Personality and Individual Differences 32 (2002) 865–875
www.elsevier.com/locate/paid
Self-esteem and perfectionism in elite athletes:
effects on competitive anxiety and self-confidence
Nathalie Koivulaa,*, Peter Hassména,b, Johan Fallbya,b,c
b
a
Department of Psychology, Stockholm University, SE-106, Stockholm, Sweden
Stockholm University College of Physical Education and Sports, Stockholm, Sweden
c
Halmstad University, Halmstad, Sweden
Received 24 November 2000; received in revised form 9 April 2001
Abstract
The setting of high standards is an integral part of elite sports, and often beneficial for the athlete’s
performance. However, individuals who are characterized by frequent cognitions about the attainment of
ideal, perfectionistic standards, have been shown to be likely to experience heightened levels of anxiety, due
to discrepancies between ideal and current self/situation. This could of course be detrimental to their sport
performance. The aim of the study was to investigate the relationship between different patterns of perfectionistic dimensions and sport-related competitive anxiety and self-confidence, for elite athletes with
different self-esteem strategies. The results revealed that the relation between self-esteem and perfectionism
differs depending on which dimensions of self-esteem and perfectionism that are being considered. Athletes
with a high self-esteem based on a respect and love for themselves had more positive patterns of perfectionism, whereas athletes who have a self-esteem that is dependent on competence aspects showed a more
negative perfectionism. Further, negative patterns of perfectionism were in the present study related to
higher levels of cognitive anxiety and lower levels of self-confidence. Hence, it seems that sport related
anxiety is positively associated to certain patterns of perfectionism, patterns that are more common in
individuals with specific self-esteem strategies. # 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Elite athletes; Perfectionism; Personality; Anxiety; Self-confidence
Emotion related research in the sport domain has indicated that both cognitive and somatic
anxiety may have differential effects on sport performance (Martens, Vealey, & Burton, 1990;
Turner & Raglin, 1996). An important area of research therefore relates to factors that may
influence precompetition anxiety. Previous empirical findings have for example shown that
* Corresponding author. Fax: +46-8-159342.
E-mail address: [email protected] (N. Koivula).
0191-8869/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0191-8869(01)00092-7
866
N. Koivula et al. / Personality and Individual Differences 32 (2002) 865–875
personal goals and standards, and interpersonal comparison and winning, are significant predictors of cognitive anxiety and self-confidence (Jones, Swain, & Cale, 1991; Lane, Terry, &
Karageorghis, 1995). It has also been suggested that feelings of worry, anxiety, and depression,
are likely to be experienced when there is a considerable discrepancy between desired goals or
future plans, and the current self and situation (Borkovec, Metzger, & Pruzinsky, 1986; Flett,
Hewitt, Blankstein, & Gray, 1998). Based on the aforementioned, it seems plausible, as has also
been suggested (e.g. Flett et al.), that especially individuals who are characterized by frequent
cognitions about the attainment of ideal, perfectionistic standards, are more likely to experience
negative emotions and heightened levels of anxiety and depression symptomatology, due to the
existing discrepancy between ideal and current self/situation. This could of course be detrimental
to sport performance, and findings indicate that anxious individuals are particularly likely to
experience thoughts that interfere with goal-directed performances (e.g. Blankstein, Toner, &
Flett, 1989).
A certain degree of perfectionistic thoughts among elite athletes are not uncommon; in fact,
most elite athletes are probably striving towards perfection. They frequently attest to the idea that
there must exist a perfect performance in their sport, whether it is a perfect hit, throw, run, or
jump. The common coaching instruction that ‘‘practice makes perfect’’ conveys the widespread
belief that given enough practice, athletes may eventually achieve the perfect performance. It has
even been suggested that the desire for perfection is essential, and may in time lead to championship performance (Ellis, 1982). Hence, setting high standards is an integral part of elite
sports, and thereby often beneficial for the athlete’s performance, but when nothing but the perfect performance is perceived to be good enough, these originally positive expectations may
instead lead to the development of a negative self-concept, and a fear-of-failure syndrome (e.g.
Williams & Leffingwell, 1996). Furthermore, it has been shown that individuals who are categorized as perfectionistic have a tendency to engage in excessive cognitive rumination about the
need to attain perfection. They also tend to be overly concerned about mistakes, to have an
exceedingly difficult time forgetting errors, and to have serious doubts about the quality of her/his
actions (e.g. Frost, Marten, Lahart, & Rosenblate, 1990).
