Defensiveness is related to increased startle magnitude

Personality and Individual Differences 37 (2004) 1441–1451 www.elsevier.com/locate/paid Defensiveness is related to increased startle magnitude Steven D. LaRowe a,* , John P. Kline b, Christopher J. Patrick c a c Center for Drug and Alcohol Programs, Medical University of South Carolina, 67 President Street, P.O. Box 250861, Charleston, SC 29425, USA b Department of Psychology, 209 Psychology Building, Florida State University, Tallahassee, FL 32306, USA Department of Psychology, University of Minnesota, Elliot Hall, 75 East River Road, Minneapolis, MN 55455, USA Received 12 January 2003; received in revised form 12 December 2003; accepted 4 February 2004 Available online 19 March 2004 Abstract Previous work has demonstrated that individuals with a repressive coping style (i.e. those high in defensiveness and low in trait anxiety) tend to show higher levels of physiological reactivity than nonrepressors. The present study examined archival data from a two-session experiment to assess whether repressors would show greater reactivity as measured by the eye blink component of the acoustic startle response. Following Weinberger et al.’s (1979) fourfold repression taxonomy, and using the EPQ-L and EPQ-N scales as measures of defensiveness and anxiety, participants were identified as repressors, true low anxious, true high anxious, and defensive high anxious. Results revealed that defensiveness, but not repression per se, was associated with greater startle magnitude. The effect was observed in the first session, but not in the second. Implications for these findings are discussed.
2004 Elsevier Ltd. All rights reserved. Keywords: Repression; Defensiveness; EPQ-L; EPQ-N; Acoustic startle 1. Introduction The current conceptualization of the ‘‘repressive coping style’’ is based on a 2 · 2 taxonomy described by Weinberger, Schwartz, and Davidson (1979). This taxonomy defines repressive coping as a construct comprised of high defensiveness, the tendency to endorse unlikely virtues and deny minor faults, combined with low levels of reported trait anxiety. The construct is typically measured using scales assessing defensiveness and self-reported anxiety. Weinberger et al. * Corresponding author. Tel.: +1-843-792-5717; fax: +1-843-792-5728. E-mail address: larowe@musc.edu (S.D. LaRowe). 0191-8869/$ - see front matter
2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.paid.2004.02.001 1442 S.D. LaRowe et al. / Personality and Individual Differences 37 (2004) 1441–1451 (1979) proposed that social desirability scales (e.g. the Marlowe–Crowne Social Desirability Scale; Crowne & Marlowe, 1964) might tap an orientation toward repressive–defensive coping. Individuals scoring high on defensiveness scales and low on anxiety scales are generally classified as ‘‘repressors’’, whereas those scoring low on defensiveness and low on anxiety are classified as ‘‘true low anxious’’. Those scoring low on defensiveness scales and high on anxiety are classified as ‘‘true high anxious’’, while individuals scoring high on both defensiveness and anxiety scales are known as ‘‘defensive high anxious’’. While ‘‘repressors’’ by definition report low levels of trait anxiety, they often paradoxically display higher levels of physiological reactivity (e.g. Derakshan & Eysenck, 1997, 2001; Weinberger et al., 1979). For instance, a number of studies have demonstrated a relationship between repressive coping and greater cardiovascular activity. In their 1979 study, Weinberger, Schwartz, and Davidson measured heart rate (HR) increases during a phrase association task. They found that repressors showed HR increases that were comparable to those observed in high anxious individuals and greater than those observed in low anxious individuals. A later study by KiecoltGlaser and Greenberg (1983) found greater HR responses in repressors relative to low and high anxious individuals during a role-playing task. In another study, Asendorpf and Scherer (1983) had repressors, true low and high anxious groups, as well as a defensive high anxious group perform a potentially anxiety-arousing free association task. Repressors showed higher HRs relative to the true low anxious group. Newton and Contrada (1992) demonstrated that repressors, relative to high and low anxious groups, showed increased HR during a task in which participants were led to believe they were being watched by a group of observers. The eye blink component of the reflexive startle response, which