Journal of Personality and Social Psychology 1976, Vol. 33, No. 5, 632-646

Evidence of Stress as a Function of Experimentally Altered Appraisal of Stimulus Aversiveness and Coping Adequacy Richard W. J. Neufeld University of Western Ontario Ontario, Canada The present study examined the effects of initial appraisal of stimulus aversiveness and coping adequacy on subsequent evidence of stress. Results indicated that level of appraised coping efficacy had the most pronounced effect during stressor exposure and that the pattern of this effect was dependent on whether the source of the appraised coping resources was "personal" (the subject's own resources) or "environmental" (a bogus technique donated by the experimenter) . In addition, information about personal resources reduced the pre- to poststressor increment in state anxiety and lessened cognitive avoidance of stressor-related stimuli. Results were discussed with respect to differential sensitivity of components of stress response to variation in environmental versus personal sources of coping efficacy, and the specific aspects of cognitive structuring associated with variation in physiological reaction (muscle tension) were noted. The importance of factors associated with appraised coping efficacy was clearly supported, but factors involving the objective (in contrast to defensive) aspects of appraised stressor potency required further investigation, and suggestions for such future research were made.

& Holmes, 1974; Monat, Averill, & Lazarus, 1972), "denying" the aversive properties of the situation or "intellectualizing" about them (cf. Lazarus & Alfert, 1964; Neufeld, 1975), and so on. The present study was primarily concerned with further investigating the former category of cogniitve mediation of stress reaction: the cognitive appraisal of the stressor situation and the appraised potential efficacy of coping resources. Although these two components of cognitive appraisal have been given considerable emphasis at the theoretical level, empirical studies are lacking where the components have been individually manipulated so as to examine their independent and interactive effects. The study reported here was designed to examine these two aspects of appraisal by varying the informational context in which the stressor and available coping resources were presented. In addition, a factor relating This study was supported by Canada Council primarily to the appraisal of coping-resource Grant S73-1S99-XI-SI. Thanks are extended to Anne Newby, Ellen Dickson, and Jon Hartwick for efficacy was examined. This factor involved their assistance in data collection. the source from which the coping resources Requests for reprints should be sent to Richard eminated. Lazarus (1966) has pointed out W. J. Neufeld, Department of Psychology, University of Western Ontario, London, Ontario, Canada that these resources can originate from the N6A SC2. individual (e.g., active escape or nonphysical The importance of cognitive activity as a mediator of stress response has continued to receive considerable emphasis in theoretical frameworks on the effects of stressful stimulation (cf. Lazarus, 1966; Lazarus, Averill, & Opton, 1974; Spielberger, 1972). Such cogniitve activity can be viewed, broadly, as falling into two categories. The first is the individual's appraisal of the potential danger or harm in a given situation (cf. Janis, 1967; Lazarus & Averill, 1972; Spielberger, 1972) in conjunction with appraisal of the available counterharm resources (e.g., Janis, 1967, p. 173; Lazarus, 1966, p. 160). The second is associated with efforts at coping with ongoing stress involving such processes as not thinking about the stressor situation, or on the other hand, dwelling on it in a "vigilant" sense (e.g., Averill & Rosenn, 1972; Houston

632

EVIDENCE OF STRESS

passive escapes involving cognitive coping strategies such as deploying attention away from the stressor, vigilance, denial, etc., as mentioned above), or they can originate from the environment (e.g., the assistance of others, physical conditions that could be to the person's advantage in dealing with the stressor, etc.). Thus, the sources of the appraised coping resources were selected so as to vary somewhat along this environmental-personal dimension. With respect to appraised stressor potency and appraised coping-resource efficacy, certain expectations were advanced, since stress reaction has been considered, in part, as a function of the balance between these two components (Lazarus, 1966, p. 160). If stress response varies with the ratio of appraised stressor potency or aversiveness (SA) to appraised coping efficacy (CE) and if evidence of stress reflects in part the size of this ratio, one might expect variation in stress reaction to correspond to variation in either component individually. However, an interaction between these two factors might also be expected in the following sense. If two different levels of appraised coping efficacy are denoted CE and CE', the difference in the ratios under a given level of appraised stressor aversiveness, SA, would be SA/CE - SA/CE' or SA(1/CE - 1/CE'). Note that as SA increases, the bracketed difference is enhanced, so that the variation in the ratio from CE to CE' varies according to the prevailing level of appraised stimulus aversiveness. In this way, the size of the effects of appraised coping adequacy on evidence of stress might be expected to vary with the magnitude of appraised stressor aversiveness, SA. A similar argument could show that the effects of varied stressor aversiveness might be enhanced under conditions of low appraised coping efficacy and diminished under conditions of high appraised coping efficacy. To summarize the main aims of the study, then, evidence of stress was investigated as a function of (a) appraisal of stressor aversiveness, (b) appraisal of coping-resource efficacy, (c) the source of the appraised coping resources, and (d) the possible interactive effects among these experimental factors, especially the first two.