The study of perfectionism and related constructs has resulted in the suggestion that perfectionism must be viewed as a multidimensional concept. It includes both negative dimensions
(such as concern over mistakes, doubts about actions, and fear of failure) and positive dimensions
(high personal standards, positive achievement strivings, followed by a sense of satisfaction and
enhanced self-esteem); see, for example, Frost et al. (1990), Hamachek (1978), Stumpf and Parker
(2000), and Terry-Short, Owens, Slade, and Dewey (1995). It seems plausible that the study of
healthy, positive perfectionism, as well as neurotic, negative perfectionism, among elite athletes
could increase our understanding of the psychological dimensions of sport performance and shed
light on how different anxiety patterns develop.
One of the few studies on perfectionism performed so far in the context of sport revealed that
athletes who scored high in concern over mistakes, also reported more anxiety and negative
thinking before competition, less self-confidence in sports, greater difficulty in concentrating, as
well as negative reactions to mistakes. Athletes with a higher level of doubts about actions
exhibited lower self-confidence in athletic contexts, and reported more images of mistakes and
worry about audience reactions (Frost & Henderson, 1991). In a study on perfectionism and
performance anxiety among musicians, it was shown that a certain pattern of perfectionism (high
N. Koivula et al. / Personality and Individual Differences 32 (2002) 865–875
867
concern over mistakes, high doubts about actions, and low personal standards) together with low
self-esteem correlated with performance anxiety (Sinden, 1999).
Some researchers propose a mediation model where self-esteem is considered an important
mediator between perfectionism and mental health (Blatt, 1995; Preusser, Rice, & Asby, 1994;
Rice, Ashby, & Slaney, 1998). This model is based on results indicating that aspects of maladaptive perfectionism are associated with lower self-esteem and higher depression, leading to the
conclusion that maladaptive perfectionists perhaps only experience depression when they also
experience chronic feelings of low self-worth and inadequacy (Rice et al.). But this indicates, as
Rice et al. argue, that also maladaptive perfectionists could have high self-esteem. A suggestion
that somewhat contradicts the ideas of perfectionism as either an antecedent or a consequence of
self-esteem (Adler, 1956; Hollander, 1965; Horney, 1950; Sorotzkin, 1985). Further, the finding
that adaptive perfectionism was not directly, or indirectly through self-esteem, associated with
depression, led Rice and colleagues to the suggestion that a positive association between adaptive
perfectionism and self-esteem only occurs, if at all, in performance or achievement contexts.
These somewhat theoretically incongruent findings may, however, to some extent be explained
by the method used for measuring self-esteem. Firstly, the most common measures of self-esteem
consist of items in which the respondent is asked very directly about her/his evaluation of her/
himself (cf. Johnson, 1997) but, more importantly in non-clinical settings, the self-esteem scales
most often employed consider an individual’s global sense of self-worth, which often reflects not
only a general sense of self-worth, but also self-confidence and competence aspects. However, it
has been suggested that more situation-related concepts of the self, such as self-confidence, should
be kept distinct from self-esteem, and that self-esteem is constituted of two different components.
One component reflecting the individual’s disposition to strive for success and competence for
attaining self-worth, and another that is not related to perceived skills, competencies or other’s
appraisals but to the individual’s basic self-acceptance (Johnson, 1997, 1998). Although selfesteem based on the feeling of being competent, to perform well, and to be appreciated by others,
perhaps is more readily associated with certain dimensions of perfectionism, another dimension
based on the individual’s basic self-acceptance might be related to perfectionism as well. So in
order to understand the association between self-esteem and both adaptive and maladaptive
perfectionism, a second dimension of self-esteem must probably be considered.
Consequently, recent research suggests that two dimensions of self-esteem exist with different
origins, subsequently generating different self-attitudes, one dimension being a more static one,
and defined as a more fundamental self-acceptance, love and appreciation for oneself, and referred to as basic self-esteem. The second dimension captures an individual’s need to be appreciated
and approved by others, to feel competent and in control, and to exert influence over other people (Forsman & Johnson, 1996; Johnson, 1997). This dimension is referred to as earning selfesteem, and can be regarded as reflecting self-esteem more of a state character (e.g. Heatherton &
Polivy, 1991), being more temporary and sensitive to situational factors, and a process by which
an individual can earn self-esteem conditionally. It is further suggested that these two dimensions
are distinct, asymmetrical and hierarchical phenomena. This in turn suggests that the function of
competence for acquiring self-esteem is dependent on the degree of basic self-esteem, and that
individuals can be categorized as having one of four different self-esteem strategies (Forsman &
Johnson, 1996; Johnson, 1997; Johnson & Forsman, 1995). Given that these four groups differ in
their degree of general self-esteem, need of achievement, and level of fear of failure and test
868
N. Koivula et al. / Personality and Individual Differences 32 (2002) 865–875
anxiety (Forsman & Johnson, 1996; Johnson & Forsman, 1995), it is not improbable that these
different self-esteem strategies in association with different patterns of perfectionism result in
various anxiety patterns and self-confidence levels.