has served as a useful index in studies of emotion and attention (for reviews see Lang, 1995; Hoffman, 1997), provides another potential index of physiological reactivity. The human startle response involves a ‘‘forward thrusting of the head and a descending flexor wave reaction, extending through the trunk to the knees’’ (Lang, Bradley, Cuthbert, & Patrick, 1993, p. 164) and is readily elicited by an acoustic probe with a rapid rise time and high intensity (e.g. a burst of white noise). Measurement of the startle response is obtained through electromyographic recordings of the activity of the orbicularis oculi muscle below the left eye. Properly filtered EMG recordings of startle blinks produce an output that appears as a ‘‘peak’’. The onset latency of this peak, defined as the time from probe onset to the onset of the response, as well as the magnitude of the peak itself, serve as measures of startle reactivity; shorter onset latencies and larger blink magnitudes are generally indicative of greater startle reactivity. Thus far, there appear to be no published studies that have examined startle reactivity of repressors. However, given that repressive style has been linked to higher levels of physiological reactivity, it is reasonable to hypothesize that repressors should show greater startle responding as well. The present study examined archival data from a study originally designed to examine startle habituation in a slide/startle paradigm (LaRowe, 1997). Although the Marlowe–Crowne Social Desirability Scale (MCSD; Crowne & Marlowe, 1964) is typically used for the assessment of the defensiveness dimension (e.g. Asendorpf & Scherer, 1983; Weinberger et al., 1979), this measure was not available as it was not a part of the original data collection protocol. However, a number of other scales have served as measures of defensiveness in a variety of studies, including the MMPI Lie scale (Frecska et al., 1988), the Repressive–Defensiveness scale from the Weinberger Adjustment Inventory (Brown et al., 1996), and the Eysenck Personality Questionnaire Lie scale S.D. LaRowe et al. / Personality and Individual Differences 37 (2004) 1441–1451 1443 (EPQ-L; Gudjonsson, 1981; Jamner & Schwartz, 1986; Kline, Schwartz, Fitzpatrick, & Hendricks, 1993; Kline, Schwartz, Fitzpatrick, & Hendricks, 1998). The EPQ-L scale, which correlates with the MCSD (e.g. r ¼ 0:52, Lara Cantu, 1990; Plante & Schwartz, 1990, r ¼ 0:54), was collected with the original data and served as the defensiveness measure in the current investigation. Although the Taylor Manifest Anxiety Scale (TMAS; Taylor, 1953) has often been used to assess anxiety in studies of repressive coping (e.g. Asendorpf & Scherer, 1983; Newton & Contrada, 1992), a variety of anxiety measures have been used in repression studies (Derakshan & Eysenck, 1997; Tomarken & Davidson, 1994; Warrenburg et al., 1989). The Eysenck Personality Questionnaire Neuroticism (EPQ-N) scale, which is highly related to the TMAS (r ¼ 0:74, Mathis, LaRowe, & Kline, 2001), was used as the measure of anxiety within the present study. The use of the Eysenck scales (i.e. EPQ-L and EPQ-N scales) to assess defensiveness, anxiety, and repression is not novel. This method of classification was initially used by Gudjonsson (1981), and later referred to as the ‘‘Gudjonsson method’’ by Furnham and Traynar (1999). In addition, it has been used in numerous studies by Kline and colleagues (Kline, Schwartz, Allen, & Dikman, 1998; Kline et al., 1993; Kline, Schwartz, Fitzpatrick, et al., 1998). Hypothesis: It was hypothesized that repressors would show greater levels of baseline startle reactivity (i.e. greater startle magnitude and shorter onset latencies) relative to non-repressors. 2. Method 2.1. Participants Twenty-four male and 24 female college-aged participants were recruited from among undergraduate volunteers who were enrolled in an introductory psychology course at Florida State University. All participants received course credit and monetary compensation for their involvement in the study. 2.2. Questionnaires The Eysenck Personality Questionnaire (EPQ). The EPQ (Eysenck & Eysenck, 1975) consists of 90 items with dichotomous (yes or no) response choices and is comprised of four subscales: Neuroticism (N; 23 items); Extraversion (E; 21 items); Psychoticism/‘‘tough mindedness’’ (P; 25 items); and Lie/‘‘social desirability’’ (L; 21 items). All of the scales have satisfactory reliability coefficients for normal adults, with test-retest reliabilities of 0.86 and 0.84 for the EPQ-N and EPQ-L scales respectively (Eysenck & Eysenck, 1975). Eysenck and Eysenck (1975) have also reported excellent internal consistency for the EPQ-N scale (Cronbach’s alphas of 0.84 for men and 0.85 for women) as well as the EPQ-L scale (Cronbach’s alphas of 0.81 for men and 0.79 for women). 