Impostor fears and perfectionistic concern over mistakes

Impostor fears and perfectionistic concern over mistakes

Personality and Individual Di€erences 29 (2000) 629±647 www.elsevier.com/locate/paid Impostor fears and perfectionistic concern over mistakes Ted Th...

180KB Sizes 17 Downloads 399 Views

Personality and Individual Di€erences 29 (2000) 629±647

www.elsevier.com/locate/paid

Impostor fears and perfectionistic concern over mistakes Ted Thompson*, Peggy Foreman, Frances Martin School of Psychology, University of Tasmania, GPO Box 252-30, Hobart, Tas. 7001, Australia Received 25 February 1999; received in revised form 8 September 1999; accepted 1 October 1999

Abstract Impostors are outwardly successful individuals who experience secret intense feelings of fraudulence in achievement situations. Elements of perfectionism are evident in a tendency on the part of impostors to maintain high standards for personal evaluation while being critical of their inability to realise these standards. This study utilised a 2 (impostor status: high, low)  2 (task type: high vs. low frequency of mistakes) between-subjects factorial design to investigate the connection between impostor fears and perfectionistic concern over mistakes. Sixty undergraduate students completed either a high or low frequency of mistake Stroop Colour-Word task, following which they completed items assessing perceptions of their performance, concern over mistakes, perceptions of control and anxiety, the Positive and Negative A€ect Scale and the Russell Causal Dimension Scale. Links with perfectionistic concern over mistakes and anxiety were strongly supported, with impostors reporting less control, greater anxiety, more negative a€ect and greater concern over mistakes than non-impostors irrespective of experimental condition. The roles of anxiety and perfectionist cognitions in the maintenance of impostor fears are discussed. 7 2000 Elsevier Science Ltd. All rights reserved. Keywords: Impostor fears; Concern over mistakes; Fear of failure; Perfectionism; Self-esteem; Stroop task

1. Introduction The impostor phenomenon refers to an intense feeling of intellectual phoniness experienced by many high-achieving individuals (Clance, 1985). Despite the accumulation of consistent, objective evidence to the contrary, people experiencing impostor fears harbour resilient doubts * Corresponding author. Tel.: +61-3-6226-2887; +61-3-6226-2883. E-mail address: [email protected] (T. Thompson). 0191-8869/00/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 1 9 1 - 8 8 6 9 ( 9 9 ) 0 0 2 1 8 - 4

630

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

of their own abilities, which they believe to be over-estimated by others. As a consequence, they su€er from perpetual fears that they will be found out: that others will discover that they are not truly intelligent, but are in fact, `impostors'. Reporting clinical anecdotes, Clance (1985) indicates that after an achievement-related task is assigned, these individuals are plagued by bad dreams, worry, self-doubts and anxiety, experiences which result in procrastination and immobility in the face of possible failure (e.g. Chrisman, Pieper, Clance, Holland & Glickauf-Hughes, 1995; Clance & Imes, 1978). Impostors react to this anxiety in one of two ways, either by extreme over-preparation, or by initial procrastination followed by frenzied preparation. Following success, there is elation and relief. However, the relief is short-lived. As new achievement situations are encountered, anxieties and self-doubts return and the cycle begins anew. Impostors' extreme over-preparation results in their attribution of success to e€ort (if they over prepared), or to luck (if they procrastinated). The apparent inability of impostors to harness their anxiety as motivation means that impostors are unlikely to derive any satisfaction from challenge, seeing it as threat. Their overstriving, like their immobility and procrastination, is a consequence of extreme anxiety. 1.1. Overgeneralisation of failure, self-criticism A further tendency on the part of impostors is to overgeneralise the negative implications of failure to the whole of their self-concepts (Thompson, Davis & Davidson, 1998). This tendency is all the more pronounced in the case of individuals with low self-esteem and is likely to lead to depression (e.g. Carver & Ganellen, 1983; Kernis, Brochner & Frankel, 1989; Kernis, Grannemann & Mathis, 1991). For example, Carver and Ganellen (1983) found overgeneralisation to be a powerful predictor of depression in both male and female college students. These consequences are endorsed in the case of impostors, with links between impostor fears and low self-esteem noted. Cozzarelli and Major (1990) found impostor scores and trait self-esteem signi®cantly correlated at r…106† ˆ ÿ 0:51, p < 0.001 and Thompson et al. (1998) found impostors report lower academic self-esteem and lower global self-esteem than non-impostors. Chrisman et al. (1995), exploring the relationship between depression and impostor fears, found a high correlation …r ˆ 0:62† between the Clance Impostor Phenomenon Scale (CIPS: Clance, 1985) and the Depressive Experiences Questionnaire (DEQ: Blatt, 1976; cited in Chrisman et al., 1995). Of the subscales comprising the DEQ Ð dependency, ecacy and selfcriticism Ð it is noteworthy that the self-criticism subscale attained the highest correlation …r ˆ 0:71), signifying the pivotal role of self-criticism within impostor beliefs. The tendency on the part of impostors to report negative emotions (anxiety, dissatisfaction, guilt, humiliation), together with their tendency to attribute failure internally to a greater extent than non-impostors (e.g. Thompson et al., 1998) underscores the veracity of clinical observations (e.g. Langford, 1990; Langford & Clance, 1993) and empirical ®ndings (e.g. Frost, Marten, Lahart & Rosenblate, 1990; Kolligian & Sternberg, 1991) suggesting links between impostor fears, anxiety and depression. Taken together, this evidence of over-inclusive, internal attributions following failure points convincingly to a greater susceptibility to depression among impostors (Rizley, 1978). Collectively, these ®ndings have implications in terms of an understanding of the dynamics and treatment of impostor fears.