633

METHOD The subjects were 180 female undergraduates fulfilling introductory psychology course requirements. Only one sex was used for this study, since responses to stressor stimuli on some of the measures employed have been found to differ between males and females (Neufeld & Davidson, 1974b). The experimenter was also a female.

Stressor The shop-safety film, "It Didn't Have to Happen," has been described in detail elsewhere (e.g., Lazarus, Opton, Nomikos, & Rankin, 1965). It depicts three accidents, the first involving finger lacerations, the second, a finger amputation, and the third, a death due to impalement by a board of wood. The entire film (approximately 12 minutes) was shown, as its stressor properties were considered to be partially dependent on the ordered sequence of events portrayed. There were several advantages in using this stressor for the present purposes. Scaling results (Neufeld & Davidson, 1974a) have shown it to have high values on unidimensionally and multidimensionally derived dimensions of judged stress as well as high intersubject agreement on its relationship to these dimensions. The general unfamiliarity of this stimulus to most subjects was expected to enhance ambiguity, thus facilitating potential impact of the experimental manipulations. Also, compared with other stressors (e.g., threat of electric shock), there was less likelihood that the defensive denial strategy, "The stimulus won't actually be presented," would be adopted.

Procedure Subjects were first informed about the general nature of the experiment, and permission was explicitly given for withdrawal at any time. (Thirteen subjects, dispersed approximately equally among the treatment combinations, withdrew before complete data were collected and were replaced). They were then asked to complete two questionnaires. Information about the questionnaires, before and after their completion, differed according to assigned treatment combinations (detailed below). The first questionnaire was the State-Trait Anxiety Inventory (cf. Spielberger, 1972), and the second one was the Inhibition-Facilitation Scale (Ullman, 1962). After a 10-minute break, while the experimenter excused herself to supposedly "score one of the questionnaires" or "calibrate equipment" (depending on the assigned treatment combination), the subject entered the room where the film was to be shown, and physiological recording equipment was attached with the functions explained. The subject was then given a brief description about the film to be shown. The content of this description varied according to the assigned level of appraised potential aversiveness. The subjects then went through a process of "clearing the mind and thinking about a blank wall," Prior to this activity, some subjects were given information about their performance on the second

634

RICHARD W. J. NEUFELD

questionnaire completed at the beginning of the session, whereas other subjects were given information about the "mind-clearing" technique. Whether they received information about the questionnaire or the mind-clearing technique, as well as the specific type of information they each received, was dependant on their assignmed treatment combination. When the mind-clearing exercise had been completed, subjects were shown a panel consisting of the 7 odd-numbered adjectives of the Subjective Stress Scale (cf. Berkun, Bialek, Kern, & Yagi, 1962) with adjacent response buttons. They were asked to select the adjective best describing how they felt at that time (Subjective Stress Scale baseline) and were told to repeat this response each time a green (leader film) screen appeared. After an additional minute to provide a prefilm baseline period for the physiological measures, the film commenced. Approximately IS sec of green leader followed each accident scene to permit the subject to report her subjective stress on the response panel. Following the film, a poststressor questionnaire (described below) was completed. The State-Trait Anxiety Inventory was again administered, and subjects were thoroughly debriefed. They were requested not to communicate about the procedures to others. The 2 X 2 X 3 experimental design consisted of three factors: appraised potential aversiveness (low and medium), the source of defence resources (personal and environmental), and the level of appraised defence efficacy (low, medium, and high). Subjects were randomly assigned to each of the 12 cells (w = lS).

