms). That is, self-deception might involve aprocess akin to different levels of consciousness in which self—deception is seen on verbal report but not on deeper orinvoluntary measures such as nonverbal expression. Further studyis necessary before drawing this conclusion.A second reason why the sensory pain might have been alteredmore might be related to the similar finding in other laboratorypain studies. For example, laboratory subjects undergoingnoxious heat stimulation rated the sensory dimension of theirhigher than the affective dimension (Gaston-Johansson etal., 1985; Price, Harkins & Baker, 1987). On the other hand,some clinical patients (especially those with a high threat tohealth) report greater affective discomfort than sensory painintensity (Price, Harkins & Baker, 1987). Pain that triggersaffective distress tends to be rated higher on a dimensiontapping unpleasantness than on a dimension tapping sensoryintensity (Wade et al., 1990).No evidence was gathered to support the notion that negativeconsequences amplifies the forced compliance effect. Berkowitzand Devine (1989) questioned the necessity of negativeconsequences in forced compliance manipulations. They might becorrect and/or the negative consequences manipulation in thepresent study might have been weak. Again it is difficult toensure that subjects will voluntarily concede to deceive others82especially when negative consequences will accrue to otherparticipants. This is a difficulty that future researchers willface, too. Having conducted one such study, the author remainsuncertain of how to overcome this in the modern research worldwhich allows students to withdraw from participating. Moreover,few psychology students remain naive about studies that examinetheir compliance with authority as a result of learning about theMilgram (1965) studies.The questionnaire measure of self—deception was related tokey factors found to predict adaptive coping or good functioning——absence of catastrophizing and presence of a sense of controlover pain. Thus, self-deception might be an adaptive way ofcoping with acute pain. This contradicts former conclusions thatself-deception might be a poor strategy in coping with long-termillness (Linden et al. 1986). However, this result must beviewed with caution since the questionnaire measure of self—deception did not interact with the experimental manipulationdesigned to demonstrate self—deceptive behavior.Since optimism has been shown to predict good physicalhealth (Scheier & Carver, 1985, 1987), optimism might also becorrelated to self—deception. Future studies should examine therelationship between these two constructs which might account fora more generally adaptive way of dealing with stress thanspecific coping strategies, especially in light of researchfindings indicating that any kind of coping strategy use issuperior to no use (Devine & Spanos, 1990). If so, then self—83deception or optimism or adaptive coping might represent the sameconstruct.An implication of the present literature review and study isthat present treatments for pain might include self-deceptivephenomena. For example, the demand characteristics of thetreatment situation encourages some obedient patients to statethat their pain is less of a problem as a result of thephysician’s ministrations. While this might be effective inreducing pain while the patient is telling this untruth, thepresent study suggests that the pain might return a short timelater such as when the patient leaves the physician’s office.No evidence was found to support the contention that people-who fake more pain come to convince themselves that they havemore pain. This contradicts the self-deception model propositionthat patients who repeatedly tell their pain story while “doctorshopping” become entrenched in their complaints. However, thehigh personal stakes involved in pain patients’ descriptionsmight make them more susceptible to self-deception thanlaboratory subjects who tell a pain story only once to a mirrorwhich purportedly hides a video camera capturing their statementsallegedly to be seen by some unknown fellow student. Mycolleagues and the author have observed pain patients entering aback pain program at the Workers’ Compensation Board of BritishColumbia exaggerating their complaints to ensure that “you guysknow I’m in serious pain” (a recent quote from one of mypatients). Thus, this area of exaggerated pain displaysdeserves further study. At the very least, pain displays would84focus attention on one’s pain and thereby increase sufferingthrough such a preoccupation with pain. It remains to be seenwhether self—deception is another process that can increase painand suffering.Future research might better simulate the clinical situationby having subjects tell the pain story to someone introduced asmedical specialists. Repeated episodes of describing the pain toa doctor would be expected to harden the self—deception such thatthe subject might resist attempts to dissuade him/her of thepain. Longitudinal research with actual pain patients might bethe best research approach to address these hypotheses sincepatients have real stakes involved in their pain complaints.Moreover, actual pain patients might prove to be more cooperativesubjects. For example, other researchers in the Craiglaboratories (e.g., S. Hyde and H. Hadjistavropoulos) havesucceeded in getting clinical pain patients to exaggerate anddiminish their facial expressions without undue self—consciousbehavior such as laughing.A further implication from the literature review is that ifother—deception leads to self-deception, then questionnairemeasures designed to measure these constructs might cause achange over repeated measurement trials. Over time, subjectswould be expected to answer both other— and self—deceptionquestions increasingly more similarly. Moreover, self—deceptionmay result in great difficulty distinguishing patients who are“genuinely” in pain from those who have consciously attempted todeceive others (Faust & Guilmette, 1990; Pankratz & Erickson,851990) as in the case of ongoing litigation or disabilityinsurance battles. Adversarial pressures can worsen a patient’sfunctioning through dissimulation and/or stress (Weissman, 1990).The self-deception model provides an appealing (but unproven)account of the underlying process involved in dissimulationleading to “real” altered pain.86SummaryChronic pain is a widespread and expensive problem.Moreover, pain is difficult to measure, explain and treat. It ishoped that by furthering our understanding of pain, we can beginto effect better treatments for it.Several models of pain have been forwarded to account forsome cases of chronic pain, but no one explanation can accountfor the underlying processes involved in the genesis of chronicpain in all cases. The present analysis offers the additionalfactor of self—deception, defined as a contradiction betweenone’s words or attitude and behavior which is motivated byprotecting self—esteem. For example, by attempting tosubjective pain to observers, a pain patient mayconvince him- or herself of the displayed painfulness through aprocess of self—deception. The analyses presented in this thesiswere designed to address the validity of this model. That is,does altering one’s pain expression come to change one’sexperience of pain through a process of self—deception?To explore the self—deception model of chronic pain, alaboratory analogue study was devised using female studentvolunteers who rated the painfulness of shock—induced stimuliunder conditions designed to foster self—deception, namely theforced compliance manipulation. Painfulness was measured 1)verbally by means of two visual analogue scales which reflectedthe pain intensity and affective unpleasantness and 2)nonverbally by means of quantified facial muscle movements.After determining the subject’s pain threshold and tolerance87levels, she underwent a pretest comprising five random shocksfrom her threshold to tolerance range. Next, a manipulationphase was to ask the subject to display more, less or the samedegree of pain while undergoing another random series of shocks.A final posttest was identical to the pretest and provided ameasure of the durability of the altered pain display effect.In the first of two studies, the altered pain display wasnonverbal: subjects exaggerated, diminished or did not changetheir facial expressiveness while undergoing the pain stimuli.In the second study, the altered pain display was verbal:subjects were told that at the end of the series they would berequired to tell a fellow student (via videotape) that the shockshurt more, less or about the same as what they had expected.Half of all subjects were further told that their deceptivecommunication would have negative consequences for viewers.Misleading fellow students about the pain experienced wasexpected to make the subjects feel badly, motivating them tochange their attitude or beliefs about the pain experienced.They were expected to change their pain reports in keeping withthe, deceptive communication. That is, other deception wasexpected to foster self—deception. This effect was