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

631

1.2. Links with perfectionism Within the clinical literature, perfectionism is a dominant theme, with impostors setting extremely high, often unrealistic standards for self-evaluation (Imes & Clance, 1984). Thompson et al. (1998) found elements of perfectionism evident in the di€erences between impostors and non-impostors in a tendency to discount positive feedback and maintain high standards for personal evaluation while being critical of their inability to realise these standards. These elements are central to the identi®cation of perfectionism and emphasised in much of the literature on this topic (e.g. Burns, 1980; Ferguson & Rodway, 1994; Frost et al., 1990; Hamachek, 1978; Pacht, 1984). As noted above, impostors are likewise di€erentiated from non-impostors in terms of a tendency to overgeneralise the negative e€ects of failure, a further cognitive distortion associated with perfectionism (Barrow & Moore, 1983; Hamachek, 1978; Pacht, 1984). Likewise, Imes and Clance (1984) note a tendency among impostors to overgeneralise any mistake as implying incompetence and Cromwell, Brown, Sanchez-Huceles and Adair (1990) found that impostors were signi®cantly more likely than non-impostors to set unreasonably high expectations and were high in their need for other peoples' approval. In a similar vein, it has been suggested that an overly critical evaluation of one's performance is at the crux of psychopathology associated with perfectionism (Frost et al., 1990). Whereas high achieving behaviour alone is not unhealthy, allowing oneself little latitude for mistakes may well be (Frost et al., 1990). It has also been proposed that the tendency to be overly preoccupied with making mistakes suggests that the motivation driving perfectionists is fear of failure rather than a need for achievement (Hamachek, 1978). On these bases, it seemed to be warranted to explore links between impostor fears and perfectionistic concern over mistakes.

1.3. The present study In a situation designed to generate a high frequency of mistakes, it was expected that impostors would demonstrate impaired performance relative to non-impostors. On the other hand, in a situation designed to generate a low frequency of mistakes, no performance di€erences were expected. Di€erences were also expected in terms of perceptions of performance in a number of ways, with (for example) impostors estimating having made a greater number of mistakes, reporting a greater concern over mistakes and stating a higher criterion for what would represent `a very good performance' than non-impostors. Our predictions were based on several considerations. One was a greater tendency on the part of impostors to experience greater anxiety during a time-limited performance situation which evokes evaluative threat relative to non-impostors. This anxiety was expected to impair performance in a situation designed to give rise to a high frequency of mistakes. A second consideration was that perfectionistic concern over mistakes associated with impostor fears was expected to add to an already high level of anxiety in situations which presage a high likelihood of mistakes. The Stroop Colour-Word Test (Stroop, 1938) was chosen in order to create a situation characterised by either a high or low frequency of mistakes. Choice of the Stroop task was

632

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

predicated on the fact that anxiety is known to be a factor a€ecting reaction time for this task (e.g. Jorgenson, 1977; Pallak, Pittman, Thane, Heller & Munson, 1975). Findings from a study by Frost et al. (1995) gave us further reason to consider the Stroop task to be particularly well suited to our purposes. These researchers investigated a€ective and cognitive reactions to making mistakes for participants high and low in perfectionist concern over mistakes. They found that following a task characterised by a high frequency of mistakes, individuals high in perfectionistic concern over mistakes reacted more negatively than individuals low in perfectionistic concern over mistakes. On this basis, we included a measure of concern over mistakes in order to determine the relative status of impostor scores on key dependent measures while controlling for concern over mistakes. Di€erences between impostors and non-impostors were also expected in terms of anxiety and perceptions of control, with impostors likely to experience greater anxiety in both high and low frequency of mistakes conditions and report less control. As these di€erences were expected to be particularly apparent in the high frequency of mistakes condition, we expected signi®cant interactions involving impostor status and frequency of mistakes. On the same basis, impostors were expected to report greater negative a€ect in the high frequency of mistakes condition relative to non-impostors. Finally, on the assumption that performance di€erences would be evident between non-impostors and impostors in the high frequency of mistakes condition, we expected impostors to attribute their performance more to internal, stable and uncontrollable factors than non-impostors.

2. Method 2.1. Experimental design Two groups of students: impostors and non-impostors were randomly assigned to either a high or low frequency of mistakes task. This rendered the experiment a 2 (status: impostor, non-impostor)  2 (frequency of mistakes: high, low) between subjects factorial design, the major dependent measures being items assessing perceptions of control and anxiety, positive and negative a€ect, concern over mistakes, performance and perceptions of performance. 2.2. Participants Participants were 60 undergraduate psychology students …N ˆ 49 females; N ˆ 11 males† with an age range of 17 to 47 yr (M=21 yr, S.D.=7 yr). Participation in the study was voluntary, with ®rst year students receiving course credit for participation. 2.3. Pre-experimental measures 2.3.1. Clance impostor phenomenon scale The impostor scale used in this study was a modi®ed version of the Clance Impostor Phenomenon scale (CIPS; Clance, 1985). In accordance with King and Cooley's (1995)

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

633

recommendation, several items on Clance's Impostor Scale were altered in order to tailor the scale to academic achievement situations. The CIPS is a 20-item self-report instrument designed to assess the extent to which individuals experience impostor fears. Respondents indicate their endorsement of items on a 5-point Likert scale with end-point designations ranging from (1) ``not at all true'' to (5) ``very true''. Items include statements such as ``I can give the impression that I'm more competent than I really am'' and ``Sometimes I'm afraid others will discover how much knowledge or ability I really lack''. The CIPS has a high level of internal consistency with reported alpha values ranging from 0.84 (Prince, 1989) to 0.96 (Holmes, Kertay, Adamson, Holland & Clance, 1993). 2.3.2. Concern over mistakes The Concern over Mistakes subscale (CM) of the Multidimensional Perfectionism Scale (MPS; Frost et al., 1990) is a 9-item measure which re¯ects negative reactions to mistakes, a tendency to interpret any mistake as equivalent to failure and a tendency to believe that one will lose the respect of others following failure. Items include statements such as ``People will probably think less of me if I make a mistake'' and ``If I do not do well all the time, people will not respect me.'' Respondents indicate the degree of their endorsement of each item on a 5-point Likert scale with end-point designations ranging from (1) ``strongly disagree'' to (5) ``strongly agree''. The CM subscale contributes 25% of the variance to the MPS and is considered the central component of perfectionism (Frost et al., 1990). 2.3.3. Fear of negative evaluation The Fear of Negative Evaluation Scale (FNE) was originally developed as a 30-item measure by Watson and Friend (1969). It was designed to assess the extent to which individuals experience apprehension at being evaluated by others. The brief form of the FNE developed by Leary (1983) was used in this study and comprises 12 of the original 30 items. Respondents are required to indicate how characteristic each statement is of them on a 5-point Likert scale with end-point designations ranging from (1) ``not at all'', to (5) ``extremely characteristic of me''. Sample items are ``Sometimes I think I am too concerned with what other people think of me'' and ``I am afraid that others will not approve of me''. Internal consistency for the brief form is high: r ˆ 0:90 (Leary, 1983). 2.4. Experimental measures 2.4.1. Global self-esteem The global self-esteem subscale of the Marsh (1990) Self-Descriptive Questionnaire III consists of 12 items. Sample items include ``Overall, I am not very accepting of myself'' and ``Overall I have pretty positive feelings about myself''. Respondents indicate their agreement with each item on an 8-point scale with end-point designations ``de®nitely false'' and ``de®nitely true''. Marsh (1990) reports a coecient alpha of 0.88 and test-retest reliability of 0.85. Possible scores range from 12 to 72. 2.4.2. The state-trait anxiety inventory Odd- and even-numbered items from the State-Trait Anxiety Inventory (STAI: Spielberger,