Levels of Presented Aversiveness The description of the film (just prior to its showing) for the low-appraised-aversiveness factor was as follows: This portion of the experiment is designed to study the effects of a film on physiological and psychological reactions. The film you are about to see is a portion of a wood shop safety film. The film portrays three accidents caused by the carelessness of the workmen. In the first accident a worker lacerated his finger in a planer. The second accident resulted in the amputation of two joints of a finger in a milling machine. The final accident occurred when a worker allowed a circular saw to drive a board through the abdomen of an unsuspecting workman. This led to his death in the working surroundings. The film was designed to impress upon the wood-mill workers the importance of good safety practices. The corresponding film description for the mediumappraised-aversiveness factor was as follows: This portion of the experiment is designed to study the effects of a film on physiological and psychological reactions. The film portrays three accidents caused by the carelessness of workmen. The first accident involved the painful laceration of a worker's finger in a planer. The second accident resulted in the traumatic amputation of two

joints of a workman's finger in a milling machine. The final accident occurred when a worker allowed a circular saw to drive a board through the abdomen of an unsuspecting workman. This led to a slow, gory death in the cold working surroundings. The actual frequency of such accidents is high enough to require the explicit message of this film. To illustrate what is meant, in previous research, several unpleasant stimuli, including this one, were presented to subjects such as yourself. They judged this one to be one of the more stressful. Several subjects reported being disturbed for a short period of time after viewing the film. The appraised-threat passages were similar regarding the amount of detail revealed about the stressor, and neither encouraged a specific cognitive defensive orientation, for example, "intellectualization," "denial," vigilance," etc. Effort was made to affect primarily the expected aversiveness of the stressor. Thus, the principle mode of variation in the preceding cognitive manipulations was related mainly to expected stimulus quality (cf. Grings, 1973).

Environmental Versus Personal Source of Defence Resource Subjects given the "environmental-source" manipulation were provided information designed to vary appraisal of 'the prefilm "mind-clearing" technique, whereas those given the "personal-source" manipulation were given information supposedly relating to their responses to the second questionnaire at the beginning of the session. When the experimenter left for 10 minutes after the questionnaires had been completed, subjects in the environmental source condition were told that "equipment was to be calibrated," whereas the personal-defence subjects were told that the experimenter's departure was for the sake of scoring the second questionnaire (the Inhibition-Facilitation Scale). The information about the questionnaires given to those administered the environmental-source treatment was that one of the possible interests of the experimenter was "personality correlates which may or may not be considered after the experiment." Information about the mindclearing technique given to those who received the personal-source treatment was that the procedure helped to quell physiological activity for the sake of polygraph calibration. The addressed sources were considered to vary along the personal-environmental dimension, since the personal-source information (detailed more fully below) purported to relate to the subject's "psychological constitution" as it pertained to the stressor, and the environmental source focused on a technique "donated by the experimenter."

Levels of Presented Coping

Efficacy

The level of appraised personal-resource defence was varied by using the following passage: I have just completed scoring the second questionnaire that you have filled out. It is called the

EVIDENCE OF STRESS I-F Scale [Inhibition-Facilitation Scale]. In the past, we have done research assessing the relationship between scores on the I-F Scale and the ability to deal with the kinds of stress shown in the upcoming movie. We have found that the I-F Scale is a fairly good predictor in that it correlated 0.92 with subjects' ability to deal with the kind of stress shown in the upcoming movie (and 1.0 is a perfect predictor). Therefore the I-F Scale is effective in discriminating between those who have the ability to deal with the kinds of stress involved in today's movie and those who don't. One of the purposes of today's experiment is to determine the effects of feedback regarding where you scored on the I-F Scale on your reactions to viewing the upcoming film. According to our past research with undergraduates your score fell at the percentile rank on the scale. This means that if there were 100 other people in the room, there would be other people better able than yourself to deal with the kinds of stress shown in the movie. For the high-defence manipulation, the reported percentile rank was 92; for the medium-defence treatment, it was 49; and for the low-defence group, it was 13. The passage addressed the subject's performance on the Inhibition-Facilitation Scale rather than the State-Trait Anxiety Inventory, since the former measures repression sensitization (cf. Ullman, 1962) and not anxiety. Thus, it was expected that this procedure would reduce the likelihood of disagreement between the subject's estimate of her own performance on a test and her purported performance as fed back by the experimenter. The Inhibition-Facilitation Scale consists of 44 Minnesota Multiphasic Personality Inventory items. Its concurrent validity as a measure of repression sensitization is supported by its correlations of .94 (psychiatric patients) and .76 (student population) with Byrne's (1961) 156-item repression-sensitization scale (UHmann, 1962). For the environmental-source manipulation, the following passage was used: I would like you to clear your mind and think of a blank wall for 4 minutes before you view this film. This procedure is one of 14 such stressreducing procedures researched for effectiveness in enabling people to deal with the kind of events portrayed by the upcoming scenes. This method is used because it was thought to quiet physiological activity prior to the stimulus, permitting greater "resources for coping during the stimulus presentation." The passage ended as follows for the high-defence treatment: This theory was confirmed, and the research had indicated that this technique ranked second best among all 14 techniques tested. One of the pur-