expected toendure and it was expected to be greatest for those in thenegative consequences condition.The results of the first study showed that exaggeratedfacial expressions of pain appear as an amplification of normalpain expression. However, increases or decreases in facial88expression did not bring about changes in verbal report of painperceived, calling into question the facial feedback hypothesis.The results of the second study suggested that pain wasaltered only for subjects who prepared to state that their painfelt less painful than expected. This effect reachedsignificance on the pain intensity visual analogue scale for lowintensity shocks. This effect did not carry over into theposttest phase, nor were negative consequences effective inamplifying the manipulation, leaving the theoretical mechanismunderlying the change in pain unclear.One particular facial movement (brow lowerer) wasconsistently related to the verbal pain reports, attesting to thevalidity of facial expression as a measure of pain. However,there was generally a dearth of findings for the dependentmeasure of facial activity.Self—reported self—deception was related to factors found topredict good functioning in chronic pain patients, namely a senseof control over pain and the absence of catastrophizing thoughts.Thus, the construct of self—deception might hold promise as anadaptive coping style.One major problem with research involving forced compliance(or more generally, subject deception) is that subjects mayappear to comply with experimental instructions to deceive otherswhile avoiding personal responsibility for their actions througha variety of mechanisms yet to be determined. The author remainsconvinced that one such mechanism is self—deception, but this isdifficult to prove. It is especially difficult to separate self-89deception from other related or similar constructs includingcognitive dissonance, self—perception, impression management,attentional and memory biases, and maybe even optimism——anothernew construct recently linked to health.90ReferencesAdelmann, P.K., & Zajonc, R.B. (1989). Facial efference and theexperience of emotion. Annual Review of Psychology, 40,249—280.Affleck, G., Tennen, H., Pfeiffer, C., & Fifield, J. (1987a).Appraisals of control and predictability in adapting to achronic disease. Journal of Personality and SocialPsychology, 53, 273—279Affleck, G., Tennen, H., Pfeiffer, C., Fifield, J., & Rowe, J.(1987b). Downward comparison and coping with seriousmedical problems. American Journal of Orthopsychiatry, 57,570—578.Ainslie, G. (1986). Beyond microeconomics. Conflict amonginterests in a multiple self as a determinant of value. InJ. Elster (Ed.), The multiple self (pp. 133—175).Cambridge: Cambridge University Press.American Psychiatric Association (1987). Diagnostic andstatistical manual of mental disorders (3rd ed., revised).- Washington, D.C. - American -Psycjüatric Association. -Arena, J.G., Sherman, R.A., Bruno, G.M., & Young, T.R. (1991).Electromyographic recordings of low back pain subjects andnon—pain controls in six different positions: effect ofpain levels. Pain, 45, 23—28.Aronson, E. (1989). Analysis, synthesis, and the treasuring ofthe old. Personality and Social Psychology Bulletin, 15,508—512.Auden, W.H., & Kronenberger, L. (1966). The Viking book ofaphorisms——A personal selection. New York: Viking Press.Audi, R. (1985). Self-deception and rationality. In M.W. Martin(Ed.), Self-deception and self-understanding (pp. 169-194).Lawrence, Kansas: University Press of Kansas.Bach, K. (1980/1981). An analysis of self—deception. Philosophyand Phenomenological Research, 41, 351—370.Bach, K. (1985). More on self-deception: Reply to Heilman.Philosophy and Phenomenological Research, 45, 611—614.Bandler, R., Madaras, G., & Bem, D. (1968). Self—observation asa source of pain perception. Journal of Personality andSocial Psychology, 9, 205—209.91Bandura, A. (1986). Self-efficacy. In A. Bandura (Ed.), SocialFoundations of Thought and Action (pp. 390-453). EnglewoodCliffs, NJ: Prentice-Hall.Bandura, A., Taylor, C.B., Williams, S.L., Mefford, I.N., &Barchas, J.D. (1985). Catacholamine secretion as afunction of perceived coping self—efficacy. Journal ofClinical and Consulting Psychology, 53, 406—414.Bayer, T.L. (1985). Weaving a tangled web: The psychology ofdeception and self deception in psychogenic pain. SocialScience and Medicine, 20, 517-527.Bern, D. (1972). Self-perception theory. In L. Berkowitz (Ed.),Advances in experimental social psychology (Vol. 6, pp. 1-62). New York: Academic Press.Berkowitz, L., & Devine,.P.G. (l989a). •Research traditions,analysis, and synthesis in social psychological theories:The case of dissonance theory. Personality and SocialPsychology Bulletin, 15, 493-507.Berkowitz, L., & Devine, P.C. (1989b). Some comments in reply:Analysis, synthesis, and contemporary social psychology.Personality and Social Psychology Bulletin, 15, 530-532.Bernard, L.C. (1990). Prospects for faking believable memorydeficits on neuropsychological tests and the use ofincentives in simulation research. Journal of Clinical andExperimental Neuropsychology, 12, 715-728.Billings, A.G., & Moos, R.H. (1981). The role of copingresponses and social resources in attenuating the stress oflife events. Journal of Behavioral Medicine, 4, 139-157.Blackwell, B., & Gutmann, N. (1986). The management of chronicillness behaviour. In S. McHugh & T.M. Vallis (Eds.),Illness behavior: A multidisciplinary model (pp. 401-408).New York: Plenum.Blew, A.F., Patterson, D.R., & Quested, K.A. (1989). Frequencyof use and rated effectiveness of cognitive and behaviouralcoping responses to burn pain. Burns, Including ThermalInlury, 15, 20—22.Blumer, D., & Heilbronn, M. (1981). The pain-prone disorder: Aclinical and psychological profile. Psychosomatics, 22,395—402.Bok, S. (1980). The self deceived. Social Science Information,19, 923—935.92Brodwin, P.E., & Kleinman, A. (1987). The social meanings ofchronic pain. In G.D. Burrows, D. Elton, & G.V. Stanley(Eds.), Handbook of chronic pain management (pp. 109—119).Amsterdam: Elsevier Science Publishers.Brown, G.K. (1990). A causal analysis of chronic pain anddepression. Journal of Abnormal Psychology, 99, 121-137.Buck, R. (1980). Nonverbal behavior and the theory of emotion:The facial feedback hypothesis. Journal of Personality andSocial Psychology, 38, 811—824.Chanowitz, B., & Langer, E.J. (1985). In M.W. Martin (Ed.),Self-deception and self-understanding (pp. 117-135).Lawrence, Kansas: University Press of Kansas.Clark, W.C., & Yang, J.C. (1983). Applications of sensorydecision theory to problems in laboratory and clinical pain.In R. Meizack (Ed.), Pain measurement and assessment (pp.15-25. New York: Raven.Cody, M.J., & O’Hair, H.D. (1983). Nonverbal communication anddeception: Differences in deception cues due to gender and- communicator dominance. Communication Monographs, -50, 175—192.Colby, C.Z., Lanzetta, J.T., & Kieck, R.E. (1977). Effects ofthe expression of pain on autonomic and pain toleranceresponses to subject-controlled pain. Psychophysiology, 14,537—540.Cooper, J., & Fazio, R.H. (1984). A new look at dissonancetheory. Advances in Experimental Social Psychology, 17,229—266.Cooper, J., & Fazio, R.H. (1989). Research tradition, analysis,and synthesis: Building a faulty case around misinterpretedtheory. Personality and Social Psychology Bulletin, 15,519—529.Craig, K.D. (1978). Social modeling influences on pain. In R.A.Sternbach (Ed.), The psychology of pain (pp. 73-109). NewYork: Raven.Craig, K.D. (1980). Ontogenetic and cultural influences on theexpression of pain in man. In H.W. Kosterlitz & L.Y.Terenius (Eds.), Pain and society (pp. 37-52). Weinham:Verlag Chemie GmbH.Craig, K.D. (1983). Modeling and social learning factors inchronic pain. In J.J. Bonica, U. Lindblom, & A. Iggo(Eds.), Advances in pain research and therapy (Vol. 5, pp.813-827). New York: Raven.93Craig, K.D. (1986). Social modeling influences: Pain incontext. In R.A. Sternbach (Ed.), The Psychology of Pain(2nd ed., pp. 67-96). New York: Raven.Craig, K.D. (1989). Emotional aspects of pain. In P.D. Wall &R. Melzack (Eds.), Textbook of pain, 2nd ed. (Pp. 220-229).Edinburgh: Churchill Livingstone.Craig, K.D., Hyde, S.A., & Patrick, C.J. (1991). Genuine,suppressed, and faked facial behavior during exacerbation ofchronic low back pain. Pain, 46, 161-171.Craig, K.D., & Patrick, C.J. (1985). Facial expression duringinduced-pain. Journal of Personality and Social Psychology,4., 1089—1091.Craig, K.D., & Prkachin, K.M. (1978). Social modeling influenceson sensory decision theory and psychophysiological indexesof pain. Journal of Personality and Social Psychology. 