634

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

Gorsuch, Lushene, Vagg & Jacobs, 1983) were used to create parallel forms as a means of assessing participants' anxiety immediately before and after completing the Stroop task. Each split half form contains 10 items. Respondents are required to indicate how true each statement is for them at the present moment on a 4-point scale with end-point designations ``not at all'' (1) and ``very much so'' (4). Sample items are ``I feel calm'' and ``I am worried.'' Spielberger et al. (1983) report a median internal consistency (KR-20) of 0.93 for the full scale form of the STAI and licence parallel forms based on odd- and even-numbered items. This option was followed in order to minimise any tendency on the part of participants to respond to items on the second occasion of administration in a manner which was consistent with their responses on the ®rst occasion. This was a genuine concern as the two administrations were completed within a short period of time (10 min or so). Based on a prior study by the ®rst author (Thompson & le Fevre, 1999), the KR-20 for the ®rst administration of the STAI was 0.89, while that for the second administration of the STAI was 0.88. Possible scores for each split half of the STAI range from 10 to 40. 2.4.3. Causal dimensions scale The Russell (1982) Causal Dimensions scale comprises nine items, with three items addressing each of the three attributional dimensions: internality, stability and controllability. End-point designations for internality were ``re¯ects an aspect of the situation'' (1) and ``re¯ects an aspect of yourself'' (9). For stability, end-point designations were ``temporary'' (1) and ``permanent'' (9), while for controllability, end-point designations were ``uncontrollable by you or other people'' (1) and ``controllable by you or other people'' (9). Coecient alphas for internality, stability and controllability are 0.87, 0.84 and 0.73, respectively. In this study, respondents were asked to state a cause for their performance on the Stroop task and then assess the internality, stability and controllability of that cause. Possible scores range from three to 27 for each subscale. 2.4.4. Positive and negative a€ect schedule (PANAS) The PANAS Scales (Watson, Clark & Tellegen, 1988) were developed as a brief measure of positive and negative a€ect. The PANAS is comprised of 20 items. Respondents indicate the extent to which they experience these feelings ``right now at the present time'' on a scale with end-point designations of (1) ``very slightly'' or ``not at all'' to (5) ``extremely''. Adjective descriptors include ``interested'', ``excited'', ``afraid'', ``ashamed''. The internal consistency of the PANAS ranges from 0.86 to 0.90 for positive a€ect and from 0.84 to 0.87 for negative a€ect. Possible scores range from 20 to 100. 2.4.5. Attitudes towards mistakes Four items modelled on those used by Frost et al. (1995) were used to assess participants' attitudes to making mistakes. Participants were ®rst asked to estimate the number of mistakes they had made on the Stroop task, following which they were asked to indicate the number of mistakes they thought would represent ``a very good performance''. In each case, participants elected one of seven response options, as follows: (0±5), (6±11), (12±17), (18±23), (24±29), (36± 40), with these options designed to capture the minimum and maximum mistakes possible on the Stroop task.

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

635

Participants were then asked to estimate the extent to which the number of mistakes they had made was ``less or more than average compared with other people'' on a 7-point scale with endpoint designations ``much more than average'' (1) and ``much less than average'' (7). Finally, participants were asked to describe their ``level of concern about making a mistake'' on a 7-point scale with end-point designations ``not at all concerned'' (1) to ``very concerned'' (7). 2.4.6. Perceptions of performance To assess perceptions of their performance participants were asked to rate their satisfaction with their performance on a 7-point scale with end-point designations ``very satis®ed'' (1) and ``very dissatis®ed'' (7). Participants also reported the extent to which they felt their performance had been a success, with end-point designations ``a total failure'' (1) and ``a total success'' (7) and indicated how con®dent they felt about their ability. For the latter item, end-point designations were ``not at all con®dent'' (1) and ``very con®dent'' (7). Finally, participants were asked to indicate how motivated they were to perform with end-point designations ``not at all motivated'' (1) and ``very motivated'' (7). 2.4.7. Single-item measures of control and anxiety Individual items assessed perceptions of control and anxiety following high and low frequency of mistakes tasks for impostors and nonimpostors. On the one hand, participants were asked to rate the extent to which they felt they were in control of the outcome of their performance on a 7-point scale with end-point designations ``not at all in control'' (1) through to ``very much in control'' (7). On the other hand, participants were asked to rate how anxious they felt while performing on the Stroop task on a 7-point scale with end-point designations ``not at all anxious'' (1) through to ``very anxious'' (7). 2.4.8. Stroop task A computerised version of the Stroop Colour Naming Task (Stroop, 1938) was used. The Stroop procedure involves naming the ink colours in which words or groups of letters are written. For the present study, participants were required to respond to either: (1) the colour of the ink that the word was written in, or (2) the text of the word. Four stimulus words and colours were used: red, blue, green and yellow. Each participant was given a practice session of 12 stimulus presentations. The experimental session consisted of 40 presentations in four blocks of 10 presentations with ten second breaks between blocks. Each presentation consisted of an initial 500ms duration cue statement appearing on the screen which indicated the criterion on which the participant was to respond (colour of the ink or word text). The stimulus word then appeared for 3000 ms during which time the participant was required to respond by pressing a numeric key corresponding to the correct colour or word indicated by a panel of four numbered colours on the screen. Response errors were made more salient by an ``uh-oh'' sound generated by the computer. Reaction time to correct responses and the number of errors were recorded by the computer program for the Stroop task. 2.5. Procedure Prior to the experiment 318 undergraduate students were screened using the Clance Impostor