635

poses of this experiment is to attempt a replication of this finding. For the medium-defence treatment, the ending was as follows: This theory was neither confirmed nor disconfirmed. The research has indicated that this technique ranked 8th from the top among all 14 techniques tested. One of the purposes of the experiment is to investigate this finding. The low-defence treatment consisted of the following passage: Some recent research has raised some questions concerning the effectiveness of this technique as a coping strategy. One study has indicated that this technique ranked second from the bottom among the 14 stress-reducing techniques tested. One of the purposes of this experiment is to investigate this finding. Preexperimental Scaling of Passages An important consideration for adopting these types of passages in testing the null hypotheses was whether or not they had the intended appraisalmodifying effects (Greenwald, 1975). Thus to test the validity of the passages, the following procedure was adopted: 90 undergraduate judges (not participating in the study) scaled three passages each: one of three aversiveness-appraisal passages (one of which was subsequently discarded), one of three personal-resource-defence passages, and one of three environmental-resource-defence passages. Only one passage from each type of manipulation was presented so as to prevent the subjects from simply rank ordering them relative to each other. It should be noted that the judged passages were not identical to the passages subsequently used (as presented above). The appraisal-modifying content was the same, however, in describing the film, reference was made to the last accident scene only.1 The judges were asked to rate (a) "how threatened (anxious, uneasy)" they would feel about viewing the film after having read the appraised aversiveness passage, (b) how adequate they felt the environmental-source technique would be in dealing with the upcoming film, and finally, (c) how "confident and relaxed" they would feel about viewing the film after having been informed of their score on the test according to the personal-resource passage. The judgments were done on 9-point scales tailored to the specific questions. Table 1 presents the F values for each type of manipulation. Duncan's new-multiple-range pro1 It was originally our intention to expose subjects to only the last accident during the actual experiment. However, since the film's stressor properties were subsequently considered to be partly dependent on the ordered sequence of events portrayed (see the above description of the stressor), the entire film was shown.

636

RICHARD W. J. NEUFELD

TABLE 1 JUDGMENTS ABOUT APPRAISAL-MODIFYING PASSAGES

Type of passage

F(2, 87)

Proportion of variance accounted for by the betweenpassage factor

Stimulus aversiveness Personal-source resources Environmental-source resources

44.39*

.489

99.14*

.684

29.26*

.392

* p < .005.

cedure for post-analysis-of-variance comparisons indicated significant differences (p < .01) among all three passage means for each type of manipulation. Table 1 also shows the best-estimate proportions of rating variance (cf. Kirk, 1968) accounted for by the between-passage factors. 2 These judgements confirmed that the substance of each passage had the appraisal-modifying effect intended.

Comment on Experimental Design It should be noted that the levels of appraised resource adequacy for the environmental source involved different "types" of overt manipulations compared with the personal source. Thus the "morphology" of the defence adequacy manipulation was necessarily "confounded" with the source of the defence resource per se. It was thus possible that any main effects or interactions involving the source of defense resources factor could reflect differential effectiveness of the defence adequacy manipulation. On the other hand, tests on the main effect of appraised defence adequacy would be clearly interpretable, as would individual simple main effects tests on defence adequacy within each level of the source of defence factor. This latter consideration was important in assessing whether or not appraised defence adequacy had some effect, regardless of whether the defence eminated from the individual or the environment.

Measures Subjective Stress Scale. This measure consisted of 7 adjectives from an original set of 14 (cf, Berkun et al., 1962) scaled for their association with stress. The adjectives were presented on a response panel in scrambled order with respect to their location on the stress scale. Actual scale values assigned to the adjectives consisted of those obtained from a successive-intervals analysis of judgments from an undergraduate sample (Neufeld & Davidson, 1972). A single dimension of subjective stress has been shown to account for perceived interadjective differences according to multidimensional scaling results with undergraduates (Neufeld, 1973). The measure has been shown to be relatively sensitive to variation in perceived stressor properties for this population (Davidson & Neufeld, 1974; Neufeld & Davidson, 1974b).