Based on the aforementioned, the aim of the present study was to investigate the relation of
different patterns of perfectionistic dimensions to sport-related competitive anxiety and self-confidence, for elite athletes with varying degrees of basic and earning self-esteem. The dimensions
‘‘Concern over Mistakes’’ and ‘‘Doubt about Action’’ have shown to have stronger associations
with depression, anxiety, stress symptoms, and low self-esteem, than other dimensions of perfectionism (e.g. Cheng, Chong, & Wong, 1999; Frost & Henderson, 1991). It was therefore expected
that especially these two dimensions would have an important role in contributing to competitive
anxiety and self-confidence. Although the study on perfectionism in female athletes by Frost and
Henderson revealed several associations between specific dimensions of perfectionism and reactions to athletic competition, the results warrant further research on how certain perfectionistic
patterns, such as scoring relatively high or low on ‘‘Personal Standards’’ in combination with
high or low levels on ‘‘Concern about mistakes’’ and ‘‘Doubts about Actions’’, are related to
competitive anxiety and self-confidence. Also, the inventory employed to measure anxiety prior to
athletic competition in the study by Frost and Henderson, is a measure of trait anxiety, as well as
a measure that combines cognitive and somatic aspects of anxiety. In order to increase our
understanding of the association between perfectionism and anxiety in sport, it could be beneficial to use a measure of state anxiety, which is believed to significantly influence the quality of
the athletic experience (Martens, Burton, Vealey, Bump, & Smith, 1990). To distinguish between
cognitive and somatic anxiety could also be helpful since it has been shown that these have different antecedents, and that performance expectations held before evaluation are more highly
correlated with cognitive than somatic state anxiety (e.g. Alexander & Krane, 1996; Liebert &
Morris, 1967; Morris, Brown, & Halbert, 1977; Morris & Liebert, 1970).
Predicted results were that low basic self-esteem, especially in combination with high earning
self-esteem, would be associated with more negative patterns of perfectionism. It was also
expected that individuals with more negative patterns of perfectionism would report lower sportrelated self-confidence and higher competitive anxiety. In particular cognitive anxiety was expected to be higher for individuals with negative perfectionism because of the tendency of neurotic
perfectionists to ruminate about the inability to reach perfectionistic standards, and because
cognitive anxiety is thought to emanate, as suggested by Martens, Burton et al. (1990), from
evaluative cues, negative feedback, and negative performance expectations.
1. Methods
1.1. Respondents
The Swedish Olympic Committee has established a program that includes all individual athletes
in Sweden that are considered to be possible contenders at the Summer Olympics in 2004, or the
Winter Olympics in 2006. In order to qualify for the program, each athlete must be judged by
their own sport federation to presently be among the top eight athletes in their sport in Sweden.
Hence, all 178 athletes in the present study were elite athletes of high standard, and either already
N. Koivula et al. / Personality and Individual Differences 32 (2002) 865–875
869
Olympians or considered as potential future Olympians. In total, 69 female (mean age=20.3,
S.D.= 3.2) and 109 male (mean age=21.0, S.D.= 3.2) athletes completed the inventories.
1.2. Instruments
Self-esteem was measured by shortened versions of the Basic and Earning Self-Esteem Scales
(Forsman & Johnson, 1996). Out of a total of 42 original items on the Basic Self-Esteem Scale
(SE-B), 12 were used in the present study (e.g. ‘‘I’m satisfied with being the person I am’’). This
scale is comprised of two dimensions; one reflecting warm relations to others and free access to
emotions, the other representing a self-assertiveness and independence in expressing needs and
opinions. Earning self-esteem was measured by 14 items from the original Earning Self-Esteem
Scale (SE-E) (e.g. ‘‘When people like me, my self-esteem is strengthened quite a lot’’). This scale
epitomises three related dimensions; one reflecting the sense of self-esteem that is dependent on
meeting certain standards, one reflects the high ambitions and demands that the individual places
on her/himself, and the third represents the individual’s need for control and power in order to be
able to achieve the high standards he/she has. The longer and shorter versions of these two scales
have been shown to have adequate internal consistency (Cronbach’s alphas ranging between 0.73
and 0.92 for different subsamples) and appears to be reliable and valid measures (Forsman &
Johnson; Johnson, 1997, 1998; Johnson & Forsman, 1995). The items used in the present version
of the scales are chosen so that they represent all the different dimensions of the scales and also
have about the same number of items that have reversed and non-reversed coding. The items
from each scale are presented in a randomized order in a single questionnaire and the respondents
are asked to indicate, on a five-point scale, the extent of agreement with each of the 26 statements; the scale ranging from 1 (‘‘Strongly agree’’) to 5 (‘‘Strongly disagree’’). The participants
were categorized in to one of four self-esteem groups based on their scoring on the two selfesteem scales. A mean split was used to categorize the participants as having a high or low score
on each scale.