636

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

Phenomenon Scale (CIPS: Clance, 1985), the Fear of Negative Evaluation Scale (FNE: Leary, 1983) and the Concern over Mistakes subscale (CM) of the Multidimensional Perfectionism Scale (MPS: Frost et al., 1990). In accordance with recommendations by Holmes et al. (1993) respondents obtaining scores of 62 and above on the CIPS were classi®ed as impostors. From the screened sample, a total of 60 students were selected for experimental participation. Of these, 30 students were classi®ed as impostors (M=74, S.D.=6.0) and 30 as non-impostors (M=46, S.D.=6.0). Score ranges for the impostor group were from 63 to 90, while the score range for the nonimpostor group was from 29 to 57. Once experimental participants were identi®ed on the basis of the above criteria, individuals were randomly assigned to experimental conditions. Half of the participants in each of the impostor and non-impostor groups …N ˆ 15† were exposed to the high frequency of mistakes condition and half to the low frequency of mistakes condition. Participants were tested individually. On arrival at the laboratory, students signed a consent form and completed the global self-esteem subscale of the Self-Descriptive Questionnaire III (SDQ III). They then completed the ®rst of two parallel forms of the STAI (Spielberger et al., 1983). All participants were informed that the intention of the experiment was to investigate peoples' thoughts and feelings after working on a task. Participants were then seated in front of a computer monitor and told they were required to perform a timed computer task. Instructions were given by the experimenter and displayed on the computer screen. Participants in the high frequency of mistakes condition were instructed to respond to either the name of a colour word displayed on the computer screen or the ink colour of that word. This was cued by one of two prompts: ``colour'' or ``word'' presented on the computer screen for 500 ms for participants in the high frequency of mistakes condition. Participants in the low frequency of mistakes condition were instructed to respond to the colour of a patch displayed on the screen. After a short practice trial consisting of a block of 12 trials, participants indicated their readiness to begin the computerised Stroop Colour-Word Test by clicking the mouse on a designated area of the computer screen and then completed the experiment. The stimuli (either colour patches or words) were presented with an interstimulus interval (ISI) of 200 ms. Immediately following the task, participants completed the second form of the STAI, items assessing perceptions of their performance, concern over mistakes, perceptions of control and anxiety, the PANAS and the Russell Causal Dimensions Scale. Participants were then fully debriefed by the experimenter and thanked for their participation.

3. Results 3.1. Relationships between personality measures Based on data from the N ˆ 318 individuals screened in order to identify those with either high or low scores on the modi®ed version of the Clance Impostor Scale (CIPS) revealed signi®cant positive correlations between respondents' scores on this scale and scores on the FNE scale r…317† ˆ 0:633, p < 0.0001 and the CM subscale of the MPS: r…317† ˆ 0:597, p <

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

637

0.0001. The FNE and CM scales were also signi®cantly correlated at r…317† ˆ 0:546, p < 0.0001. A one-way ANOVA for impostor status for global self-esteem scores (administered to experimental participants only) revealed that impostors reported signi®cantly lower global selfesteem (M=60.3, S.D.=16.23) than non-impostors (M=75.67, S.D.=11.30), F…1, 59† ˆ 18:19, p < 0.001. There was also a signi®cant di€erence between the impostor (M=25.067, S.D.=8.403) and non-impostor groups (M=15.950, S.D.=5.578) in CM scores: F…1, 58† ˆ 24:51, p < 0.0001 and fear of negative evaluation from others: F…1, 58† ˆ 16:703, p < 0.0001. 3.2. Analysis strategy The strategy followed in analysing the experimental data was to submit each dependent variable to separate 2 (impostor status: impostor, non-impostor)  2 (task type: high vs. low frequency of mistake) ANOVAs. As the numbers of males in cells was low (ranging from two to four), analyses involving gender were not feasible. In cases where statistically signi®cant results for the dependent measure in question could have been reasonably expected to be an outcome of perfectionistic concern over mistakes, analyses of covariance were undertaken with CM scores entered as a covariate. In all analyses, the alpha level was set at 0.05. 3.3. Perceptions of control and anxiety Separate 2 (impostor status: impostor, non-impostor)  2 (task type: high vs. low frequency of mistakes) ANOVAs were performed for the single items assessing perceptions of control and anxiety. It was anticipated that the high frequency of mistakes condition would elicit greater anxiety and give rise to feelings of diminished control relative to the low mistake condition. As these e€ects could be reasonably expected to be greater for impostors relative to nonimpostors, it was expected that signi®cant interactions may arise for impostor status and task type. Failing this, signi®cant main e€ects were expected for task type. While for perceptions of control a signi®cant interaction between impostor status and frequency of mistakes failed to materialise, there was, as expected, a signi®cant main e€ect for task type: F…1, 56† ˆ 18:27, p < 0.001, indicating that all participants felt less in control following the high mistake task. There was also a signi®cant main e€ect for impostor status: F…1, 56† ˆ 8:68, p ˆ 0:005 indicating that impostors felt less in control of their performance than non-impostors (see Tables 1 and 2 for descriptive statistics for impostor status and task type, respectively). For anxiety, a 2 (impostor status: impostor, non-impostor)  2 (task type: high vs. low frequency of mistakes) ANOVA revealed the predicted interaction: F…1, 56† ˆ 4:22, p < 0.05 (see Fig. 1). This arose by virtue of the fact that impostors felt signi®cantly more anxious in the high frequency of mistake condition than in the low frequency of mistake condition: F…1, 29† ˆ 12:37, p ˆ 0:002: There was also a trend towards a main e€ect for impostor status, with impostors reporting greater anxiety across both high and low frequency of mistake conditions than non-impostors: F…1, 28† ˆ 3:75, p ˆ 0:063:

638

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

3.4. Performance measures Separate 2 (status: impostor, non-impostor)  2 (task type: high frequency of mistakes, low frequency of mistakes) ANOVAs were performed on the two dependent variables, reaction time and number of errors. Neither signi®cant interactions nor main e€ects involving impostor status were evident. While impostors made more mistakes and were slower in their reaction time than non-impostors in both high and low frequency of mistake conditions, the di€erences