Heart rate. A baseline value for heart rate was taken after the treatments had been administered and consisted of the fastest beat during the 10-sec interval prior to film commencement. The response to the last two accident scenes was taken as the fastest beat during a 30-sec interval, which consisted of the 10 sec during the accident scene, the 10 sec preceding the accident scene, and the 10 sec following the accident scene. For the first accident scene, the 10 sec following the accident scene interval was omitted, since it overlapped with the portion of green leader film and the subject's response to the Subjective Stress Scale for that accident. Skin conductance. A baseline value for skin conductance (in micromhos) was taken (after the treatment manipulations) just preceding film commencement. For the last two accident scenes, the value of this response index consisted of the highest conductance during the 30 sec while the accident scene was being shown. Again, the postaccident 10-sec interval was omitted for the first accident scene. Frontalis muscle tension. Baseline values of frontalis muscle tension were taken as the area under the Grass 7P3 preamplifier integrated output curve (see the Apparatus section below) during the 10-sec interval preceding film commencement. Muscle tension for each accident scene was measured as the maximum of the areas under the integrated curve during the 10-sec intervals corresponding to those used for scoring skin conductance and heart rate. This index was added to the more conventional psysiological measures of stress (heart rate and skin conductance), as several investigators have considered skeletal muscle reactions to be potentially important indices of response to affective stimuli (cf. Martin, 1970). This measure has previously been found to discriminate among stressed and nonstressed normal subjects whether or not heart rate and skin conductance have done so (Neufeld & Davidson, 1974b; Newman, 1962) and is especially sensitive for females (Davidson & Neufeld, 1974; Neufeld, 1972).

Apparatus Physiological recording was done with a Grass Model 7 polygraph. A Grass Model 7P3 preamplifier with an integrator circuit was used to monitor muscle tension while a Grass 7P4 preamplifier and cardiotachometer monitored heart rate, A Grass Model 7P1 preamplifier was used in recording skin conductance. Frontalis muscle tension was recorded with surface electrodes placed midway between the hairline and each eyebrow. Cardiac activity was monitored as pulse rate using a photoelectric plethysmograph placed on the fifth finger of the nonpreferred hand and shielded with a dark cloth. 2 The high-aversiveness-appraisal passage was discarded, despite the scaling results, since ensuing pilot work indicated that its believability in the actual experiment might be in question, as it required a description that sounded "overly affected."

EVIDENCE OF STRESS The skin conductance electrodes were covered with K-Y sterile lubricant and were placed on the first and fourth fingers of the nonpreferred hand.

Poststressor

Questionnaire

The poststressor questionnaire was designed to assess the experimental effects on the subject's evaluation of stimulus aversiveness and, separately, on her confidence about dealing with the stimulus. In addition, the questionnaire assessed the possible effects on cognitive "avoidance" versus "vigilance" during the film presentation. The question directed toward evaluated stimulus aversiveness was as follows: "After hearing the description of the film, to what extent did you feel the film would be disturbing or distressing to the average person?" The subject was asked to provide three separate ratings to this question as it applied to her while she was viewing each of the three accidents. The rating for each accident was made on a 9-point scale anchored by "extremely" (9) and "not at all" (1). Subjects given the environmental-source manipulation were asked, "How did the procedure of clearing your mind and thinking of a blank wall for 4 minutes affect how confident you felt about viewing the film?" Ratings for each accident were obtained on a 9-point scale anchored by "not at all confident" (9) and "highly confident" (1). Similar ratings were obtained from those given the personal-source manipulation, who were asked, '"How did the information about where you scored on the second questionnaire affect how confident you felt about viewing the film?" The questions relating to vigilance versus avoidance were each answered on a 9-point scale and were fashioned after those of Monat et al. (1972). There were three questions for vigilance (e.g., "I tried to attend to the film, noting my own reactions to the scenes") and three questions for avoidance (e.g., "I thought about things not related to the film"). The questionnaire was given after the stressor presentation in order to reduce disruption of the ongoing sequence of events while the film was being presented.

RESULTS Reaction to each accident scene for each physiological measure was taken as the maximum response to the scene (as described earlier) minus the baseline value for that measure. (Preliminary analyses indicated no experimental effects on any of the baseline values). The reactivity score for the Subjective Stress Scale was the scale value for the adjective selected following the scene minus the baseline scale value taken just before film commencement. In addition, the evaluated stimulus aversiveness and the value for confidence in dealing with the stimulus (coping confidence) were included in the battery of measures assessing the stress response of each

637

scene. The value of trait anxiety from the preexperimental Trait-State Anxiety Inventory was taken as a covariate, since it measures individual differences in the tendency to experience any stressor as dangerous or threatening (Spielberger, 1972, p. 39). Evidence of stress was assessed separately at each accident scene by means of a 2 X 2 X 3 (Level of Aversiveness X Source of Defence X Level of Defence) multivariate analysis of covariance.8 This type of multivariate analysis was employed because it helps to summarize experimental effects when multiple dependent variables are used (Woodward & Overall, 1975). Since skin conductance and heart rate tend to be relatively insensitive measures of laboratory stress response