2.3. Stimulus materials A series of 24 slides depicting simple black and white geometric shapes were presented. These neutral slides were used so that basic baseline startle responding could be examined in the absence 1444 S.D. LaRowe et al. / Personality and Individual Differences 37 (2004) 1441–1451 of the modulating effects of affective foreground stimuli. Slides were projected onto a 1 m · 1 m opaque slide screen using a Kodak Ektagraphic III-B slide projector, positioned behind a one-way mirror in an adjoining equipment room. The slide screen was situated 2.5 m in front of the subject. Participants were administered acoustic startle probes consisting of a 50 ms burst of 105 dB white noise with immediate (<10 ms) rise time. Startle probes were presented binaurally to participants through Telephonics TDH-49 stereo headphones. 2.4. Stimulus presentation Presentation and timing of slide and startle stimuli and collection and storage of session data were controlled by a Gateway 386 microcomputer, using the VPM software package (Cook, Atkinson, & Lang, 1987). Two different stimulus presentation orders were used, with one being the reverse of the other. Each consisted of 24 slide presentations, 23 inter-trial intervals (ITIs), and 24 acoustic startle probes. Slides were presented for 6 s each. Startle probes occurred during 12 of the 24 slide presentations at one of three times following slide onset (1.8, 3.5, 4.5 s). Probes were also administered during 12 ITIs of varying lengths (from 10 to 20 s) at one of four onset times (4, 6, 8, or 10 s). ITI lengths varied in order to reduce probe and slide onset predictability. 2.5. Physiological measures Physiological signals were recorded using Coulbourn bioelectric recording modules connected to a Gateway microcomputer that controlled sampling, digitizing, and storage of data. Blink responses to the startle probes were measured from Beckman miniature Ag–AgCl electrodes positioned at the orbicularis oculi muscle beneath the left eye, following the standardized methods recommended by Fridlund and Cacioppo (1986). The raw blink EMG signal was amplified using a Coulbourn S75-01 High Gain Bioamplifier and then rectified and integrated with a Coulbourn S76-01 Contour-Following Integrator with a time constant of 80 ms. Digital sampling began 3 s prior to slide onset at 20 Hz and continued until 50 ms prior to probe onset, at which time the sampling rate increased to 1000 Hz. Sampling at this faster rate continued for 250 ms following probe onset, after which sampling resumed at 20 Hz and continued until 2 s following slide offset. Blink responses were scored off-line for baseline EMG and startle magnitude in arbitrary A/D units using a scoring program developed by Curtin (1996). This scoring program provided startle onset latency data (in milliseconds) as well. 2.6. Procedure Informed consent was obtained at the beginning of the first session of the experiment. Participants then completed a brief information form inquiring about medical problems/medications, and difficulties in hearing and vision. Participants were seated in a reclining chair. The experimenter cleaned the electrode placement sites and placed the electrodes. Experimental instructions were then read aloud. Participants were told that they would be viewing a series of slides and they were instructed to watch each slide the entire time it was on the screen. They were also told they would occasionally hear a brief noise click through headphones that they could simply disregard. S.D. LaRowe et al. / Personality and Individual Differences 37 (2004) 1441–1451 1445 Once the instructions were read, the computer was started and 30 s of baseline data were collected before any stimuli were administered. The stimuli were then presented over the course of approximately 10–15 min. After physiological recording was complete, the EPQ was administered, along with other individual difference measures not reported here. At the end of the first session, participants were scheduled for a second session to be held a week later. The procedures within the second session were identical to those in the first. 