Table 1 Means and standard deviations for trait personality measures, perceptions of control and anxiety, performance measures, perceptions of performance, a€ective and attributional reactions for impostors and non-impostors Impostor M Personality measures Clance Impostor Scale Fear of Negative Evaluation Scale Concern over Mistakes Scale Global Self-esteem Perceptions of control and anxiety Control Anxiety Performance measures Reaction time(ms) Number of errors Concern over mistakes Concern over mistakes Estimation of own mistakes Mistakes compared with others Estimation of a `very good' performance Perceptions of performance Satisfaction with performance Performance success Con®dence in performance Motivation PANAS Positive mood Negative mood Attributions Internality Stability Controllability STAI STAI-1 STAI-2

Non-impostor SD

M

SD

78.833 33.867 25.067 60.3

6.566 5.894 8.403 16.23

48.750 27.467 15.950 75.67

6.058 6.231 5.578 18.17

4.733 4.867

1.837 1.776

5.767 3.767

1.837 1.431

1084.339 4.433

241.968 4.606

1039.412 3.467

207.723 3.401

3.933 1.900 4.133 1.133

2.164 1.094 1.408 0.346

2.500 1.533 4.567 1.067

1.408 0.730 1.331 0.254

3.933 4.267 4.300 5.700

1.946 1.680 1.643 1.418

2.600 5.367 5.133 5.600

1.499 1.273 1.224 1.192

27.933 16.200

8.196 6.815

32.733 11.533

6.626 2.097

18.800 11.233 17.133

5.378 5.992 5.329

19.333 11.900 17.733

6.008 6.707 5.854

20.600 21.633

5.969 7.107

16.167 14.767

3.204 3.481

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

639

were nonsigni®cant. Table 3 shows F values for impostor status, frequency of mistakes and the interaction term. Nevertheless, there were signi®cant main e€ects for task type (high, low frequency of mistakes) for each of the performance measures. Participants in the high frequency of mistake condition demonstrated higher response latencies: F…1, 56† ˆ 25:7, p < 0.001 and made more errors: F…1, 56† ˆ 31:6, p < 0.001, than participants in the low frequency of mistake condition, o€ering further con®rmation of the e€ectiveness of the experimental manipulation. 3.5. Concern over making mistakes Separate 2 (impostor status: impostor, non-impostor)  2 (task: high mistake, low mistake) ANOVAs were performed on items addressing participants' concern over making mistakes,

Table 2 Means and standard deviations for task type (high, low frequency of mistakes) for perceptions of control and anxiety, performance, perceptions of performance, concern over mistakes, PANAS and attributions Low mistake M Perceptions of control and anxiety Control Anxiety Performance measures Reaction time (ms) Number of errors Concern over mistakes Concern over mistakes Estimation of own mistakes Mistakes compared with others Estimation `very good' performance Perceptions of performance Satisfaction with performance Performance success Con®dence in performance Motivation PANAS Positive mood Negative mood Attributions Internality Stability Controllability STAI STAI-1 STAI-2

High mistake S.D.

M

S.D.

4.500 4.500

1.717 1.776

6.000 4.133

1.114 1.613

1186.091 6.333

182.769 3.111

937.660 1.567

193.426 3.441

3.333 2.233 3.900 1.167

1.971 0.971 1.213 0.379

3.100 1.200 4.800 1.033

1.954 0.551 1.400 0.183

4.067 4.100 4.567 5.567

1.660 1.296 1.406 1.251

2.467 5.333 4.867 5.733

1.697 1.525 1.592 1.363

30.033 14.233

7.156 6.510

30.633 20.467

8.463 5.444

17.667 10.633 16.700

5.610 6.820 6.047

20.467 12.500 18.167

5.444 5.728 5.018

17.600 20.433

5.222 6.497

19.167 15.967

5.246 5.893

640

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

Fig. 1. Single-item reports of anxiety for impostors and nonimpostors following high and low frequency of mistakes tasks.

estimations of their own mistakes and the number of mistakes relative to other people. For participants' ratings of their level of concern over making mistakes, there was a signi®cant main e€ect for impostor status: F…1, 56† ˆ 8:96, p < 0.01, with impostors signi®cantly more concerned about making mistakes than non-impostors. There was also a trend for impostors to estimate that they had made a higher number of mistakes than non-impostors: F…1, 56† ˆ 3:32, p ˆ 0:074: Finally, for the item which addressed participants' estimations of whether the number of mistakes they made was ``less or more than average compared with other people'', there was a signi®cant main e€ect for task type: F…1, 58† ˆ 7:04, p ˆ 0:01, indicating that participants estimated a higher number of mistakes compared with other people in the high frequency of mistakes task than in the low frequency of mistakes task. However, there was neither a signi®cant interaction nor main e€ects for estimations of the number of mistakes that would represent a ``very good performance''. 3.6. Covariate analyses involving concern over mistakes Analyses of co-variance (ANCOVAs)1 were subsequently performed using CM scores as a covariate for items assessing concern over making mistakes and estimations of mistakes. Selection of these dependent measures was premised on the likely mediating roles of CM scores based on essentially similar assessments by Frost et al. (1995). ANCOVAs using CM scores as 1 In each case in which ANCOVAs were used (a) the F-test on the covariate was signi®cant and (b) the assumption of homogeneity of slopes was met.

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

641

Table 3 Results of two-factor analyses of variance for perceptions of control and anxiety, performance, perceptions of performance, concern over mistakes, PANAS items and attributions by impostors and non-impostors in high and low frequency of mistakes conditionsa Interaction term Perceptions of control and anxiety Control Anxiety Performance measures Reaction time (ms) Number of errors Concern over mistakes Concern over mistakes Estimation of own mistakes Mistakes compared with others Estimation of a `very good' performance Perceptions of performance Satisfaction with performance Performance success Con®dence in performance Motivation PANAS Positive mood Negative mood Attributions Internality Stability Controllability STAI STAI-1 STAI-2 a

p < 0.05;



p < 0.01;