36,805—815.Craig, K.D., & Prkachin, K.M. (1980). Social influences onpublic and private components of pain. In I.G. Sarason &C.D. Spielberger (Eds.), Stress and Anxiety (Vol. 7, pp. 57-72). Washington: Hemisphere Publishing Corp.Craig, K.D., & Prkachin, K.M. (1983). Nonverbal measures ofpain. In R. Melzack (Ed.), Pain measurement and assessment(pp. 173—179). New York: Raven.Craig, K.D., & Weiss, S.M. (1971). Vicarious influences on painthreshold determinations. Journal of Personality and SocialPsychology, 19, 53-59.Crisson, J.E., & Keefe, F.J. (1988). The relationship of locusof control to pain coping strategies and psychologicaldistress in chronic pain patients. Pain, 35, 147-154.Crook, J., Tunks, E., Rideout, E., & Browne, G. (1986).Epidemiologic comparison of persistent pain sufferers in aspecialty pain clinic and in the community. Archives ofPhysical and Medical Rehabilitation, 67, 451—455.Davidson, D. (1986). Deception and division. In J. Elster(Ed.), The multiple self (pp. 79-92). Cambridge: CambridgeUniversity Press.Davis, M.A. (1981). Sex differences in reporting osteoarthriticsymptoms: A sociomedical approach. Journal of Health andSocial Behavior, 22, 290—310.Demos, R. (1960). Lying to oneself. Journal of Philosophy. 57,588—595.94Devine, D.P., & Spanos, N.P. (1990). Effectiveness of maximallydifferent cognitive strategies and expectancy in attenuationof reported pain. Journal of Personality and SocialPsychology, 58, 672—678.Dobson, K., & Franche, R-L. (1989). A conceptual and empiricalreview of the depressive realism hypothesis. CanadianJournal of Behavioral Sciences, 21, 419—433.Dolce, J.J. (1987). Self-efficacy and disability beliefs inbehavioral treatment of pain. Behavioural Research andTherapy, 25, 289-299.Douglas, W., & Gibbons, K. (1983). Inadequacy of voicerecognition as a demonstration of self—deception. Journalof Personality and Social Psychology, 44, 589—592.Dunkel-Schetter, C., Folkman, S., & Lazarus, R.S. (1987).Correlates of social support receipt. Journal ofPersonality and Social Psychology, 53, 71-80.Eisenberg, L. (1977). Disease and illness: Distinctions betweenprofessional and popular ideas of sickness. Culture,Medicine and Psychiatry, 1, 9--23.Ekman, P. (1985). Telling Lies: Clues to Deceit in theMarketplace, Politics, and Marriage. New York: Norton.Ekman, P., & Friesen, W.V. (1978a). Investigator’s guide to theFacial Action Coding System. Palo Alto, California:Consulting Psychologists Press.Ekman, P., & Friesen, W.V. (1978b). Manual for the Facial ActionCoding System. Palo Alto, California: ConsultingPsychologists Press.Ekman, P., & Friesen, W.V. (1982). Measuring facial movementwith the Facial Action Coding System. In P. Ekman & W.V.Friesen (Eds.), Emotion in the human face (2nd ed., pp. 178-211). Cambridge: Cambridge University Press.Ekman, P., Friesen, W.V., & Ellsworth, P. (1982). What emotioncategories or dimensions can observers judge from facialbehavior? In In P. E]cman & W.V. Friesen (Eds.), Emotion inthe human face (2nd ed., pp. 39-55). Cambridge: CambridgeUniversity Press.Ekman, P., Levenson, R.W., & Friesen, W.V. (1983). Autonomicnervous system activity distinguishes among emotions.Science, 221, 1208—1210.95Escobar, J.I., Burnam, M.A., Karno, M., Forsythe, A., & Golding,J.M. (1987). Somatization in the community. Archives ofGeneral Psychiatry, 44, 713-718.Elster, J. (1986). Introduction. In J. Elster (Ed.),multiple self (pp. 1-34). Cambridge: Cambridge UniversityPress.Engel, G.L. (1977). The need for a new medical model: Achallenge for medicine. Science, 196, 129-136.Faust, D., & Guilmette, T.J. (1990). To say it’s not so doesn’tprove that it isn’t: Research on the detection ofmalingering. Reply to Bigler. Journal of Consulting andClinical Psychology, 58, 248—250.Fazio, R.H. (1986). How do attitudes guide behavior? In R.M.Sorrentino & E.T. Higgins (Eds.), Handbook of Motivation andCognition (pp. 204-243). New York: Guilford.Fazio, R.H. (1990). Multiple processes by which attitudes guidebehavior: The MODE model as an integrative framework.Advances in Experimental Social Psychology, 23, 75—109.Fazio, R.H.,-Zanna,M.P., & Cooper, J (19L77). Dissonance andself—perception: An integrative view of each theory’sproper domain of application. Journal of ExperimentalSocial Psychology, 13, 464—479.Felton, B.J., Revenson, T.A., Hinrichsen, G.A. (1984). Stressand coping in the explanation of psychological adjustmentamong chronically ill adults. Social Science and Medicine,18, 889—898.Fernandez, E. (1986). A classification system of cognitivecoping strategies for pain. Pain, 26, 141-151.Fernandez, E., & Turk, D.C. (1989). The utility of cognitivecoping strategies for altering pain perception: A metaanalysis. Pain, 38, 123—135.Festinger, L. (1957). A theory of cognitive dissonance.Stanford, Calif.: Stanford University Press.Festinger, L., & Carismith, J.M. (1959). Cognitive consequencesof forced compliance. Journal of Abnormal and SocialPsychology, 58, 203—210.Fingarette, H. (1985). Alcoholism and self—deception. In M.W.Martin (Ed.), Self-deception and self-understanding (pp. 52-67). Lawrence, Kansas: University Press of Kansas.96Flor, H., & Turk, D.C. (1984). Etiological theories andtreatments for chronic back pain. I. Somatic models andinterventions. Pain, 19, 105—121.Folkman, S., & Lazarus, R.S. (1980). An analysis of coping in amiddle-aged community sample. Journal of Health and SocialBehavior, 21, 219—239.Fordyce, W.E. (1976). Behavioral methods for chronic pain andillness. St. Louis: C.V. Mosby.Fordyce, W.E. (1989). The cognitive/behavioral perspective onclinical pain. In J.D. Loeser & K.J. Egan (Eds.), Managingthe chronic pain patient (pp. 51-64). New York: Raven.Fordyce, W.E., Roberts, A.H. & Sternbach, R.A. (1985). Thebehavioral management of chronic pain: A response tocritics. Pain, 22, 113—125.Gaston—Johansson, F., Johansson, G., Felldin, R., & Sanne, H.(1985). A comparative study of pain description, emotionaldiscomfort and health perception in patients with chronicpain syndrome and rheumatoid arthritis. ScandinavianJournal of Rehabilitation Medicine, 17, 109-119.Gergen, K.J. (1985). The ethnopsychology of self-deception. InM.W. Martin (Ed.), Self-deception and self-understanding(pp. 228—243). Lawrence, Kansas: University Press ofKansas.Gilbert, D.T., & Cooper, J. (1985). Social psychologicalstrategies of self-deception. In M.W. Martin (Ed.), Self-deception and self-understanding (pp. 75-94). Lawrence,Kansas: University Press of Kansas.Girodo, M., & Wood, D. (1979). Talking yourself out of pain:The importance of believing that you can. Cognitive Therapyand Research, 3, 23-33.Goebel, R.A. (1983). Detection of faking on the Halstead-ReitanNeuropsychological Test Battery. Journal of ClinicalPsychology. 39, 59-69.Gracely, R.H. (1989). Pain psychophysics. In C.R. Chapman &J.D. Loeser (Eds.), Advances in pain research and therapy.Vol. 12. Issues in pain measurement (pp. 211-229). NewYork: Raven.Gracely, R.H., McGrath, P., & Dubner, R. (1978a). Ratio scalesof sensory and affective verbal pain descriptors. Pain, 5,5—18.97Gracely, R.H., McGrath, P., & Dubner, R. (1978b). Validity andsensitivity of ratio scales of sensory and affective verbalpain descriptors: Manipulation of affect by diazepam.Pain, 5, 19—29.Gracely, R.H., McGrath, P., & Dubner, R. (1979). Narcoticanalgesia: Fentanyl reduces the intensity but not theunpleasantness of painful tooth pulp sensations. Science,3, 1261—1263.Greenwald, A.G., Pratkanis, A.R., Leippe, M.R., & Baumgardner,M.H. (1986). Under what conditions does theory obstructresearch progress? Psychological Review, 93, 216—229.Greenwald, A.G., & Ronis, D.L. (1978). Twenty years of cognitivedissonance: Case study of the evolution of a theory.Psychological Review, 85, 53-57.Grunau, R.V.E., & Craig, K.D. (1987). Pain expression inneonates: Facial action and cry. Pain, 28, 395—410.Grunau, R.V.E., Johnston, C.C., & Craig, K.D. (1990). Neonatalfacial and cry responses to invasive and non—invasiveprocedures. Pain,42, 295-305.-Gunther, M.S. (1984-1985). Deceived and betrayed. The Annual ofPsychoanalysis, 12—13, 177—219.Gur, R.C., & Sackeim, H.A. (1979). Self—deception: Aconcept in search of a phenomenon. Journal of Personalityand Social Psychology, 37, 147-169.Gustafson, L.A., & Orne, M.T. (1963). Effects of heightenedmotivation on the detection of deception. Journal ofApplied Psychology, 47, 408-411.Haight, M.R. (1985). Tales from a black box. In M.W. Martin(Ed.), Self-deception and self-understanding (pp. 244-260).Lawrence, Kansas: University Press of Kansas.Haythornthwaite, J.A., Sieber, W.J., & Kerns, R.D. (1991).Depression and the chronic pain experience. Pain, 46, 177—184.Heaton, R.K., Smith, H.H., Lehman, R.A.W., & Vogt, A.T. (1978).Prospects for faking believable deficits onneuropsychological testing. Journal of Clinical andConsulting Psychology, 46, 892-900.Heft, M.W., Gracely, R.H., Dubner, R., & McGrath, P. (1980). Avalidation model for verbal descriptor scaling of humanclinical pain. Pain, 9, 363—373.98Hellman, N. (1983). Bach on self-deception. Philosophy andPhenomenological Research, 44, 113-120.Heyneman, N.E., Fremouw, W.J., Gana, D., Kirkland, F., & Heiden,L. (1990). Individual differences and the effectiveness ofdifferent coping strategies for pain. Cognitive Therapy andResearch, 14, 63-77.Higgins, E.T. (1989). Self-discrepancy theory: What patterns ofself beliefs cause people to suffer? Advances inExperimental Social Psychology, 22, 93-136.Hilgard, E.R. (1949). Human motives and the concept of the self.American Psychologist, 4, 374-382.Hocking, J.E., & Leathers, D.C. (1980). Nonverbal indicators ofdeception: A new theoretical perspective. CommunicationMonographs, 47, 119-131.Hyman, R. (1989). The psychology of deception. Annual Review ofPsychology, 40, 133-154.International Association for the Study of Pain (1986). Painterms: A current list with definitions and notes on usage.Pain, Supp. 3, S215—S221.Izard, C.E. (1990). Facial expressions and the regulation ofemotions. Journal of Personality and Social Psychology, 58,487—498.Jamison, R.N., Rock, D.L., & Parris, W.C.V. (1988). Empiricallyderived Symptom Checklist 90 subgroups of chronic painpatients: A cluster analysis. Journal of BehavioralMedicine, 11, 147—158.Jamner, L.D., & Schwartz, G.E. (1986). Self-deception predictsself—report and endurance of pain. Psychosomatic Medicine,4k., 211—223.Keefe, F.J., Caldwell, D.S., Queen, K., Gil, K.M., Martinez, S.,Crisson, J.E., Ogden, W., & Nunley, J. (1987a).Osteoarthritic knee pain: A behavioral analysis. Pain, 28,309—321.Keefe, F.J., Caldwell, D.S., Queen, K.T., Gil, K.M., Martinez,S., Crisson, J.E., Ogden, W., & Nunley, J. (l987b). Paincoping strategies in osteoarthritis patients. Journal ofClinical and Consulting Psychology, 55, 208-212.Keefe, F.J., Dunsmore, J., & Burnett, R. (in press). Behavioraland cognitive—behavioral approaches to chronic pain: Recentadvances and future directions. Journal of Clinical andConsulting Psychology.99Kerns, R.D, Jr., Turk, D.C., Holzman, A.D., & Rudy, T.E. (1986).Comparison of cognitive—behavioral and behavioral approachesto the outpatient treatment of chronic pain. The ClinicalJournal of Pain, 1, 195—203.King-Farlow, J., & Bosley, R. (1985). Self-formation and themean (Programmatic remarks on self—deception). In M.W.Martin (Ed.), Self-deception and self-understanding (pp.195—220). Lawrence, Kansas: University Press of Kansas.Kipp, D. (1985). Self-deception, inauthenticity, and weakness ofwill. In M.W. Martin (Ed.), Self—deception and self-understanding (pp. 261-283). Lawrence, Kansas: UniversityPress of Kansas.Kieck, R.E., Vaughan, R.C., Cartwright-Smith, J., Vaughan, K.B.,Colby, C.F., & Lanzetta, J.T. (1976). Effects of beingobserved on expressive, subjective and physiologicalresponses to painful stimuli. Journal of Personality andSocial Psychology, 34, 1211-1218.Kolm, S-C. (1986). (trans. by N. Thom). The Buddhist theory of‘no-self’. In J. Elster (Ed.), The multiple self (pp. 233-265]). Cambridge: Cambridge University Press.Koopman, C., & Eisenthal, S. (1984). Ethnicity in the reportedpain, emotional distress and requests of medicaloutpatients. Social Science and Medicine, 18, 487—490.Kopel, S., & Arkowitz, H. (1974). Role playing as a source ofself—observation and behavior change. Journal ofPersonality and Social Psychology, 29, 677—686.Kotarba, J. (1983). Chronic pain: Its social dimensions.Beverly Hills CA: Sage Press.Kraut, R.E. (1982). Social presence, facial feedback, andemotion. Journal of Personality and Social Psychology, 42,853—863.Krebs, D., Denton, K., Higgins, N.C. (1988). On the evolutionof self-knowledge and self-deception. In K. McDonald (Ed.),Sociobiological perspectives on human development (pp. 105-129). New York: Springer Verlag.Lack, D.Z. (1982). Women and pain: Another feminist issue.Women and therapy, 1, 55-64.Laird, J. (1984). The real role of facial response in theexperience of emotion: A reply to Tourangeau and Ellsworth,and others. Journal of Personality and Social Psychology,.4.1, 909—917.100Lander, J., & Fowler-Kerry, S. (1989). Gender effect in painperception. Perceptual and Motor Skills, 68, 1088-1090.Lanzetta, J.T., Cartwright-Smith, J., & Kieck, R.E. (1976).Effects of nonverbal dissimulation on emotional experienceand autonomic arousal. Journal of Personality and SocialPsycholoqy, 33, 354—370.Large, R.G. (1986). DSM-III diagnoses in chronic pain:Confusion or clarity? The Journal of Nervous and MentalDisease, 174, 295—303.Lawson, K., Reesor, K.A., Keefe, F.J., & Turner, J.A. (1990).Dimensions of pain—related cognitive coping: Cross—validation of the factor structure of the Coping StrategyQuestionnaire. Pain, 43, 195-204.Lazarus, R.S. (1979). Positive denial: The case for not facingreality. Psychology Today, 13, 44-60.Lazarus, R.S., & Folkman, S. (1984). Stress, appraisal, andcoping. New York: Springer.Lazarus, R.S., & Folkman, S. (1987). Transactional theory andresearch on emotions and coping. In L. Laux & G. Vo-ssel(Eds.), Personality in biographical stress and copingresearch. European Journal of Personality.Lee, D. (1985). Facial action determinants of pain iudqment.Unpublished doctoral dissertation, University of BritishColumbia, Vancouver.LeResche, L., & Dworkin, S.F. (1988). Facial expressions ofpain and emotions in chronic TMD patients. Pain, , 71-78.Lerman, C.E. (1987). Rheumatoid arthritis: Psychologicalfactors in the etiology, course, and treatment. ClinicalPsychology Review, 7, 413-425.Linden, W., Paulhus, D.L., & Dobson, K.S. (1986). Effects ofresponse styles on the report of psychological and somaticdistress. Journal of Clinical and Consulting Psychology,54, 309—313.Livingstone, W.K. (1943). Pain mechanisms: A physiologicinterpretation of causalgia and its related states. NewYork: Macmillan.Lockard, J.S., & Paulhus, D.L. (1988). Self—deception: Where dowe stand? In J.S. Lockard & D.L. Paulhus (Eds.), Selfdeception: An adaptive mechanism? (pp. 251-255). EnglewoodCliffs, NJ: Prentice Hall.101Lord, C.G. (1989). The “disappearance” of dissonance in an ageof relativism. Personality and Social Psychology Bulletin,15, 513—518.Lykken, D.T. (1984). Detecting deception in 1984. AmericanBehavioral Scientist, 27, 481-499.Martelli, M.F., Auerbach, S.M., Alexander, J., & Mercuri, L.G.(1987). Stress management in the health care setting:Matching interventions with patient coping styles. Journalof Clinical and Consulting Psychology, 55, 201-207.Martin, M.W. (1985). In M.W. Martin (Ed.), Self-deception andself—understanding (pp. 1—27). Lawrence, Kansas:University Press of Kansas.Matsumoto, D. (1987). The role of facial response in theexperience of emotion: More methodological problems and ameta—analysis. Journal of Personality and SocialPsychology, 52, 769—774.McCrae, R.R. (1984). Situational determinants of copingresponses: Loss, threat, and challenge. Journal ofPersonality and Social Psychology, 46,919-928.-Mechanic, D. (1961). The concept of illness behavior. Journalof Chronic Disease, 15, 189—194.Mechanic, D. (1986). Illness behaviour: An overview. In S.McHugh & T.M. Vallis (Eds.), Illness behavior: Amultidisciplinary model (pp. 101-109). New York: Plenum.Mele, A.R. (1987). Irrationality: An essay on akrasia, self-deception and self—control. Oxford: Oxford UniversityPress.Mellor, S., Conroy, L., & Masteller, B.K. (1986). Comparativetrait analysis of long—term recovering alcoholics.Psychological Reports, 58, 411-418.Meizack, R. (1984). Measurement of the dimensions of painexperience. In B. Bromm (Ed.), Pain measurement in man.Neurophysiological correlates of pain (pp. 327-348).Amsterdam: Elsevier Science Publishers.Meizack, R. (1975). The McGill Pain Questionnaire: Majorproperties and scoring methods. Pain, 1, 277-299.Meizack, R., & Wall, P. (1982). The challenge of pain.Harmondsworth, England: Penguin Books.Milgram, S. (1965). Some conditions of obedience anddisobedience to authority. Human Relations, 18, 57-76.102Miller, M.E., & Bowers, K.S. (1986). Hypnotic analgesia andstress inoculation in the reduction of pain. Journal ofAbnormal Psychology, 95, 6-14.Nillham, J., & Kellogg, R.W. (1980). Need for social approval:Impression management or self—deception? Journal ofResearch in Personality, 14, 445-457.Motokawa, T. (1989). Sushi science and hamburger science.Perspectives in Biology and Medicine, 32, 489-504.Nisbett, R.E., & Wilson, T.D. (1977). Telling more than we canknow: Verbal reports on mental processes. PsychologicalReview, 84, 231—259.Parkes, K.R. (1984). Locus of control, cognitive appraisal, andcoping in stressful episodes. Journal of Personality andSocial Psychology, 46, 655—668.Patrick, C.J., Craig, K.D., & Prkachin, K.M. (1986). Observerjudgments of acute pain: Facial action determinants.Journal of Personality and Social Psychology, 50, 1291-1298.Pankratz, L., & Erickson, R.C. (1990). Two views of malingering.The -Clinical Neuropsychologist, 4, 379-389.Paulhus, D. (1982). Individual differences, self—presentation,and cognitive dissonance: Their concurrent operation inforced compliance. Journal of Personality and SocialPsychology, 43, 838-852.Paulhus, D.L. (1984). Two-component models of socially desirableresponding. Journal of Personality and Social Psychology,4, 598—609.Paulhus, D.L. (1986). Self-deception and impression managementin test responses. In A. Angleitner & J.S. Wiggins (Eds.),Personality assessment via questionnaires (pp. 143-159).New York: Springer-Verlag.Paulhus, D.L., Fridhandler, B., & Hayes, S. (in press).Psychological defense: Contemporary theory and research.In S.R. Briggs, R. Hogan, & W. Jones (Eds.), Handbook ofpersonality psychology. New York: Academic Press.Paulhus, D.L., & Reid, D.R. (1991). Enhancement and denial insocially desirable responding. Journal of Personality andSocial Psychology, 60, 