In order to measure perfectionism, the Multidimensional Perfectionism Scale (MPS), consisting
of six subscales was used (Frost et al., 1990). Each subscale corresponds to one of the six
dimensions of the Frost et al.’s (1990) model of perfectionism: 1. Concern over Mistakes (CM), 2.
Personal Standards (PS), 3. Doubts about Actions (DA), 4. Organization (O), 5. Parental
Expectations (PE), and 6. Parental Criticism (PC). This instrument has been shown to have adequate internal consistency and validity for the individual scales and the whole instrument (e.g.
Frost, Heimberg, Holt, Mattia, & Neubauer, 1993; Frost & Marten, 1990; Frost et al., 1990). In
the present study, a Swedish translation by Saboonchi and Lundh (see Lundh, Broman, Hetta, &
Saboonchi, 1994; Saboonchi & Lundh, 1997) was used, which also appears to be a reliable and
valid measure (e.g. Saboonchi, 2000). The MPS is a questionnaire in which the respondents are
asked to indicate, on a five-point scale, the extent to which the respondent agrees with each of the
35 statements. The response scale ranges from 1 (‘‘Strongly disagree’’) to 5 (‘‘Strongly agree’’).
Only the scales PS, CM, and DA, were considered in the present study for reasons explained in
the Introduction. Mean splits were used to classify the participants as high or low scorers on each
scale.
Finally, competitive anxiety and self-confidence were assessed by the 27-item Competitive State
Anxiety Inventory-2 (CSAI-2) devised by Martens, Vealey et al. (1990). This sport-specific
870
N. Koivula et al. / Personality and Individual Differences 32 (2002) 865–875
instrument has been shown to measure three relatively independent states, namely self-confidence, cognitive state anxiety, and somatic state anxiety. The CSAI-2 appears to be a valid
instrument with desirable psychometric properties (e.g. Alexander & Krane, 1996; Gould,
Petlichkoff, & Weinberg, 1984; Martens, Burton, et al., 1990). The Cronbach’s alphas for the
Swedish translations of the CSAI-2 subscales used in the present study were: a=0.88 (Self-Confidence), a=0.88 (Cognitive Anxiety), and a=0.85 (Somatic Anxiety). The CSAI-2 is a questionnaire in which the respondents are asked to rate on a four-point scale, with one representing
‘‘not at all’’ and four representing ‘‘very much so’’. The respondents were instructed to select the
most important competition that they had participated in during the previous season, and then to
carefully recall their feelings as they appeared immediately before the chosen competition. The
majority of competitions selected were World and European Championships, together with a few
Nordic Championships. Studies have shown that athletes are in retrospect able to reliably recall
their anxiety level during a competition (e.g. Hanin, 1980, 1986; Imlay, Garda, Stanbrough, &
O’Connor, 1995; Liukkonen, 1991; Raglin & Morris, 1994; Raglin & Turner, 1993; Turner &
Raglin, 1996; Wilson, Raglin, & Harger, 2000).
1.3. Procedure
The questionnaires were administered to approximately 30 athletes at a time in a group setting.
After the athletes were assembled and before they were allowed to commence, standardized verbal and written instructions were given. Ample time was allowed for completing the questionnaires.
2. Results
All variables were tested for normality. When values of skewness and kurtosis were considered
significant (=0.05), transformations were conducted in order to normalize the variables.
A one-way multivariate analysis of variance (MANOVA), with SE-category as the independent
variable and the three MPS-scales as the dependent variables, was performed, revealing a significant effect of SE-category (Pillai–Bartlett trace =0.36, F9,519=7.94, P<0.0001). One followup univariate one-way ANOVA was conducted for each MPS-scale in order to target differences
detected by the MANOVA. To control for Type I errors, when performing multiple ANOVAs,
the Bonferroni method was used to determine the significance level alpha: /c=0.0167. Pairwise
comparisons (Tukey) were conducted following significant results (=0.05).