2.7. Design and analyses For pre-startle baseline EMG and startle onset latency data, a Defensiveness (2) · Anxiety (2) · Sex (2) · Session (2) mixed-model analysis of variance (ANOVA) was performed, with Defensiveness, Anxiety, and Sex serving as between-subject measures, and Session serving as a withinsubjects variable. For startle magnitude, pre-startle baseline EMG measures for Sessions 1 and 2 were included as covariates (yielding a 2 · 2 · 2 · 2 ANCOVA) in order to rule out whether any observed results were mere artifacts of differences in baseline EMG levels. 3. Results Average startle magnitude data for each subject within each session were calculated. These values were examined across sessions to determine if any participant displayed extremely different values across sessions. A difference score was calculated by subtracting Session 1 magnitude from Session 2 magnitude. The absolute value of the difference was produced and the distribution of the absolute magnitude difference was inspected for outliers. Three participants (all female) showed absolute difference scores that were 2.5 standard deviations from the mean and were therefore excluded. Because the startle data distributions of Sessions 1 and 2 were significantly positively skewed, even after these outliers were omitted, the magnitude scores were square root transformed to normalize the data. Pre-startle baseline EMG data also tended to be skewed and were also square root transformed. The Pearson product moment correlation between square-root transformed average startle magnitude for Sessions 1 and 2 was significant, rð45Þ ¼ 0:71, p < 0:001, as were the between-session correlations for square-root transformed pre-startle baseline rð45Þ ¼ 0:47, p < 0:01 and startle onset latency, rð45Þ ¼ 0:79, p < 0:001, suggesting that these measures were fairly stable across sessions once outliers were removed. The EPQ-L was positively and significantly related to Session 1 EMG baseline, rð45Þ ¼ 0:38, p < 0:05, as well as Session 1 startle magnitude, rð45Þ ¼ 0:37, p < 0:05 (Table 1). An examination of sex differences revealed a significant difference for startle onset latency in Session 1, F ð1; 43Þ ¼ 4:21, p < 0:05, as the mean for the male group (M ¼ 45:59, SD ¼ 4.79, n ¼ 24) was greater than that of the female group (M ¼ 42:50, SD ¼ 5.31, n ¼ 21). A significant difference was also observed for the EPQ-L, F ð1; 43Þ ¼ 6:44, p < 0:05; the mean for the female group (M ¼ 7:57, SD ¼ 3.96, n ¼ 21) was higher than that of the male group (M ¼ 4:88, SD ¼ 3.17, n ¼ 24). This difference did not appear to affect the relationship between the EPQ-L and startle magnitude, as post-hoc analysis revealed that the relationship remained significant after controlling for the effects of Sex, rð42Þ ¼ 0:33, p < 0:05. 1446 S.D. LaRowe et al. / Personality and Individual Differences 37 (2004) 1441–1451 Table 1 Intercorrelations between questionnaire data and startle data Measure 1 2 3 4 5 6 7 8 1. 2. 3. 4. 5. 6. 7. 8. – )0.26 – 0.38* )0.19 – 0.08 )0.06 0.47** – 0.37* )0.17 0.51** 0.34* – 0.13 )0.02 0.19 0.59** 0.71** – )0.22 0.08 )0.42** )0.29 )0.68** )0.57** – )0.12 0.08 )0.30* )0.48** )0.41** )0.61** 0.79** – EPQ-L EPQ-N SQRT BASELINE 1 SQRT BASELINE 2 SQRT MAG 1 SQRT MAG 2 ONSET LATENCY 1 ONSET LATENCY 2 *p < 0:05; **p < 0:01. ANOVA/ANCOVA results. Prior to ANOVA analyses, subjects were classified as high or low defensive and high or low anxious. While the methods of classification vary across studies (e.g. Derakshan & Eysenck, 1997; Newton & Contrada, 1992), Weinberger (1990) has recommended selecting the top third to top quarter of the defensiveness dimension when attempting to identify those high in defensiveness, and this method has been used in numerous repression studies (i.e. Brown et al., 1996; Kline, Schwartz, Allen, et al., 1998; Kline et al., 1993; Kline, Schwartz, Fitzpatrick, et al., 1998; Tomarken & Davidson, 1994). Because males and females had significantly different means on the EPQ-L scale, the top tercile was identified separately for each sex. The cutoff scores for males and females were 7 and 12, respectively, which produced a ‘‘high defensive’’ group containing 7 males and 6 females, and a ‘‘low defensive’’ group containing 17 males and 15 females. High and low anxiety was determined using median splits on the EPQ-N; those who scored 13 or above (n ¼ 21, 10 female) were labeled ‘‘high anxious’’ while the remaining participants were labeled ‘‘low anxious’’. For pre-startle baseline, the Session · Defensiveness interaction was significant, F ð1; 37Þ ¼ 5:12, p < 0:05. The