307—317.Paulhus, D.L., & Suedfeld, P. (1988). A dynamic complexity modelof self-deception. In J.S. Lockard & D.L. Paulhus (Eds.),Self—deception: An adaptive mechanism? (pp. 132-145).Englewood Cliffs, NJ: Prentice Hall.103Pearlin, L.I., & Schooler, C. (1978). The structure of coping.Journal of Health and Social Behavior, 19, 2-21.Pears, D. (1986). The goals and strategies of self—deception.In J. Elster (Ed.), The multiple self (pp. 59-77).Cambridge: Cambridge University Press.Pennebaker, J.W., & Chew, C.H. (1985). Behavioral inhibitionand electrodermal activity during deception. Journal ofPersonality and Social Psychology, 49, 1427—1433.Pennebaker, J.W. (1982). The psychology of physical symptoms.New York: Springer Verlag.Pennebaker, J.W., & Skelton, J.A. (1978). Psychologicalparameters of physical symptoms. Personality and SocialPsychology Bulletin, 4, 524-530.Penner, L.A. (1986). Social psychology: Concepts andapplications. St Paul: West Publishing.Philips, H.C. (1983). Assessment of chronic headache behavior.In R. Meizack (Ed.), Pain measurement and assessment (pp.155-163). New York: Raven.Philips, H.C. (1987). Avoidance behaviour and its role insustaining chronic pain. Behaviour Research and Therapy,25, 273—279.Power, P.W. (1984). Adolescent reaction to parental neurologicalillness: Coping and intervention strategies. PediatricSocial Work, 3, 45-52.Price, D.D. (1984). Roles of psychophysics, neuroscience, andexperiential analysis in the study of pain. In L. Kruger &J. C. Liebeskind (Eds), Advances in pain research and therapy(Vol. 6, pp. 341-355). New York: Raven.Price, D.D., & Barber, J. (1987). An analysis of factors thatcontribute to the efficacy of hypnotic analgesia. Journalof Abnormal Psychology, 96, 45-51.Price, D.D., Barrell, J.J., & Gracely, R.H. (1980). Apsychophysical analysis of experiential factors thatselectively influence the affective dimension of pain.Pain, 8, 137—149.Price, D.D., Harkins, S.W., & Baker, C. (1987). Sensoryaffective relationships among different types of clinicaland experimental pain. Pain, 28, 297-307.104Price, D.D., McGrath, P.A., Raf ii, A., & Buckingham, B. (1983).The validation of visual analogue scales as ratio scalemeasures for chronic and experimental pain. Pain, 17, 45—56.Prkachin, K.M., & Cameron, R. (1990). Behavior and cognitivetherapies. In E. Tunks, A. Bellissimo & R. Roy (Eds.),Chronic pain: Psychosocial factors in rehabilitation (2nded.; pp. 144—183). Melbourne, Florida: Krieger.Prkachin, K.M., & Mercer, S.R. (1989). Pain expression inpatients with shoulder pathology: Validity, properties andrelationship to sickness impact. Pain, 39, 257—265.Quattrone, G.A., & Tversky, A. (1984). Causal versus diagnosticcontingencies: On self—deception and on the voter’sillusion. Journal of Personality and Social Psychology, ,237—248.Rachman, S.J., & Philips, C. (1980). Psychology and behavioralmedicine. Cambridge: Cambridge University Press.Reesor, K.A., & Craig, K.D. (1988). Medically incongruentchronic bacic pain: Physical limitations, suffering, andineffective coping. Pain, 32, 35-45.Revenson, T.A. (1981). Coping with loneliness: The impact ofcausal attributions. Personality and Social PsychologyBulletin, 7, 565—571.Romano, J.M., & Turner, J.A. (1985). Chronic pain anddepression: does the evidence support a relationship?Psychological Bulletin, 97, 18—34.Rorty, A.O. (1980). Self-deception, akrasia, and irrationality.Social Science Information, 19, 905-922.Rorty, A.O. (1986). Self-deception, akrasia and irrationality.In J. Elster (Ed.), The multiple self (pp. 115-131).Cambridge: Cambridge University Press.Rosenstiel, A.K., & Keefe, F.J. (1983). The use of copingstrategies in chronic low back pain patients: Relationshipto patient characteristics and current adjustment. Pain,1Z, 33—44.Ross, S.L., Gil, K.M., & Keefe, F.J. (1988). Learned responsesto chronic pain: Behavioral, cognitive, andpsychophysiological. In R.D. France & K.R.R. Krishnan(Eds.), Chronic pain (pp. 228-243). Washington D.C.:American Psychiatric Press.105Ross, S.L., Keefe, F.J., & Gil, K.M. (1988). Behavioral conceptsin the analysis of chronic pain. In R.D. France & K.R.R.Krishnan (Eds.), Chronic pain (pp. 104-114). WashingtonD.C.: American Psychiatric Press.Rowat, K.M., & Jeans, M.E. (1989). A collaborative model ofcare: Patient, family and health professionals. In P.D.Wall & R. Meizack (Eds.), Textbook of pain (2nd ed., pp.1010-1014). New York: Churchill Livingstone.Sackeim, H.A., & Gur, R.C. (1978). Self-deception, Self-confrontation, and consciousness. In G.E. Schwartz & D.Shapiro (Eds.), Consciousness and self-regulation (pp. 139-197). New York: Plenum.Sackeim, H.A., & Gur, R.C. (1979). Self-deception, other-deception, and self-reported psychopathology. Journal ofConsulting and Clinical Psychology, 47, 213—215.Sakai, H., & Andow, K. (1980). Attribution of personalresponsibility and dissonance reduction. JapanesePsychological Research, 22, 32-41.Scheier, M.F., & Carver, C.S. (1985). Optimism, coping, andhealth: Assessment and implications of generalized outcomeexpectancies. Health Psychology. 4, 219-247.Scheier, M.F., & Carver, C.S. (1987). Dispositional optimism andphysical well—being: The influence of generalized outcomeexpectancies on health. Journal of Personality, 55, 169-210.Schelling, T.C. (1986). The mind as a consuming organ. In J.Elster (Ed.), The multiple self (pp. 177—195). Cambridge:Cambridge University Press.Scher, S.J., & Cooper, J. (1989). Motivational basis ofdissonance: The singular role of behavioral consequences.Journal of Personality and Social Psychology, 56, 899-906.Schlenker, B.R. (1982). Translating actions into attitudes: Anidentity-analytic approach to the explanation of socialconduct. In L. Berkowitz (Ed.), Advances in experimentalsocial psychology (Vol. 15, pp. 194-247). New York:Academic.Sigall, H., & Page, R. (1971). Current stereotypes: A littlefading, a little faking. Journal of Personality and SocialPsychology, 18, 247—255.Silver, M., Sabini, J., & Miceli, M. (1989). On knowing self—deception. Journal for the Theory of Social Behaviour, 19,213—227.106Smith, T.W., Aberger, E.W., Follick, M.J., & Ahern, D.K. (1986).Cognitive distortion and psychological distress in chroniclow back pain. Journal of Consulting and ClinicalPsychology, 54, 573—575.Snyder, C.R. (1985). Collaborative companions: Therelationship of self—deception and excuse making. In M.W.Martin (Ed.), Self—deception and self-understanding (pp. 35-51). Lawrence, Kansas: University Press of Kansas.Spielberger, C.D., Gorsuch, R.L., Lushene, R.E. (1970). Manualfor the State-Trait Anxiety Inventory. Palo Alto, Calif.:Consulting Psychologist Press.Spinhoven, P., & Linssen, A.C.G. (1991). Behavioral treatment ofchronic low back pain. I. Relation of coping strategy use tooutcome. Pain, 45, 29-34.Steele, C.M. (1988). Psychology of self-affirmation: Sustainingthe integrity of the self. Advances in Experimental SocialPsychology, 21, 261—302.Stone, A.A., & Neale, J.M. (1984). New measure of daily coping:Development and preliminary results. Journal of Personalityand Social Psychology, 4&, 892-906.-Suls, J., & Fletcher, B. (1985). The relative efficacy ofavoidant and nonavoidant coping strategies. A meta—analysis. Health Psychology, 4, 249—288.Swalm, D., & Craig, K.D. (submitted for publication). Placeboeffects on verbal and nonverbal expression of pain.Szabados, B. (1985). The self, its passions and self—deception.In M.W. Martin (Ed.), Self-deception and self-understanding(pp. 143-168). Lawrence, Kansas: University Press ofKansas.Taylor, S.E., & Brown, J.D. (1988). Illusion and well-being: Asocial psychological perspective on mental health.Psychological Bulletin, 103, 193—210.Tedeschi, J.T., Gaes, G.G., Norman, N., & Melburg, V. (1986).Pills and attitude change: Misattribution of arousal orexcuses for negative actions? The Journal of GeneralPsychology, 113, 309—328.Tennen, H., & Herzberger, S. (1985). Ways of Coping Scale. InD.J. Keyser & R.C. Sweetland (Eds.), Test critiques (Vol.III; pp. 686-697). Kansas City: Test Corp. of America.Tesser, A., & Paulhus, D. (1983). The definition of self:Private and public self—evaluation management strategies.Journal of Personality and Social Psychology, 44, 672-682.107Tetlock, P.E., & Manstead, A.S.R. (1985). Impression managementversus intrapsychic explanations in social psychology: Auseful dichotomy? Psychological Review, 92, 59-77.Totman, R. (1975). Cognitive dissonance and the placeboresponse: The effect of differential justification forundergoing dummy injections. European Journal of SocialPsychology, 5, 441-456.Tunks, E., & Bellissimo, A. (1988). Coping with the copingconcept: A brief comment. Pain, 34, 171-174.Tunks, E., & Roy, R. (1990). Chronic pain and the occupationalrole: Clinical issues. In E. Tunks, A. Bellissimo & R. Roy(Eds.), Chronic pain: Psychosocial factors inrehabilitation (2nd ed., pp. 59-73). Melbourne, Florida:Krieger.Turk, D.C., Meichenbaum, D., & Genest, M. (1983). Pain andbehavioral medicine: A cognitive—behavioral perspective.New York: Guilford.Turk, D.C., & Rudy, T.E. (1986). Assessment of cognitive factorsin chronic pain:- A worthwhile- enterprise? Journal o-fClinical and Consulting Psychology. 