Impostor status

Frequency of mistakes

1.091 4.215

8.669 7.344

18.267 0.816

0.213 0.683

0.841 1.302

25.719 31.699

0.005 0.247 0.010 3.111

8.963 3.322 1.632 0.778

0.238 26.380 7.039 3.111

0.441 0.239 2.333 0.236

11.013 10.414 5.046 0.850

15.858 17.682 0.654 0.36

4.044 1.638

6.452 12.842

0.008 0.002 1.948 0.001 2.639

0.135 0.163 0.175 12.758 27.338

0.101 0.317 3.726 1.276 1.047 1.593 11.567

p < 0.001.

a covariate were also performed for all other single-item measures where main e€ects for impostor status were evident. These were perceptions of control, anxiety, satisfaction with performance, estimations of performance success, con®dence in performance, positive mood, negative mood, STAI-1 and STAI-2. The assumption guiding these analyses was that di€erences between impostors and nonimpostors on these items may be due to di€erences in concern over mistakes rather than impostor scores per se. For concern over mistakes, the e€ect of the covariate was nonsigni®cant at F…2, 52† ˆ 0:031, p > 0.8, while the residual di€erence due to impostor status was marginally signi®cant at: F…1, 52† ˆ 3:54, p < 0.066. This pattern was repeated for the second administration of the STAI, where the e€ect of the covariate was again nonsigni®cant at F…2, 52† ˆ 0:473, p > 0.4, the residual di€erence due to impostor status being marginally signi®cant at F…1, 52† ˆ 3:59, p < 0.064. For negative mood, the e€ect of the covariate was nonsigni®cant at F…2, 52† ˆ 0:030, p >

642

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

0.8, while the residual di€erence due to impostor status was signi®cant at F…1, 52† ˆ 4:17, p < 0.046. On these bases, the residual marginally signi®cant di€erence due to impostor status for these dependent measures cannot be attributed solely to scores on the CM subscale of the MPS. For the remaining dependent measures: perceptions of control, anxiety, satisfaction with performance, estimations of performance success, con®dence in performance, positive mood and STAI-1, neither the e€ect of the covariate nor the residual di€erence due to impostor status was signi®cant. As such, impostor status adds little to di€erences evident for these dependent measures once scores on the CM subscale of the MPS are taken into account. 3.7. Perceptions of performance Separate 2 (impostor status: impostor, non-impostor)  2 (task type: high mistake, low mistake) ANOVAs were conducted on ratings for satisfaction, estimations of success, con®dence and motivation. There were signi®cant main e€ects for impostor status for satisfaction: F…1, 56† ˆ 11:01, p < 0.01, con®dence: F…1, 56† ˆ 5:05, p ˆ 0:029 and estimations of success: F…1, 56† ˆ 10:414, p ˆ 0:002, with impostors reporting more dissatisfaction with their overall performance, less con®dence in their ability to perform well on the task and lower estimations of success than non-impostors. There was also a signi®cant main e€ect for task type for estimations of success: F…1, 56† ˆ 17:682, p < 0.001 and satisfaction with performance: F…1, 56† ˆ 15:858, p < 0.001, indicating (not surprisingly) that all participants felt less satis®ed with their performance following the high mistake task and made lower estimations of success on that task. There were no signi®cant interactions or main e€ects for motivation. (See Tables 1 and 2). 3.8. PANAS Separate 2 (impostor status: impostor, non-impostor)  2 (task type: high vs. low frequency of mistakes) analyses of variance for positive and negative mood ratings revealed a signi®cant interaction for positive mood: F…1, 56† ˆ 4:04, p < 0.05. This arose by virtue of the fact that impostors reported signi®cantly less positive mood than non-impostors following the high frequency of mistakes condition: F…1, 28† ˆ 16:67, p < 0.001 (see Fig. 2). For negative mood ratings, a main e€ect for impostor status was evident, with impostors reporting signi®cantly more negative mood than non-impostors F…1, 56† ˆ 12:84, p < 0.001. 3.9. Attributional dimensions Separate 2 (impostor status: impostor, non-impostor)  2 (task type: high vs. low frequency of mistakes) ANOVAs were conducted on subscale scores for internality, controllability and stability. For the internality dimension there was a trend towards a main e€ect for task type: F…1, 59† ˆ 3:72, p ˆ 0:059, with participants more likely to attribute their performance to external factors in the high frequency of mistake condition. Neither signi®cant interactions nor main e€ects were evident in relation to the stability and controllability dimensions (see Tables 1 and 2).

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

643

Fig. 2. Positive mood ratings on the PANAS for impostors and nonimpostors following high and low frequency of mistakes tasks.

3.10. State anxiety Table 4 shows means and standard deviations for parallel forms of the STAI for impostors and non-impostors. As the STAI-1 was completed prior to the Stroop task, scores on this measure are not di€erentiated on the basis of experimental condition. Accordingly, a univariate F-test indicated that impostors were signi®cantly more anxious than non-impostors prior to the Stroop task: F…1, 58† ˆ 12:848, p < 0.0007. For STAI-2 scores, a 2 (impostor status: high, low)  2 (task type: high, low frequency of mistakes) ANOVA was performed. Despite that fact that the interaction term failed to reach signi®cance: F…1, 56† ˆ 2:639, p ˆ 0:1099, a signi®cant main e€ect was revealed for impostor status, with impostors reporting higher levels of anxiety than non-impostors: F…1, 59† ˆ 27:34, p < 0.0001. There was also a signi®cant main e€ect for task type: F…1, 56† ˆ 11:567, p < Table 4 Means and standard deviations for anxiety ratings on the STAI-1 and STAI-2 by impostors and non-impostors following the high and low frequency of mistakes taska

Impostors Non-impostors a

STAI-1 (N = 30)

STAI-2 (Low mistake) (N = 15)

STAI-2 (High mistake) (N = 15)

M

S.D.

M

S.D.

M

S.D.

20.60 16.17

5.97 3.20

18.33 13.60

7.15 2.97

24.93 5.93

5.49 3.65

STAI-1 scores were not di€erentiated on the basis of experimental condition.

644

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

0.0012, with both impostors and non-impostors experiencing greater anxiety in the high frequency of mistake condition.