The ANOVA performed on the MPS subscale ‘‘Personal Standards’’ (PS) revealed a significant
effect of SE-category (F3,173=8.78, P<0.0001). Post hoc tests showed that HH- (high basic and
high earning self-esteem) and LH- (low basic, high earning) individuals had significantly higher
scores on the PS subscale than HL- and LL-individuals, as shown in Table 1.
The ANOVA performed on the MPS subscale ‘‘Concern over Mistakes’’ (CM) disclosed a significant effect of SE-category (F3, 173=8.95, P< 0.0001). Both HH- and HL-individuals had
significantly lower scores on the CM subscale than LH-individuals (see Table 1).
The one-way analysis of variance performed on the MPS subscale ‘‘Doubts about Actions’’
(DA) resulted in a significant effect of SE-category (F3,173=11.11, P<0.0001). HH- and
871
N. Koivula et al. / Personality and Individual Differences 32 (2002) 865–875
HL-individuals had significantly lower scores on the DA-subscale than LH- and LL-individuals
(Table 1).
A one-way multivariate analysis of variance (MANOVA), with MPS-category as the independent variable and the three CSAI-2 subscales as the dependent variables, was performed, revealing a significant effect of MPS-category (Pillai–Bartlett trace =0.29, F21,510=2.55, P< 0.0005).
One-way analyses of variance, with MPS-categorization as the independent variable, were subsequently conducted for each CSAI-2 subscale in order to target differences detected by the
MANOVA. Again, the Bonferroni method was used to determine the significance level alpha: /
c=0.0167. Pairwise comparisons (Tukey) were conducted following significant results (=0.05).
The ANOVA performed on the CSAI-2 Self-Confidence subscale was significant
(F7,170=3.73, P<.001). The LHH-group (low Personal Standards, high Concern over Mistakes,
and high Doubts about Actions) had a significantly lower mean score than the LLL, HLL-, and
HLH-groups.
The one-way analysis of variance conducted on the CSAI-2 Cognitive Anxiety was also significant (F7,170=7.01, P<.0001). The LHH-group (low Personal Standards, high Concern over
Mistakes, and high Doubts about Actions) had a significantly higher mean than all other groups
except for the HHH-group (high Personal Standards, high Concern over mistakes, and high
Doubts about Actions). The HHH-group had a significantly higher mean compared to the HLLand LLL-groups.
However, the differences between the groups on the CSAI-2 Somatic Anxiety subscale were
small, as can be verified in Table 2. Consequently, a non-significant result of the one-way
ANOVA (F7,170=1.59, P=0.09) was obtained.
Table 1
Means (standard deviations in parentheses) of MPS-subscales for each SE-category (Basic/Earning)
Basic/Earning SE
PS
CM
DA
High/High (HH; n=55)
High/Low (HL; n=48)
Low/High (LH; n=37)
Low/Low (LL; n=37)
3.9 (.5)
3.4 (.6)
3.8 (.5)
3.4 (.7)
1.9 (.4)
1.8 (.4)
2.3 (.5)
2.0 (.6)
1.5
1.5
1.9
1.9
(.5)
(.4)
(.5)
(.5)
Table 2
Means (standard deviations in parentheses) of CSAI-subscales for each MPS-category
MPS-category PS/CM/DA
Self-confidence
Cognitive Anxiety
Somatic Anxiety
High/High/High (HHH; n =37)
High/High/Low (HHL; n=22)
High/Low/High (HLH; n=15)
High/Low/Low (HLL; n=24)
Low/High/High (LHH; n=14)
Low/High/Low (LHL; n=16)
Low/Low/High (LLH; n=22)
Low/Low/Low (LLL; n=28)
27.5
29.3
31.5
30.7
24.9
28.2
27.4
30.3
16.0
13.0
12.3
10.8
18.9
12.9
14.0
12.5
16.4
14.4
14.2
14.5
18.0
14.3
16.4
15.1
(5.0)
(4.4)
(4.3)
(5.2)
(5.1)
(3.8)
(5.0)
(5.3)
(5.7)
(3.8)
(2.7)
(2.0)
(7.4)
(3.9)
(3.6)
(4.0)
(4.8)
(4.8)
(2.7)
(4.5)
(4.9)
(2.5)
(5.0)
(5.8)
872
N. Koivula et al. / Personality and Individual Differences 32 (2002) 865–875
3. Discussion
The aim of the study was to inquire about the relation between various self-esteem strategies
and different dimensions of perfectionism. The purpose was also to study how these different
patterns of perfectionistic dimensions are related to competitive anxiety and sport confidence.