54, 760-768.Turk, D.C., Rudy, T.E., & Flor, H. (1985). why a familyperspective on pain? International Journal of FamilyTherapy, 7, 223—234.Turner, J. (1990). Coping and chronic pain. Paper presented atthe 1990 International Association for the Study of Pain.Tursky, B. (1974). Physical, physiological, and psychologicalfactors that affect pain reaction to electric shock.Psychophysiology, 10, 95-112.Tursky, B., & O’Connell, D. (1972). Reliability andinterjudgment predictability of subjective judgments ofelectrocutaneous stimulation. Psychophysiology. 9, 290-295.Violon, A. (1985). Family etiology of chronic pain.International Journal of Family Therapy, 7, 223—234.Vitaliano, P.P., Russo, J., Carr, J.E., Maiuro, R.D., & Becker,J. (1985). The Ways of Coping Checklist: Revision andpsychometric properties. Multivariate Behavioral Research,20, 3—26.Wack, J.T., & Turk, D.C. (1984). Latent structure of strategiesused to cope with nociceptive stimulation. HealthPsychology, 3, 27-43.108Wade, J.B., Price, D.D., Hamer, R.M., Schwartz, S.M., & Hart,R.P. (1990). An emotional component analysis of chronicpain. Pain, 40, 303—310.Warner, R. (1980). Deception and self-deception in shamanismand psychiatry. International Journal of Social Psychiatry,26, 41—52.Weiner, B. (1986). Attribution, emotion, and action. In R.M.Sorrentino & E.T. Higgins (Eds.), Handbook of motivation andcognition (pp. 281-312). New York: Guilford.Weissman, H.N. (1990). Distortions and deceptions in selfpresentation: Effects of protracted litigation in personalinjury cases. Behavioral Sciences and the Law, 8, 67—74.Welles, J.F. (1981). Sociobiology of self-deception. HumanEthology Newsletter, 3, 14-19.Werth, L.F., & Flaherty, J. (1986). A phenomenological approachto human deception. In R.W. Mitchell & N.S. Thompson(Eds.), Deception: Perspectives on human and non—humandeceit. Albany: State University of New York Press.Wesley, A.L., Gatchel, R.J., Polatin, P.B., Kinney, R.K., &Mayer, T.G. (1990). Differentiation between somatic andcognitive/affective components in commonly used measurementsof depression in patients with chronic low—back pain.Spine, 16(6 Supplement), S213—S215.White, N.F. (1990). A socio-ecological model for pain. In E.Tunks, A. Bellissiirio & R. Roy (Eds.), Chronic pain:Psychosocial factors in rehabilitation (2nd ed., pp. 74-103). Melbourne, Florida: Krieger.Wilson, T.D. (1985). Self-deception without repression: Limitson access to mental states. In M.W. Martin (Ed.), Self—deception and self-understanding (pp. 95-116). Lawrence,Kansas: University Press of Kansas.Wolff, B.B. (1978). Behavioural measurement of human pain. InR.A. Sternbach (Ed.), The psychology of pain (pp. 129-168).New York: Raven.Zimbardo, P.G., Cohen, A.R., Weisenberg, M., Dworkin, L., &Firestone, I. (1966). Control of pain motivation bycognitive dissonance. Science, 151, 217—219.Zimbardo, P.G., Cohen, A.R., Weisenberg, M., Dworkin, L., &Firestone, I. (1969). The control of experimental pain. InP.G. Zimbardo (Ed.), The cognitive control of motivation.The consequences of choice and dissonance (pp. 100—125).Glenview, Ill.: Scott, Foresman & Co.109Zuckerman, M., DePaulo, B.M., Rosenthal, R. (1981). Verbal andnonverbal communication of deception. Advances inExperimental Social Psychology, 14, 1-59.Zuckerman, M., DePaulo, B.M., Rosenthal, R. (1986). Humans asdeceivers and lie detectors. In P.D. Blanck, R. Buck, R.Rosenthal (Eds.), Nonverbal communication in the clinicalcontext (pp. 13-35). University Park, PA: PennsylvaniaState University Press.Appendices110111A. Visual Analogue Scales (Adapted from Price et al., 1983)Painful: Yes NoSensory intensity:the most intense painsensation imaginableUnpleasantness:the most unpleasantfeeling imaginable112B. FACS Pain-related Action Units (Ekman & Friesen, 1978)AU FACS Name Weight from previous study*1 Inner brow raise —.0652 Outer brow raise -.0754 Brow lowerer .3375 Upper lid raise N/A6 Cheek raiser .4647 Lid tightener N/A (used AU 6 weight)10 Upper lip raise N/A11 Nasolabial deepen N/A12 Lip corner pull .33214 D-impler- -.029-20 Lip stretcher N/A23 Lip tight —.04125 Lips part .20626 Jaw drops N/A (used AU 25 weight)27 Mouth stretches N/A (used AU 25 weight)43 Eyes close N/A45 Blink .188*Swalm & Craig, 1991, used factor score coefficients derived fromthe first factor of a Principle Components Factor Analysis of themost frequently occurring AUs in shock-induced pain. These arelisted where available, otherwise weights were used from asimilarly occurring AU (as in the cases of AU 7, 26, and 27) orelse weights were not used since they were not available (listedas N/A in the table).113C. Experimental Participation Consent Form: Studies 1 & 2Study: Pain DisplayThe principle investigator of this study is Dr. Kenneth D. Craigof the Psychology Department, telephone 228-3948. Funding hasbeen provided by the SSHRC.We are attempting to define the normal responses to pain stimuliso that we might contrast these with chronic pain patients’responses. For that reason, we are using a standard procedurethat involves carefully measured pain stimuli——electric stimuli——delivered to the forearm through a 1-inch electrode.The first series of shocks will begin at current intensities thatcannot be detected and will increase each time in small amounts,up to a level that is no longer tolerable. At any time duringthis sequence of shocks, simply saying “stop” will end the seriesof shocks, and no further increases will be delivered. Followingthe series of increasing shocks, another series of 5 shocks willbe administered; these will be within the tolerable range andwill be delivered in a random order of intensities. These 5shocks wilL be repeated twice-more under varying conditions.After each stimulus, you will rate the discomfort on a couple ofscales. A videotaping of the procedure will capture yournonverbal reactions. Your videotape will be shown to a selectfew research assistants and students who are involved in thisresearch project. You will also be asked to complete somequestionnaires designed to measure how you deal with theexperience. The entire procedure should take less than 60minutes for which you may receive course credit if yourinstructor has agreed, or you may be paid $6.00 for yourparticipation. All of your responses will be kept confidentialby using a number rather than your name.If you have any questions at any time, please feel free to askthe experimenter, Deiphin Swalm.I hereby consent to participate in the study as described. Iunderstand that the risks to me as a subject are minimal. Ifurther acknowledge that I am aware that I can withdraw fromparticipation in the study at any time and I may refuse to answerany questions without prejudice or loss of remuneration.Date:_____________________Signature:__________________114D. Protocol pertaining to manipulation phase for Study 1.(Read to subject:)THE NEXT SERIES IS MORE STIMULI IN A MIXED UP ORDER.BUT THIS TIME YOU MAY BE ASKED TO CHANGE YOUR FACIAL EXPRESSION.THE REASON FOR THIS IS THAT WE WISH TO USE YOUR TAPE FOR TRAININGSTUDENTS AS DATA CODERS. STUDENTS LIKE YOURSELF WILL BE TAUGHTTO JUDGE PAIN PATIENTS’ FACES FOR THE DEGREE OF PAIN THAT THEYFEEL, AND TO DETECT WHEN THE PATIENTS FAKE MORE OR LESS PAIN.PLEASE THROW A DIE TO DETERMINE WHICH CONDITION YOU’LL BE IN:(hand die to S)*A 1 OR 2 MEANS YOU’RE IN THE EXAGGERATE CONDITION,A 3 OR 4 MEANS YOU’RE IN THE DIMINISH CONDITION,WHILE A 5 OR 6 MEANS YOU’RE IN THE CONTROL CONDITION.(give S die to toss; put up sign for Exaggerate, Diminish, orControl on mirror)WHAT THIS MEANS IS THAT YOU ARE TO{EXAGGERATE YOUR FACIAL EXPRESSION / DIMINISH YOUR FACIALEXPRESSION / JUST ACT NATURALLY AS BEFORE} DURING THIS NEXTSERIES OF STIMULI WHILE I VIDEOTAPE YOU.(MAKE IT LOOK AS IF YOU’RE IN MUCH {MORE/LESS} DISCOMFORT THANYOU REALLY ARE IN.)-NEXT, I NEED YOU TO TOSS THE DIE TO DETERMINE WHETHER YOUR TAPEWILL BE USED AS A TRAINING TAPE OR AS A TESTING TAPE.(give S die to toss; put up sign for Training or Testing on tablein front of subject, facing mirror; explain the nature of thatcondition:)(training condition:) THIS MEANS THAT YOUR TAPE WILL BE USED TOTRAIN OUR STUDENT DATA CODERS. THEY WILL LEARN TO CODE THEDEGREE OF PAIN, AND WHETHER OR NOT PEOPLE ARE FAKING MORE OR LESSPAIN USING YOUR TAPE AND THE TAPES OF OTHER SUBJECTS.(testing condition:) THIS MEANS THAT YOUR TAPE WILL BE USED TOTEST OUR STUDENT DATA CODERS. THEY WILL BE TESTED ON WHETHERTHEY CAN CODE THE DEGREE OF PAIN, AND WHETHER OR NOT PEOPLE AREFAKING MORE OR LESS PAIN USING YOUR TAPE AND THE TAPES OF OTHERSUBJECTS. THOSE CODERS WHO CAN’T PASS THE TEST WON’T BE HIRED.*In a study of cognitive dissonance and shock—induced pain, Sakai& Andow (1980) had subjects toss a die to enhance their sense ofresponsibility for the subsequent behavior that they engaged in.115E. Protocol pertaining to manipulation phase for Study 2.