4. Discussion The results of this study provide support for the expectation that relative to non-impostors, impostors would report a greater concern over mistakes and tend to overestimate the number of mistakes made …p ˆ 0:074). They also report less satisfaction with performance, rate the success of their performance as lower, express less con®dence in their performance, report greater negative a€ect and lower perceptions of control. In addition, they reported greater anxiety both prior to and following the Stroop task on the STAI-1 and STAI-2 measures, respectively. Discrepant results for the single-item measure of anxiety and that for the STAI-2 (the interaction term being signi®cant for the single-item measure of anxiety but not for the STAI2) were evident. While the STAI-2 assessed state anxiety following the Stroop task, the singleitem measure asked for a retrospective report of anxiety experienced during task completion, the actual wording being ``How anxious did you feel while performing on the Stroop task?'' The assessments then, were somewhat di€erent, with the single item measure assessing anxiety experienced during task performance, with the STAI-2 assessing end-state anxiety. On the basis of the single item measure of self-reported anxiety, there was also a trend …p ˆ 0:063† on the part of impostors to report greater anxiety relative to non-impostors. It is noteworthy that these results occurred (1) irrespective of whether the task involved a high or low frequency of mistakes and (2) despite the absence of performance di€erences between impostors and nonimpostors in either the high or low frequency of mistakes condition. These ®ndings underscore the fact that the di€erences noted between impostors and non-impostors on all of the fore-mentioned bases are perceptual di€erences, without any basis in performance or type of task. Nevertheless, it needs to be stressed that once concern over mistakes scores on the MPS are taken into account, impostor status adds little to di€erences on key dependent measures noted earlier. The absence of di€erences between impostors and non-impostors in performance on the basis of either reaction time or errors is consistent with ®ndings by Frost and colleagues (Frost et al., 1995; Frost et al., 1997) for participants high or low in concern over mistakes. In the former of these studies, it is noteworthy that Frost et al. (1995) failed to ®nd signi®cant di€erences based on judgements of what constituted an average number of mistakes, what constituted a good performance and participants' estimates of their own number of mistakes. In the Frost et al. (1997) study, high CM participants failed to report making more mistakes than low CM participants, evidence bolstered by the fact that judges in assessing daily records of mistakes failed to ®nd evidence of reported di€erences in ratings of mistakes on four dimensions: importance, wrongness, harmfulness to the participant and harmfulness to others. On these bases, the di€erences for high- versus low-CM participants from the studies by Frost and associates lay not in the number or kind of mistakes made, but in perceptual di€erences between individuals high or low in CM. The ®nding that impostors report greater dissatisfaction with their performance and view it

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

645

as less successful than non-impostors is consistent with clinical observations that any less-thanperfect performance is disappointing (Clance, Dingman, Reviere & Stober, 1995; Imes & Clance, 1984) and with previous ®ndings which link impostor fears with a propensity to hold high standards for self-evaluation while being critical of their failure to attain these standards (e.g. Hamachek, 1978; Frost et al., 1990; Pacht, 1984; Thompson et al., 1998). Frost et al. (1990) observe that perfectionists, regardless of actual performance, tend to feel that they should have done better. These personal imperatives are consistent with impostor thoughts and feelings (Clance & O'Toole, 1988; Matthews & Clance, 1985). The fact that impostors experienced greater negative mood irrespective of experimental condition and less positive mood than non-impostors following a high frequency of mistakes task is consistent with the fore-mentioned studies by Frost and colleagues (Frost et al., 1995; Frost et al., 1997). It is noteworthy that these ®ndings are congruent with observations by Frost et al. (1995) that although individuals high in CM do not make more mistakes than other people, they feel much worse about the mistakes they do make. It would seem that impostors, like high CM individuals, also feel worse about making mistakes. The pattern of ®ndings for anxiety, with impostors reporting greater anxiety both prior to and immediately following the Stroop task is consistent with clinical observations and empirical research that describes the impostor phenomenon as an anxiety-related experience (Cozzarelli & Major, 1990; Topping, 1983; Topping & Kimmel, 1985). Impostors' anxiety is likely exacerbated in a situation characterised by many mistakes, in which impostors are particularly prone to overgeneralise a single mistake to mean a total failure. While the present study did not ®nd any evidence to suggest that impostors internalise failure to a greater extent than non-impostors, this result is not unexpected given the absence of performance di€erences for impostors relative to nonimpostors. 4.1. Implications and conclusions The ®ndings from the present study strongly endorse links between impostor fears and perfectionistic concern over mistakes. The anxiety and negative a€ect experienced by impostors when exposed to a situation which presages a high frequency of mistakes indicates that these individuals ®nd such situations stressful and aversive. It follows that therapeutic intervention needs to address these distorted perceptions, in particular impostors' greater estimations of mistakes, lower estimations of success, lesser satisfaction with performance, lower con®dence in performance and lower perceptions of control relative to nonimpostors. It stands to reason that these perceptions are likely to mediate impostors' greater anxiety and more negative mood in situations which forebode a high frequency of mistakes. Based on the ®nding from the present study that impostors have a high fear of negative evaluation from other people and previous research into the impostor phenomenon, it is evident that one basis of impostor fears is fear of social exposure: being exposed to others as fraudulent and lacking in ability, thereby attracting negative judgements from others. Clance and colleagues (Clance et al., 1995; Clance & O'Toole, 1988) propose that impostor fears originate from early childhood experiences whereby the child experiences shame and humiliation when she did not gain parental approval for accomplishments. As social anxiety has also been linked to concern over mistakes (Saboochi & Lundh, 1997) further investigation

646

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

into the relationship between social anxiety and impostor fears may yield further insights into the bases of impostor fears, o€ering new directions for cognitive-behavioural interventions.