Individuals scoring high on earning self-esteem also seem to have high personal standards (PS)
(see Table 1). Further, high basic self-esteem seems to be associated with low degrees of concern
over mistakes (CM) and doubts about actions (DA). Conversely, low basic self-esteem appears to
be related to high scores regarding CM and DA.
A more detailed examination of the results reveals that athletes who were categorized as having
both a high basic and a high earning self-esteem (SE-HH), also displayed high scores on the PS
scale, together with relatively low scores on the CM and DA scales. Looking at the results of the
ANOVAs on the three CSAI-2 subscales shows that the MPS category with high PS, low CM,
and low DA (MPS-HLL), exhibited high self-confidence, low levels of cognitive anxiety, and
relatively low levels of somatic anxiety. This is well in line with findings showing that a high need
for achievement, and a low level of fear of failure and test anxiety characterize the SE-HH group.
It has further been suggested that these individuals primarily engage in tasks in order to enhance
an already high self-esteem. This group is then believed to enjoy challenge, to be inspired by
success, and are expected to employ a realistic and adequate coping strategy (Forsman & Johnson, 1996; Johnson & Forsman, 1995).
Athletes with high basic and low earning self-esteem (SE-HL) scored relatively low on all three
MPS-subscales. The ANOVAs on the three CSAI-2 subscales reveals that the MPS-LLL (a person setting low personal standards, and showing low concern over mistakes and doubts about
actions) displayed high self-confidence, low levels of cognitive anxiety, and relatively low levels of
somatic anxiety. Previous research has also shown that individuals with high basic self-esteem
and low earning self-esteem, generally are characterized by a low need for achievement, a low
degree of Type A behavior, and low levels of test anxiety. These individuals are believed to be
basically contented as they are, thereby giving little importance to showing competence as well as
not being much concerned with failure (Forsman & Johnson, 1996; Johnson & Forsman, 1995).
In contrast, displaying a low basic self-esteem and a high earning self-esteem (SE-LH) seems,
among these athletes, to be related to a relatively high score on the PS subscale, as well as high
scores on the CM and DA subscales of the MPS. And the analyses of the CSAI-2 scales indicate
that those who score high on the three MPS subscales, also display a relatively low sport-related
self-confidence, and relatively high cognitive and somatic anxiety. These findings correspond well
to findings showing that individuals who are categorized as having a low basic and high earning
self-esteem, display high degrees of Type A behavior, a high need for achievement and also fear
of failure and test anxiety. It has been suggested that these individuals in particular constantly
strive to build their self-esteem by constant demands on their own performance and perfection,
and a need for approval and admiration from others. Accordingly, their need to succeed is high.
Being unsuccessful therefore becomes a severe threat to an already weak self-esteem, and their
inflated fear for failure results in increased anxiety (Forsman & Johnson, 1996; Johnson & Forsman, 1995).
The fourth group of athletes, with low basic self-esteem and low earning self-esteem (SE-LL),
scored relatively low on the PS (personal standards) scale, somewhat high on CM (concern over
N. Koivula et al. / Personality and Individual Differences 32 (2002) 865–875
873
mistakes), and high on DA (doubts about actions). Individuals both with low basic and earning
self-esteem have earlier been shown to be characterized by a low need for achievement and a low
degree of Type A behavior, but a high level of test anxiety; they are according to the literature
expected to renounce challenge (Forsman & Johnson, 1996; Johnson & Forsman, 1995).
The general patterns discussed herein are, however, accompanied by individual differences. This
indicates that the association between self-esteem and perfectionism is complex, and that other
psychological factors, such as the athlete’s coping skills and strategies for stressful situations, may
act as mediators. Something that warrants further research along these lines.
It has earlier been shown that, especially unhealthy, perfectionism is correlated with a lack of
self-esteem (e.g. Cheng et al., 1999; Stumpf & Parker, 2000), but it is still not readily evident in
what way specific dimensions of perfectionism are related to self-esteem; different suggestions
have been made. Hollander (1965) claimed that an excessive striving for perfection was a common response to a low self-worth. This is consistent with Adler (1956), who proposed the idea
that neurotic perfectionism is a self-focused tactic to protect a fragile self-esteem. Some have
instead argued that a low general self-esteem is a more or less inevitable consequence for the
perfectionist individual. Because the perfectionist’s self-worth- and -esteem is based on the
attainment of perfectionistic and often unattainable goals, the slightest negative feedback will
thereby be regarded as evidence for the discrepancy between current and ideal self and situation,
resulting in a lower self-esteem (Horney, 1950; Sorotzkin, 1985), but the findings of studies using
the mediational model, placing self-esteem as a mediator between perfectionism and mental
health, could indicate that perfectionism is not an antecedent or a consequence of self-esteem,
since perfectionistic individuals can display both high or low self-esteem.