(Read to subject:)THE NEXT SERIES IS MORE STIMULI IN A MIXED UP ORDER.BUT THIS TIME YOU MAY BE ASKED TO STATE A FALSE ATTITUDEREGARDING THE PAIN EXPERIENCED.WE’RE GOING TO SHOW YOUR TAPE TO STUDENTS LIKE YOURSELF WHO WILLBE IN A SECOND STUDY.THAT STUDY EXAMINES THE EFFECT THAT SEEING SOMEONE IN PAIN HAS ONAN OBSERVER’S PAIN EXPERIENCE.I CAN GIVE YOU A CHOICE HERE ABOUT WHICH CONDITION YOU’LL BE IN.IN THE EXPERIMENTAL CONDITIONS, YOU’D BE ASKED TO SAY THAT THEPAIN YOU’D EXPERIENCED WAS MORE OR LESS THAN WHAT YOU’D EXPECTEDIT TO BE.IN THE CONTROL CONDITION, YOU’D BE ASKED TO SAY THAT THE PAIN WASABOUT THE SAME AS WHAT YOU’D EXPECTED IT TO BE.DO YOU WISH TO MAKE A CHOICE OR WOULD YOU LIKE TO LEAVE IT UP TOCHANCE AND PICK A CONDITION FROM THIS BOX?(pause)OKAY, GO AHEAD AND PICK ONE FROM THE BOX.WHAT THIS MEANS IS THAT AFTER THE NEXT SERIES, I’LL COME IN ANDASK YOU ‘HOW WAS THE PAIN THAT TIME?’I’D LIKE YOU TO SAY:(read off of statement sheet)ARE YOU WILLING TO SAY THIS? (pause)OKAY. GO AHEAD AND PRACTISE SAYING THIS OUT LOUD FOR ME.116F. Prepared Statements for Study 2.The following statements were typed on separated strips ofpaper; half also had the negative consequences paragraph added(see below):1. My pain was much worse than I’d expected. It really hurt alot.2. My pain was much less than I’d expected. It really didn’thurt much.3. My pain was about the same as I’d expected. It felt aboutthe same.The negative consequences manipulation was an additionalparagraph on the prepared statement sheet which explained thatstudents who view the videotape would experience worse pain dueto a particular effect relevant to the statement condition.1. Subjects in the “pain worse” condition read:Expectation effect: As you know, one’s expectations are powerfulpredictors of what one experiences. Subjects who see you saythat you’re in a lot of pain will expect to feel a lot of pain,too. As a result, they’ll feel more pain after viewing yourvideotape.2. Subjects in the “pain less” condition read:Surprise effect: As you know, when one is emotionally aroused,one becomes more physically aroused and thus more sensitive topain. After seeing you say that you’re in little or no pain,subjects will be unpleasantly surprised during their shocks andthey’ll become emotionally aroused. As a result, they’ll feelmore pain after viewing your videotape.3. Subjects in the “pain same” condition read:Empathy effect: As you know, when one is emotionally aroused,one becomes more physically aroused and thus more sensitive topain. Subjects who see you say that you’re in pain will be verydistressed and emotionally aroused. As a result, they’ll feelmore pain after viewing your videotape.117G. Post-Experimental Questionnaire: Study 1Your answers to the following questions will help us to evaluatethis study; all answers will be kept confidential. Please answereach question carefully and honestly. Circle the appropriatenumber or write your answer for each question.1. In your own words, please state what you think the purpose(s)of this study was.2. How responsible do y feel for any negative consequences(e.g., not getting hired) that might occur to the trainees whoview your videotape?1.. .2...3...4...5. ..6...7not at all very responsible3. How responsible is the research team for any negativeconsequences (e.g., not getting hired) that might occur to thetrainees who view your videotape?1.. .2...3...4...5...6...7not at all very responsible4. How responsible are the student data coders themselves forany negative consequences (e.g., not getting hired) that mightoccur to them?1.. .2...3.. .4...5...6.. .7not at all very responsible1185. If you were asked to change your facial expression, pleaseanswer these questions:a. How uncomfortable did you feel about misleading data codersin training?1.. .2.. .3.. .4.. .5.. .6.. .7not at all very uncomfortableb. How did you go about changing your facial expression?c. Did this seem to alter your experience of pain?d. If so, how was your pain experience altered?119H. Post-Experimental Questionnaire: Study 2Please answer the following questions. Some are repeated fromthe earlier questionnaire.1. In your own words, please state what you think the purpose(s)of this study was.2. on average, how much did you dislike the shock stimuli?1.. .2.. .3...4...5...6...7didn’t mind it strongly disliked it3. How responsible do you feel for any negative consequences(e.g., more sensitive to pain) that might occur to the studentswho view your videotape?1...2.. .3...4...5...6...7not at all very responsible4. If you were asked to lie about your pain experience, pleaseanswer these questions:a. How uncomfortable or guilty did you feel about telling alie?1.. .2.. .3.. .4.. .5.. .6.. .7not at all very uncomfortable -b. Did lying about your pain seem to alter your experience ofpain?c. If so, how was your pain experience altered?120I. Debriefing: Study 1Study: Self—deception as a Source of Incongruous Pain DisplayDr. Kenneth D. Craig & Deiphin SwalmDepartment of Psychology, UBCThank-you very much for participating in this study. Yourcontribution to our understanding of pain will hopefully lead tobetter understanding of and treatments for pain patients.In addition to the reasons outlined earlier, this study wasdesigned to examine the impact that “telling a lie” with the facehas on subsequent pain responses. Some subjects exaggerate theirfacial response, others diminish it, while a third group does notchange their facial expression. We expect the first two groupsto experience more and less pain, respectively, as a result ofthis manipulation.Furthermore, it is believed that when subjects are given a weakreason to deceive others (who in turn might be adverselyaffected), these subjects also deceive themselves. Past studieshave shown that when people lie about their attitude on someissue, they come to believe their own lie, especially if theytell the lie to another student. We are attempting to generalizethis finding to the pain domain. This process of self-deceptionis expected to change the pain response in the direction of thedeceptive communication.Actually, we do not presently plan to show your videotape to datacoders in training, but only to trained data coders assigned tothis research project. In general, most subjects comply with theexperimenter’s request to tell a lie, and so your compliance isperfectly normal.121The various questionnaires that you completed were designed tomeasure self—deception and other constructs that might be closelyrelated (such as anxiety and coping strategies).Finally, we are interested in how the face changes during andafter deception; for that reason a trained coder will list thediscrete movements in the face captured on videotape.Since we will be asking some of your classmates or acquaintancesto participate in this study, it is important that you avoiddiscussing this study’s purpose with potential subjects. Thank-you very much for your cooperation. If you have any furtherquestions about the study, please call the experimenter, DelphinSwalm, at 228—4927.Now that you are aware of the true nature of the study, you mayrefuse to permit use of the videotape or other data withoutprejudice or loss of remuneration. Otherwise, please sign belowto indicate your consent to permit our use of your videotape andother data.Signature:122J. Debriefing: Study 2Debriefing: Study number 90-022Title: Self—deception as a Source of Incongruous Pain DisplayDr. Kenneth D. Craig & Deiphin Swaim, Dept. of Psychology, UBCThank-you very much for participating in this study. Yourcontribution to our understanding of pain will hopefully lead tobetter understanding of and treatments for pain patients.In addition to the reasons outlined earlier, this study wasdesigned to examine the impact that “telling a lie” about painhas on subsequent pain responses. Some subjects are asked to saythey have more or less pain, while a third group says that theyhave about the same pain as before. We expect the first twogroups to experience more and less pain, respectively, as aresult of this manipulation. -Furthermore, it is believed that when subjects are given a weakreason to deceive others (who in turn might be adverselyaffected), these subjects also deceive themselves. Past studieshave shown that when people lie about their attitude on someissue, they come to believe their own lie, especially if theytell the lie to another student. We are attempting to generalizethis finding to the pain domain. This process of self-deceptionis expected to change the pain response in the direction of thedeceptive communication. In general, most subjects comply withthe experimenter’s request to tell a lie, and so your complianceis perfectly normal.Actually, we do not presently plan to show your videotape toother students, but only to trained data coders assigned to thisresearch project. We are interested in how the face changesduring and after deception; for that reason a trained coder willlist the discrete movements in the faces captured on videotape.123The various questionnaires that you completed were designed tomeasure self—deception and other constructs that might be closelyrelated (such as anxiety and coping strategies).Since we will be asking some of your classmates or acquaintancesto participate in this study, it is important that you avoiddiscussing this study’s purpose with potential subjects. Thank-you very much for your cooperation. If you have any furtherquestions about the study, please call the experimenter, DeiphinSwalm, at 822—5280.Now that you are aware of the true nature of the study, you mayrefuse to permit use of the videotape or other data withoutprejudice or loss of remuneration. Otherwise, please sign belowto indicate your consent to permit our use of your videotape andother data.Signature:124K. Study 2. Manipulation phase questionnaireYour answers to the following questions will help us to evaluatethis study; all answers will be kept confidential. Please answereach question carefully and honestly. Circle the appropriatenumber or write your answer for each question.1. Did you have a choice about making a statement regarding yourpain experience? Yes No (tick off one)2. How do you think viewing your videotape will affect othersubj ects?3. How responsible do you feel for any negative consequences(e.g., more sensitive to pain) that might occur to the studentswho view your videotape?