References Barrow, J., & Moore, C. (1983). Group interventions with perfectionistic thinking. Personal and Guidance Journal, 61, 612±615. Burns D. D. (1980). The perfectionist's script for self-defeat. Psychology Today, pp. 34±52. Carver, C. S., & Ganellen, R. J. (1983). Depression and components of self-punitiveness: high standards, selfcriticism and overgeneralisation. Journal of Abnormal Psychology, 92, 330±337. Chrisman, S. M., Pieper, W. A., Clance, P. R., Holland, C. L., & Glickauf-Hughes, C. (1995). Validation of the Clance impostor phenomenon scale. Journal of Personality Assessment, 65, 456±467. Clance, P. R. (1985). The impostor phenomenon: overcoming the fear that haunts your success. Atlanta: Peachtree. Clance, P. R., Dingman, D., Reviere, S. L., & Stober, D. R. (1995). Impostor phenomenon in an interpersonal/ social context: origins and treatment. Women and Therapy, 16, 79±97. Clance, P. R., & Imes, S. A. (1978). The impostor phenomenon in high achieving women: dynamics and therapeutic intervention. Psychotherapy: Theory, Research and Practice, 15, 241±247. Clance, P. R., & O'Toole, M. A. (1988). The impostor phenomenon: an internal barrier to empowerment and achievement. Women and Therapy, 6, 51±64. Cozzarelli, C., & Major, B. (1990). Exploring the validity of the impostor phenomenon. Journal of Social and Clinical Psychology, 9, 401±417. Cromwell, B., Brown, N., Sanchez-Huceles, J., & Adair, F. L. (1990). The impostor phenomenon and personality characteristics of high school honor students. Journal of Social Behavior and Personality, 5, 563±573. Ferguson, K., & Rodway, M. R. (1994). Cognitive behavioural treatment of perfectionism: Initial evaluation studies. Research on Social Work Practice, 4, 283±308. Frost, R. O., Marten, P., Lahart, C., & Rosenblate, R. (1990). The dimensions of perfectionism. Cognitive Therapy and Research, 14, 449±468. Frost, R. O., Trepanier, K. L., Brown, E. J., Heimberg, R. G., Juster, H. R., Makris, G. S., & Leung, A. W. (1997). Self-monitoring of mistakes among subjects high and low in perfectionistic concern over mistakes. Cognitive Therapy and Research, 21, 209±222. Frost, R. O., Turcotte, T., Heimberg, R., Mattia, J., Holt, C., & Hope, D. A. (1995). Reactions to mistakes among subjects high and low in perfectionistic concern over mistakes. Cognitive Therapy and Research, 19, 195±205. Hamachek, D. E. (1978). Psychodynamics of normal and neurotic perfectionism. Psychology, 15, 27±33. Holmes, S. W., Kertay, L., Adamson, L. B., Holland, C. L., & Clance, P. R. (1993). Measuring the impostor phenomenon: A comparison of Clance's IP scale and Harvey's I-P scale. Journal of Personality Assessment, 60, 48±59. Imes, S., & Clance, P. R. (1984). Treatment of the impostor phenomenon in high-achieving women. In C. M. Brody, Women therapists working with women: new theory and process of feminist therapy. New York: Springer. Jorgenson, C. B. (1977). Visual set and anxiety on the Stroop phenomenon. Perceptual and Motor Skills, 44, 659± 667. Kernis, M. H., Brochner, J., & Frankel, B. S. (1989). Self-esteem and reactions to failure: the mediating role of overgeneralisation. Journal of Personality and Social Psychology, 57, 707±714. Kernis, M. H., Grannemann, B. D., & Mathis, L. C. (1991). Stability of self-esteem as a moderator of the relationship between level of self-esteem and depression. Journal of Personality and Social Psychology, 61, 80±84. King, J. E., & Cooley, E. L. (1995). Achievement orientation and the impostor phenomenon among college students. Contemporary Educational Psychology, 20, 304±312. Kolligian, J., & Sternberg, R. J. (1991). Perceived fraudulence in young adults: is there an `impostor syndrome'? Journal of Personality Assessment, 56, 308±326. Langford, J. (1990). The need to look smart: The impostor phenomenon and motivations for learning. Doctoral dissertation, Georgia State University. Dissertation Abstracts International, 51, 3604B.

T. Thompson et al. / Personality and Individual Di€erences 29 (2000) 629±647

647

Langford, J., & Clance, P. R. (1993). The impostor phenomenon: Recent research ®ndings regarding dynamics, personality and family patterns and their implications for treatment. Psychotherapy, 30, 495±501. Leary, M. R. (1983). A brief version of the negative evaluation scale. Personality and Social Psychology Bulletin, 9, 371±375. Marsh, H. W. (1990). Self-Descriptive Questionnaire III Manual. San Antonio: The Psychological Corporation. Matthews, G., & Clance, P. R. (1985). Treatment of the impostor phenomenon in psychotherapy clients. Psychotherapy in Private Practice, 3, 71±81. Pacht, A. R. (1984). Re¯ections on perfection. American Psychologist, 39, 391±397. Pallak, M. S., Pittman, T. S., Heller, J. F., & Munson, P. (1975). The e€ect of arousal on Stroop color-word task performance. Bulletin of the Psychonomic Society, 6, 248±250. Prince, T. J. (1989). The impostor phenomenon revisited: a validity study of Clance's IP scale. Unpublished masters, thesis, Georgia State University, Atlanta. Rizley, Z. (1978). Depression and distortion in the attribution of causality. Journal of Abnormal and Social Psychology, 87, 32±48. Russell, D. (1982). The Causal Dimensions Scale: a measure of how individuals perceive causes. Journal of Personality and Social Psychology, 87, 32±48. Saboochi, F., & Lundh, L. G. (1997). Perfectionism, self-consciousness and anxiety. Personality & Individual Di€erences, 22, 921±928. Spielberger, C. D., Gorsuch, R. L., Lushene, R., Vagg, P. R., & Jacobs, G. A. (1983). State-trait anxiety inventory for adults. California: Consulting Psychologists Press. Stroop, J. R. (1938). Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 6, 643± 661. Thompson, T., Davis, H., & Davidson, J. (1998). Attributional and a€ective responses of impostors to academic success and failure outcomes. Personality and Individual Di€erences, 25, 381±396. Thompson, T., & le Fevre, C. (1999). Implications of manipulating post-hoc attributions on the strategy use of defensive pessimists and strategic optimists. Personality and Individual Di€erences, 26, 887±904. Topping M. E. H. (1983). The impostor phenomenon: A study of its construct validity and incidence in university faculty members. Doctoral dissertation, University of South Florida. Dissertation Abstracts International, 44, 1948B-1949B. Topping, M. E. H., & Kimmel, E. B. (1985). The impostor phenomenon: Feeling phony. Academic Psychology Bulletin, 7, 213±225. Watson, D., Clark, L., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative a€ect: the PANAS scales. Journal of Personality and Social Psychology, 54, 1063±1070. Watson, D., & Friend, R. (1969). Measurement of social-evaluative anxiety. Journal of Consulting and Clinical Psychology, 33, 448±457.