pattern of results was consistent with the pattern of initial correlations, i.e. defensiveness appeared to be associated with higher levels of pre-startle baseline in Session 1, but not Session 2, though post-hoc analysis revealed no significant results. There was a main effect for Sex, F ð1; 37Þ ¼ 4:29, p < 0:05, indicating that women had greater baseline levels on average. However, follow-up analyses revealed that the Sex effect was significant only in the 2 · 2 · 2 · 2 ANOVA, and was not observed in any other post-hoc analysis. The Anxiety · Sex · Session interaction was significant, F ð1; 37Þ ¼ 4:12, p < 0:05, but while the pattern of data suggested that low anxious females showed the highest baseline levels in Session 1, the follow-up analysis did not yield statistically significant results (p < 0:07). Results of the four-way ANCOVA for startle magnitude revealed that Pre-Startle Baseline levels significantly covaried with levels of startle magnitude within Session 1, F ð1; 35Þ ¼ 24:38, p < 0:001, as well as Session 2, F ð1; 35Þ ¼ 20:19, p < 0:001. In addition, a significant Session · Defensiveness interaction was observed, F ð1; 35Þ ¼ 6:89, p < 0:05. Post-hoc univariate tests revealed that within Session 1 only, there was a significant effect for defensiveness on startle magnitude, F ð1; 44Þ ¼ 4:15, p < 0:05; post-hoc analyses revealed that within Session 1, high defensive individuals (M ¼ 28:46, SD ¼ 9.32, n ¼ 13) showed larger startle magnitudes relative to low defensive individuals (M ¼ 20:74, SD ¼ 7.78, n ¼ 32). In contrast, within Session 2, the S.D. LaRowe et al. / Personality and Individual Differences 37 (2004) 1441–1451 1447 Fig. 1. Two way interaction of defensiveness by session (mean and standard error). magnitudes of high defensive individuals (M ¼ 21:06, SD ¼ 7.71, n ¼ 13) were comparable to those of the low defensive individuals (M ¼ 21:76, SD ¼ 9.87, n ¼ 32) (Fig. 1). Repeated measures ANOVAs were performed separately for both high and low defensive groups and indicated a main effect for Session within the high defensive group, F ð1; 12Þ ¼ 46:97, p < 0:001, but not the low defensive group, F ð1; 31Þ ¼ 0:73, p < 0:4. This indicated that high defensive individuals showed a significant drop in startle magnitude from Session 1 to Session 2. No effects were noted for Anxiety or Sex. For startle onset latency, the Defensiveness (2) · Anxiety (2) · Sex (2) · Session (2) repeated measures ANOVA yielded a main effect for Sex, F ð1; 37Þ ¼ 5:11, p < 0:05. No other effects were noted (Fig. 1). 4. Discussion The main findings within the present study showed that defensive individuals, regardless of their levels of self-reported trait anxiety, showed greater baseline levels of pre-startle EMG and greater levels of startle magnitude. Although this finding is in contrast to the prediction that repression (i.e. high defensiveness combined with low anxiety) would be associated with the greatest level of reactivity, it is consistent with a growing number of physiological reactivity studies that report effects for defensiveness rather than repression per se (c.f. Warrenburg et al., 1989; Brody et al., 2000; Jamner & Schwartz, 1986; Kline, Schwartz, Allen, et al., 1998; Kline et al., 1993; Kline, Schwartz, Fitzpatrick, et al., 1998). It is important to account for the fact that the relationship between defensiveness and physiological reactivity was only observed in the first session. It has long been argued (e.g. Gale, 1981; Ney & Gale, 1988) that the physiological experiment creates a socially salient interpersonal context that involves fairly intimate interaction, as experimenters must enter the personal space of their participants in order to place electrodes. This intimacy might have been particularly salient for the defensive individuals, who typically display the highest levels of physiological reactivity 1448 S.D. LaRowe et al. / Personality and Individual Differences 37 (2004) 1441–1451 during social-evaluative interaction (c.f. Asendorpf & Scherer, 1983; Barger, Kircher, & Croyle, 1997; Newton & Contrada, 1992). This fact, combined with the finding that initial visits to the laboratory are typically associated with higher levels of reactivity on a variety of physiological measures (e.g. Davis & Cowles, 1988; Hartmann et al., 1995), including the startle response (Haerich, 1997; Ornitz & Guthrie, 1989), might account for the higher levels of physiological reactivity in the defensive individuals during the first experimental session. Although the effects