- -1...2...3..-.4..5.-..6.-..7 ---not at all very responsible4. If you were asked to lie about your pain experience, pleaseanswer this question:a. How uncomfortable or guilty do you feel about telling alie?l...2...3...4...5...6...7not at all very uncomfortable125L. Study 2: Simple interactive effects analyses for theSensory ScaleFactor Held Other Factors F ratio (df), p valueConstantTime X Intensity .73 (8.432), .66Time X Intensity 1.50 (8.432), .16Time X Intensity 3.04 (8,432), .01Time 1(Pretest)Time 2(Manipulation)Time 3(Posttest)2.25 (2,108), .111.82 (2,108), .1714.97 (2,108), .012.30 (2,108), .114.79 (2,108), .01ExaggerateDiminishControlE/D/C X IntensityE/D/C X IntensityE/D/C X Intensity.413.231.55(8,216)(8,216),(8.216)Control.92• 01• 14Intensity1 X TimeIntensity2 X TimeIntensity3 X TimeIntensity4 X TimeIntensity5 X TimeTime 2(Manipulation)Intensity1 X E/D/C 5.47 (2,54), .01Intensity2 X E/D/C 8.84 (2,54), .01Intensity3 X E/D/C 1.17 (2,54), .32Intensity4 X E/D/C .77 (2,54), .47Intensity5 X E/D/C 1.12 (2,54), .33126Newman-Keuls tests for Control X Intensity3 X Time:Time1 > Time2 or Time3 (p < .01)cellmeans: 73.75 58.3 54.05Newman-Keuls tests for Control X Intensity5 X Time:Time1 > Time2 or Time3 (p < .01)82.25 72.25 74.3Newman-Keuls tests for Time2 X Intensity1 X Condition:Exaggerate or Control > Diminish (p < .01)42.4 41.1 19.95Newman-Keuls tests for Time2 X Intensity2 X Condition:Exaggerate or Control > Diminish (p < .01)51.7 58.1 29.3127N. Study 2: Simple interactive effects analyses for theUnpleasantness ScaleFactor Held Other Factors F ratio (df), p valueConstantExaggerate Time X Intensity 1.11 (8.432), .35Diminish Time X Intensity 1.56 (8.432), .13Control Time X Intensity 3.72 (8,432), .01Time 1 E/D/C X Intensity .32 (8,216), .96(Pretest)Time 2 E/D/C X Intensity 2.64 (8,216), .01(Manipulation)Time 3 E/D/C X Intensity 1.49 (8.216), .16(Posttest)Control Intensity1 X Time .77 (2,108), .47Intensity2 X Time 2.24 (2,108), .11Intensity3 X Time 14.02 (2,108), .01Intensity4 X Time 4.86 (2,108), .01Intensity5 X Time 5.95 (2,108), .01Time 2 Intensity1 X E/D/C 1.60 (2,54), .21(Manipulation)Intensity2 X E/D/C 3.49 (2,54), .04Intensity3 X E/D/C .65 (2,54), .53Intensity4 X E/D/C .39 (2,54), .54Intensity5 X E/D/C 1.39 (2,54), .26128Newman-Keuls tests for Control X Intensity3 X Time:Time1 > Time2 or Time3 (p < .01)cellmeans: 65.55 50.55 44.75Newman-Keuls tests for Control X Intensity4 X Time:Time1 > Time2 but not > Time3 (p < .01)71.6 61.25 65.4Newman-Keuls tests for Control X Intensity5 X Time:Time1 > Time2 or Time3 (p < .01)76.95 65.6 67.45Newman-Keuls tests for Time2 X Intensity2 X Condition:Exaggerate or Control > Diminish (p < .05; n.s. @p < .01)43.85 50.65 30.36129ANOVAR Results with Self-Deception as a GroupingVariable (Sensory Pain Ratings)- Median Split Removing Middle 9 SubjectsSS MS DFN. Study 2:ConditionSelf-DecepCond X S-DF SigofFTimeCond X TimeS-D X TimeCond X S-D X TimeIntensityCond X IntensityS-D X IntensityCond X S-D X mtTime X IntensityCond X Time X mtS-D X Time X mtCond X SD X T X I14025.65279.5511603 . 195488.751088. 631019.47126.02158850.715030.12106. 033753.79867.423153 .82639.763543.177012.82279.555801.602744.38272.16509.7431.5139712.68628.7826.51469.22108.43197. 1179. 97221.452 1.861 .072 1.542 14.244 1.412 2.654 .164 86.248 1.374 .068 1.028 .8716 1.588 .6416 1.77• 167.786.225• 000.236• 077956• 000.214.994.423.544.072.744.0331300. Study 2: ANOVAR Results with Anxiety as a GroupingVariable (Sensory Pain Ratings)- Median Split Removing Middle 2 SubjectsSS MS DF F SigofFCondition 23057.53 11528.76 2 2.96 .060Anxiety 8088.67 8088.67 1 2.08 .155Condition X Anx 5689.78 2844.89 2 .73 .486Time 6455.00 3227.50 2 14.89 .000Condition X Time 2108.78 527.19 4 2.43 .052Anxiety X Time 150.60 75.30 2 .35 .707Cond X Anx X Time 1192.18 298.05 4 1.38 .248Intensity 189628.22 47407.06 4 106.32 .000Cond X Intensity 8216.18 1027.02 8 2.30 .022Anxiety X mt 4406.52 1101.63 4 2.47 .046Cond X Anx X mt 1423.06 177.88 8 .40 .920Time X mt 1708.41 213.55 8 1.78 .080Cond X Time X mt 3516.83 219.80 16 1.83 .026Anx X Time X mt 1451.06 181.38 8 1.51 .152Cond X Anx X T X I 1528.34 95.52 16 .79 .693P. Study 2: ANOVAR Results with CatastrophizingVariable (Sensory Pain Ratings)- Median Split Removing No Subjects131as a GroupingSS MS DF F Sig of FCondition 20206.28 10103.14 2 2.67 .079Catast 247.34 247.34 1 .07 .799Condition X Cat 13343.86 6671.93 2 1.76 .182Time 5490.30 2745.15 2 12.49 .000Condition X Time 2051.82 512.95 4 2.33 .060Cat X Time 106.45 53.23 2 .24 .785Cond X Cat X Time 385.27 96.32 4 .44 .781Intensity 187427.28 46856.82 4 102.61 .000Cond X Intensity 5331.76 666.47 8 1.46 .174Catast X mt 1725.10 431.28 4 .94 .439Cond X Cat X mt 3881.33 485.17 8 1.06 .391Time X mt 1544.27 193.03 8 1.67 .103Cond X Time X mt 3022.94 188.93 16 1.64 .057Cat X Time X mt 476.90 59.61 8 .52 .845Cond X Cat X T X I 3629.33 226.83 16 1.96 .014Q. Study 2:132ANOVAR Results with Sense of Control as a GroupingVariable (Sensory Pain Ratings)- Median Split Removing One Middle SubjectSS MS DF F Sig of FCondition 25726.74 12863.37 2 3.20 .049Control 346.83 346.83 1 .09 .770Condition X Control 472.33 236.17 2 .06 .943Time 5493.23 2746.61 2 12.39 .000Cond X Time 2039.21 509.80 4 2.30 .064Control X Time 50.48 25.24 2 .11 .893Cond X Control X T 368.45 92.11 4 .42 .797Intensity 173274.89 43318.72 4 92.14 .000Cond X Intensity 6449.02 806.13 8 1.71 .096Cond X Control X I 2683.97 335.50 8 .71 .679Time X Intensity 1986.06 248.26 8 2.06 .039Cond X Time X mt 3686.79 230.42 16 1.91 .018Control X Time X I 496.37 62.05 8 .51 .846Cond X Cont X T X I 2222.98 138.94 16 1.15 .304R. Sensory Scale ANOVAR Study 2 Cell MeansExaggerate Condition133Pretest41.055. 666.375. 076.7Manipulation42.451.765.172.578. 8Posttest42 . 652. 184.474.577. 0Diminish ConditionIntensity level12345Pretest34.242.459. 670.072 . 9Manipulation19. 929.356. 164.969. 3Posttest21.532.652.467. 069.9Control ConditionPretest45.355. 473.777.5Manipulation41.158. 158. 370.1Posttest36.850.854. 073.974 . 3Intensity level12345Intensity level12345 82.2 72.2134S. Unpleasantness Scale ANOVAR Study 2 Cell MeansExaggerate ConditionIntensity level Pretest Manipulation Posttest1 35.2 33.5 37.52 52.9 43.8 45.83 64.5 58.4 54.44 73.2 70.7 68.455 79.7 78.2 77.8Diminish ConditionIntensity level Pretest Manipulation Posttest1 33.6 20.6 18.92 43.8 30.4 28.23 58.5 54.7 47.04 70.1 66.9 64.65 74.3 70.6 69.4Control ConditionIntensity level Pretest Manipulation Posttest1 35.2 32.3 30.12 44.9 50.6 42.03 65.5 50.6 44.74 71.6 61.2 65.45 76.9 65.6 67.4135T. Multiple Regression Results Using Self-Deception as aModerator VariablePretest Sensory Ratings (averaged for two lowest intensities),using Diminish (coded = 2) v. Control group (coded = 1):Correlation MatrixGroup Self-Dec Grp X S-D RatingsGroupSelf—Dec —.147Grp X S—D .357 .819Ratings—.309 —.283 —.484Multiple Regression ResultsStep Variable R2 R2 Chg Beta (Egn) Sig1 Grp X S—D .23 .23 —.484 11.65 .0022 Self—Dec .27 .04 .343 6.96 .003Manipulation Sensory Ratings -(averaged fo-r two lowestintensities), using Diminish (coded = 2) v. Control group (coded= 1)Correlation MatrixGroup Self-Dec Grp X S-D RatingsGroupSelf—Dec—. 147Grp X S—D .357 .819Ratings —.525 —.097 —.351Multiple Regression ResultsStep Variable R2 R2 Chg Beta (Eqn) Sig1 Grp .27 .27 —.524 14.42 .0012 Self—Dec .31 .03 —.178 8.17 .001136Pretest Sensory Ratings (averaged for two lowest intensities),using Diminish (coded = 2) v. Exaggerate group (coded = 1):Correlation MatrixGroup Self-Dec Grp X S-D RatingsGroupSelf—Dec—. 063Grp X S—D .436 .820Ratings —.210 —.033 —.245Multiple Regression ResultsStep Variable R2 R2 Chg Beta (Eqn) Sig1 Grp X S—D .06 .06 —.245 2.42 .1282 Self—Dec .15 .09 .512 3.15 .0543 Grp .19 .05 .418 2.86 .050Manipulation Sensory Ratings (averaged for two lowestintensities), using Diminish (coded = 2) v. Exaggerate group(coded = 1)-: -Correlation MatrixGroup Self-Dec Grp X S-D RatingsGroupSelf—Dec —.063Grp X S—D .436 .820Ratings —.461 .104 —.200Multiple Regression ResultsStep Variable R2 R2 Chg Beta (Eqn) Sig1 Grp .21 .21 —.461 10.25 .003


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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


Related Items