However, the relation between perfectionism and self-esteem may be such that some aspects of
self-esteem can affect certain dimensions of perfectionism, but also the other way around. It is not
improbable that individuals with a low basic self-esteem and a high earning self-esteem strive for
perfection as a tactic not only to protect but also to increase their weak self-esteem. The definite
possibility of being unsuccessful is nevertheless a threat to their fragile self-esteem, which results
in fear of failure and thereby in doubts about their actions and concerns about mistakes, and
higher levels of anxiety. This self-esteem archetype (low basic/high earning; SE-LH) seems to be
related to a negative pattern of perfectionism. It is, however, also possible that acts evaluated as
successful will, although temporarily, increase their self-esteem which in turn may temporarily
buffer the effects of their negative pattern of perfectionism on mental health. Further, individuals
who instead have high basic and high earning self-esteem (SE-HH) seem to enjoy setting high
personal standards, regarding them as challenges. If they succeed in attaining their goals, their
already high self-esteem will be even further enhanced, but if they do not reach their goals, this
will not pose a negative threat to their self-esteem. Since these issues were not explicitly investigated in the present research, these suggestions are only speculative, warranting further research.
In conclusion, then, it seems likely that the relation between self-esteem and perfectionism differs depending on which dimensions of self-esteem and perfectionism that are being considered. It
seems that individuals with a high self-esteem based on a respect and love for themselves exhibit
more positive patterns of perfectionism. Whereas those who have a self-esteem that is dependent
on competence aspects show a more negative perfectionism. Negative patterns of perfectionism
were in the present study related to higher levels of cognitive anxiety and lower levels of self-confidence. However, it is also possible that the effect of perfectionism on anxiety and self-confidence is
874
N. Koivula et al. / Personality and Individual Differences 32 (2002) 865–875
dependent on the individual’s self-esteem strategy, in that for example individuals with negative
perfectionistic patterns do not experience high levels of cognitive anxiety if they have a high basic
self-esteem and little need to acquire self-esteem through achievements and others’ appreciation.
This has not been analyzed in the present study, but is an intriguing aspect to study in future
research.
Acknowledgements
This study was supported by the Swedish Olympic Committee and The Swedish Center for
Sport Research.
References
Adler, A. (1956). The neurotic disposition. In H. L. Ansbacher, & R. R. Ansbacher (Eds.), The individual psychology of
Alfred Adler (pp. 239–262). New York: Harper.
Alexander, V., & Krane, V. (1996). Relationships among performance expectations, anxiety, and performance in collegiate volleyball players. Journal of Sport Behavior, 19, 246–269.
Blankstein, K. R., Toner, B., & Flett, G. L. (1989). Test anxiety and the contents of consciousness: thought listing and
endorsement measures. Journal of Research in Personality, 23, 269–286.
Blatt, S. J. (1995). The destructiveness of perfectionism: implications for the treatment of depression. American Psychologist, 50, 1003–1020.
Borkovec, T. D., Metzger, R. L., & Pruzinsky, T. (1986). Anxiety, worry, and the self. In L. M. Hartman, &
K. R. Blankstein (Eds.), Perception of self in emotional disorder and psychotherapy (pp. 219–260). New York: Plenum.
Cheng, S. K., Chong, G. H., & Wong, C. W. (1999). Chinese Frost Multidimensional Perfectionism Scale: a validation
and prediction of self-esteem and psychological distress. Journal of Clinical Psychology, 55, 1051–1061.
Ellis, A. (1982). Self-direction in sport and life. Rational Living, 17, 27–33.
Flett, G. L., Hewitt, P. L., Blankstein, K., & Gray, L. (1998). Psychological distress and the frequency of perfectionistic
thinking. Journal of Personality and Social Psychology, 75, 1363–1381.
Forsman, L., & Johnson, M. (1996). Dimensionality and validity of two scales measuring different aspects of selfesteem. Scandinavian Journal of Psychology, 37, 1–15.
Frost, R. O., Heimberg, R. G., Holt, C. S., Mattia, J. I., & Neubauer, A. L. (1993). A comparison of two measures of
perfectionism. Personality and Individual Differences, 14, 119–126.
Frost, R. O., & Henderson, K. J. (1991). Perfectionism and reactions to athletic competition. Journal of Sport and
Exercise Psychology, 13, 323–335.
Frost, R. O., & Marten, P. A. (1990). Perfectionism and evaluative threat. Cognitive Therapy and Research, 14, 559–572.