of novelty and interpersonal context cannot be directly tested using the present data, it underscores the necessity for future investigation of the effects of defensiveness within the context of psychophysiological experiments. There are a number of reasons to suspect that the present results might be an artifact of sex effects. The present study was conducted predominantly by male researchers, and over half (i.e. 14) of the female participants, but only 1 male, worked with experimenters of the opposite sex. It has been suggested (e.g. Gale, 1981; Kline, Blackhart, & Joiner, 2002) that having an opposite-sex experimenter can enhance the demand characteristics of an experiment. Consistent with this, there is evidence to suggest that having an opposite-sex experimenter is related to increased levels of physiological reactivity (Kline et al., 2002; Larkin, Ciano-Federoff, & Hammel, 1998; McCubbin et al., 1991; Turner, Beidel, & Larkin, 1986). Similarly, there is also evidence that scores on the EPQ-L scale increase when there is motivation to present oneself in a favorable light (e.g. Cowles, Darling, & Skanes, 1992; Michaelis & Eysenck, 1971), and some have argued that interacting with someone of the opposite sex can provide such a motive (Leary et al., 1994). Thus, within the present study, it may be that the female participants produced both higher levels of physiological reactivity, as indicated by shorter startle latencies, and higher EPQ-L scores because they interacted with opposite-sex experimenters more frequently than did male participants. If this was indeed the case, then it follows that the observed relationship between defensiveness and startle reactivity may have been an artifact of the effects of sex. While this may seem like a compelling argument, it does not appear to pertain to the relationship between defensiveness and startle magnitude, as the results of the correlational analyses, one-way ANOVAs, and repeated-measures consistently indicated that defensiveness was related to startle magnitude while Sex was not. In addition, the higher EPQ-L scores in females may have occurred because of an apparent tendency for females to score higher on defensiveness measures (e.g. Eysenck & Eysenck, 1975; Martin & Kirkcaldy, 1998; Plaud, Gaither, & Weller, 1998). Still, the potential influence of opposite-sex interaction may impact physiological reactivity apart from the influence of defensiveness, as evidenced by the females’ shorter startle latencies that were apparently unrelated to measures of defensiveness. Although the effects of the potential interplay between defensiveness, sex of participant, and sex of experimenter could not be fully or satisfactorily explored with this limited data set, it nevertheless remains an important topic for further empirical investigation. The present study had a number of shortcomings, the most obvious of which is that it failed to control the effects of experimenter sex. Future work examining startle and defensiveness must take care to balance experimenter sex across participants and across sessions. Another weakness of the present study is that the Marlowe–Crowne scale was not available. This makes it difficult to directly compare the present results to studies that have used the MCSD. This is not to say that the MCSD is a better measure of defensiveness than the EPQ-L scale (or vice versa); in fact, it would have been ideal to include both the EPQ-L scale as well as the MCSD, since measurement theory S.D. LaRowe et al. / Personality and Individual Differences 37 (2004) 1441–1451 1449 (e.g. Cronbach & Meehl, 1955) argues that constructs such as repression and defensiveness must be investigated using a multi-method approach. One final weakness is that the present study does not contain information about whether defensiveness/repression influences the affective modulation of startle. However, the present results highlight that, even when an experimental context is designed to be relatively neutral (i.e. through the use of neutral stimuli), there may be subtle affective influences at work. These influences must be addressed and controlled for before clear inferences can be made about individual differences in repression and defensiveness with respect to affective startle modulation. References Asendorpf, J. D., & Scherer, K. R. (1983). The discrepant repressor: differentiation between low anxiety, high anxiety, and repression of anxiety by autonomic/facial/verbal patterns of behavior. Journal of Personality and Social Psychology, 45, 1334–1346. Barger, S. D., Kircher, J. C., & Croyle, R. T. (1997). The effects of social context and defensiveness on the physiological responsiveness of repressive copers. Journal of Personality and Social Psychology, 73, 1118–1128. Brody, S., Wagner, D., Heinrichs, M., James, A., Hellhammer, D., & Ehlert, U. (2000). Social desirability scores are associated with higher morning cortisol levels in firefighters. Journal of Psychosomatic Research, 49, 227–228. Brown, L. L., Tomarken, A. J., Orth, D. N., Loosen, P. T., Kalin, N. H., & Davidson, R. J. (1996). Individual differences in repressive–defensiveness predict basal salivary cortisol levels. Journal of Personality and Social Psychology, 70, 362–371. Cook, E. W., Atkinson, L., & Lang, K. G. (1987). Stimulus control and data acquisition of IBM PC’s and compatibles. Psychophysiology, 24, 726–727. Cowles, M., Darling, M., & Skanes, A. (1992). Some characteristics of the simulated self. Personality and Individual Differences, 13, 501–510. Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52, 281–302. Crowne, D. P., & Marlowe, D. (1964). The approval motive: Studies in evaluative dependence. New York: Wiley. Curtin, J. (1996). An interactive program for scoring peak analog responses in Windows. Unpublished computer program, Florida State University, Tallahassee. Davis, C., & Cowles, M. (1988). A laboratory study of temperament and arousal: a test of Gale’s hypothesis. Journal of Research in Personality, 22, 101–116. Derakshan, N., & Eysenck, M. W. (1997). Interpretive biases for one’s own behavior and physiology in high-traitanxious individuals and repressors. Journal of Personality and Social Psychology, 73, 816–825. Derakshan, N., & Eysenck, M. W. (2001). Effects of focus of attention on physiological, behavioural, and reported state anxiety in repressors, low-anxious, high-anxious, and defensive high-anxious individuals. Anxiety Stress, and Coping, 14, 285–299. Eysenck, H. J., & Eysenck, S. B. G. (1975). Eysenck Personality Questionnaire. San Diego, CA: EDITS. Frecska, E., Lukacs, H., Arato, M., Mod, L, Alfoldi, A., & Magyer, I. (1988). Dexamethasone suppression test and coping behavior in psychosocial stress. Psychiatry Research, 33, 137–145. Fridlund, A. J., & Cacioppo, J. T. (1986). Guidelines for human electromyographic research. Psychophysiology, 23, 567–589. Furnham, A., & Traynar, J. (1999). Repression and effective coping styles. European Journal of Personality, 13, 465– 492. Gale, A. (1981). EEG studies of extraversion-introversion: what’s the next step?. In D. E. Broadbent & R. Lynn (Eds.), Dimensions of Personality: Papers in Honour of H.J. Eysenck (pp. 181–207). Oxford: Pergamon. Gudjonsson, G. H. (1981). Self-reported emotional disturbance and its relation to electrodermal reactivity, defensiveness and trait anxiety. Personality and Individual Differences, 2, 47–52. Haerich, P. (1997). Long term habituation and sensitization of the human acoustic startle response. Journal of Psychophysiology, 11, 103–114. 1450 S.D. LaRowe et al. / Personality and Individual Differences 37 (2004) 1441–1451 Hartmann, A., Krumrey, K., Dietl, T., Vogl, L., Zhou, Y., Dirlich, G., & Heurser-Link, M. (1995). Long-term habituation of brain evoked potential responses and pituitary-adrenal secretion with repeated (placebo) testing. Psychoneuroendocrinology, 20, 865–877. Hoffman, H. S. (1997). Attentional factors in the elicitation and modification of the startle reaction. In P. J. Lang, R. F. Simons, & M. T. Balaban (Eds.), Attention and orienting: Sensory and motivational processes (pp. 185–204). Mahwah, NJ: Erlbaum. Jamner, L. D., & Schwartz, G. E. (1986). Self-deception predicts self-report and endurance of pain. Psychosomatic Medicine, 48, 211–223. Kiecolt-Glaser, J. K., & Greenberg, B. (1983). On the use of physiological measures in assertion research. Journal of Behavioral Assessment, 5, 97–109. Kline, J. P., Blackhart, G. C., & Joiner, T. E. (2002). Sex, lies, and electrode caps: an interpersonal context for defensiveness and anterior electroencephalographic asymmetry. Personality and Individual Differences, 33, 459–478. Kline, J. P., Schwartz, G. E., Allen, J. J. B., & Dikman, Z. V. (1998). Perceptual and electroencephalographic registration of masked emotional words in defensiveness: an exploratory study. Personality and Individual Differences, 24, 499–512. Kline, J. P., Schwartz, G. E., Fitzpatrick, D. F., & Hendricks, S. E. (1993). Defensiveness, anxiety and the amplitude/ intensity function of auditory-evoked potentials. International Journal of Psychophysiology, 15, 7–14. Kline, J. P., Schwartz, G. E., Fitzpatrick, D. F., & Hendricks, S. E. (1998). Repressive/defensive coping and identification thresholds for pleasant and unpleasant words. Imagination, Cognition, and Personality, 17, 283–291. Lang, P. J. (1995). The emotion probe: studies of attention and motivation. American Psychologist, 50, 372–385. Lang, P. J., Bradley, M. M., Cuthbert, B. N., & Patrick, C. J. (1993). Emotion and psychopathology: Startle probe analysis. In L. Chapman & D. Fowles (Eds.), Progress in experimental personality and psychopathology research (Vol. 16, pp. 163–199). New York: Springer. Lara Cantu, M. A. (1990). Reliability and validity of the Marlowe and Crowne Social Desirability Scale in an adult population. Salud Mental, 13, 35–38. Larkin, K. T., Ciano-Federoff, L. M., & Hammel, D. (1998). Effects of gender of observer and fear of negative evaluation on cardiovascular reactivity to mental stress in college men. International Journal of Psychophysiology, 29, 311–318. LaRowe, S. D. (1997). Startle habituation and overall reactivity: quantification, reliability, and individual differences. Unpublished masters thesis, Florida State University, Tallahassee. Leary, M. R., Nezlek, J. B., Downs, D., Radford-Davenport, J., Martin, J., & McMullen, A. (1994). Self-presentation in everyday interactions: effects of target familiarity and gender composition. Journal of Personality and Social Psychology, 67, 664–673. Martin, T., & Kirkcaldy, B. (1998). Gender differences on the EPQ-R and attitudes to work. Personality and Individual Differences, 24, 1–5. Mathis, K. I., LaRowe, S. D., & Kline, J. P. (2001). Socially desirable responding, self deception, and impression management: defensiveness by any other name. Poster presented at the annual convention of the American Psychological Society, Toronto, Ont., Canada. McCubbin, J. A., Wilson, J. F., Bruehl, S., Brady, M., Clark, K., & Kort, E. (1991). Gender effects on blood pressures obtained during an on-campus screening. Psychosomatic Medicine, 53, 90–100. Michaelis, W., & Eysenck, H. J. (1971). The determination of personality inventory factor patterns and intercorrelations by changes in real-life motivation. Journal of Genetic Psychology, 118, 223–334. Newton, T., & Contrada, R. J. (1992). Repressive coping and verbal-autonomic dissociation: the influence of social context. Journal of Personality and Social Psychology, 62, 159–167. Ney, T., & Gale, A. (1988). A critique of laboratory studies of emotion with particular reference to psychophysiological aspects. In H. L. Wagner (Ed.), Social psychophysiology and emotion: Theory and clinical applications. Chichester, NY: Wiley. Ornitz, E., & Guthrie, D. (1989). Long-term habituation and sensitization of the acoustic startle response in the normal adult human. Psychophysiology, 26, 166–173. Plante, T. G., & Schwartz, G. E. (1990). Defensive and repressive coping styles: self-presentation, leisure activities, and assessment. Journal of Research in Personality, 24, 173–190. S.D. LaRowe et al. / Personality and Individual Differences 37 (2004) 1441–1451 1451 Plaud, J. J., Gaither, G. A., & Weller, L. A. (1998). Gender differences in the sexual rating of words. Journal of Sex and Marital Therapy, 24, 13–19. Taylor, J. A. (1953). A personality scale of manifest anxiety. Journal of Abnormal and Social Psychology, 48, 285–290. Tomarken, A. J., & Davidson, R. J. (1994). Frontal brain activation in repressors and nonrepressors. Journal of Abnormal Psychology, 103, 339–349. Turner, S. M., Beidel, D. C., & Larkin, K. T. (1986). Situational determinants of social anxiety in clinic and nonclinic samples: physiological and cognitive correlates. Journal of Consulting and Clinical Psychology, 54, 523–527. Warrenburg, S., Levine, J., Schwartz, G. E., Fontana, A. F., Kerns, R. D., Delaney, R., & Mattson, R. (1989). Defensive coping and blood pressure reactivity in medical patients. Journal of Behavioral Medicine, 12, 407–424. Weinberger, D. A. (1990). The construct validity of the repressive coping style. In J. L. Singer (Ed.), Repression and dissociation: Implications for personality theory, psychopathology, and health (pp. 37–386). Chicago: University of Chicago Press. Weinberger, D. A., Schwartz, G. E., & Davidson, R. J. (1979). Low-anxious, high-anxious, and repressive coping styles: psychometric patterns and behavioral and physiological responses to stress. Journal of Abnormal Psychology, 88, 369–380.