Frost, R. O., Marten, P., Lahart, C., & Rosenblate, R. (1990). The dimensions of perfectionism. Cognitive Therapy and
Research, 14, 449–468.
Gould, D., Petlichkoff, L., & Weinberg, R. S. (1984). Antecedents of temporal changes in and relationships between
the CSAI-2 subcomponents. Journal of Sport Psychology, 6, 289–304.
Hamachek, D. E. (1978). Psychodynamics of normal and neurotic perfectionism. Psychology, 15, 27–33.
Hanin, Y. L. (1980). A study of anxiety in sports. In W. F. Straub (Ed.), Sport psychology: an analysis of athlete
behavior (pp. 236–249). Ithaca, NY: Mouvement Publications.
Hanin, Y. L. (1986). State-trait anxiety research on sports in the USSR. In C. D. Spielberger, & R. Diaz-Guerrero
(Eds.), Cross-cultural anxiety(Vol. 2) (pp. 45–64). Washington, DC: Hemisphere Publishing.
Heatherton, T. F., & Polivy, J. (1991). Development and validation of a scale for measuring state self-esteem. Journal
of Personality and Social Psychology, 60, 895–910.
Hollander, M. H. (1965). Perfectionism. Comprehensive Psychiatry, 6, 94–103.
Horney, K. (1950). Neurosis and human growth: the struggle toward self-realization. New York: Norton.
N. Koivula et al. / Personality and Individual Differences 32 (2002) 865–875
875
Imlay, G. J., Garda, R. D., Stanbrough, M. E., & O’Connor, P. J. (1995). Anxiety and athletic performance: a test of
zone of optimal function theory. International Journal of Sport Psychology, 26, 295–306.
Johnson, M. (1997). On the dynamics of self-esteem. Empirical validation of basic self-esteem and earning self-esteem.
Doctoral dissertation, Stockholm University, Stockholm.
Johnson, M. (1998). Self-esteem stability: the importance of basic self-esteem and competence strivings for the stability
of global self-esteem. European Journal of Personality, 12, 103–116.
Johnson, M., & Forsman, L. (1995). Competence strivings and self-esteem. Personality and Individual Differences, 19, 417–430.
Jones, G., Swain, A., & Cale, A. (1991). Gender differences in precompetition temporal patterning and antecedents of
anxiety and self-confidence. Journal of Sport and Exercise Psychology, 13, 1–15.
Lane, A., Terry, P., & Karageorghis, C. (1995). Antecedents of multidimensional competitive state anxiety and selfconfidence in duathletes. Perceptual and Motor Skills, 80, 911–919.
Liebert, R. M., & Morris, L. W. (1967). Cognitive and emotional components of test anxiety: a distinction and some
initial data. Psychological Reports, 20, 975–978.
Liukkonen, J. (1991). Application of the STAI for Finnish athletes and coaches. Paper presented at the 12th International Conference of the Society for Test Anxiety Research in Budapest, Hungary, July 1991.
Lundh, L. G., Broman, J. E., Hetta, J., & Saboonchi, F. (1994). Perfectionism and insomnia. Scandinavian Journal of
Behavioral Therapy, 23, 3–18.
Martens, R., Burton, D., Vealey, R., Bump, L., & Smith, D. (1990). The development of the Competitive State Anxiety
Inventory-2 (CSAI-2). In R. Martens, R. S. Vealey, & D. Burton (Eds.), Competitive anxiety in sport (pp. 117–190).
Champaign, IL: Human Kinetics.
Martens, R., Vealey, R., & Burton, D. (1990). Competitive anxiety in sport. Champaign, IL: Human Kinetics.
Morris, L.W., Brown, N.R., Halbert, B. (1977). Effects of symbolic modeling on the arousal of cognitive and affective
components of anxiety in preschool children. In C.D. Spielberger & I.G. Sarason (Eds.), Stress and anxiety (Vol. 4).
Washington, DC: Hemisphere.
Morris, L. W., & Liebert, R. M. (1970). The relationship of cognitive and emotional components of test anxiety to
physical arousal and academic performance. Journal of Consulting and Clinical Psychology, 35, 332–337.
Preusser, K. J., Rice, K. G., & Ashby, J. S. (1994). The role of self-esteem in mediating the perfectionism–depression
connection. Journal of College Student Development, 35, 88–93.
Raglin, J. S., & Morris, M. J. (1994). Precompetition anxiety in women volleyball players: a test of zof theory in a team
sport. British Journal of Sports Medicine, 28, 47–51.
Raglin, J. S., & Turner, P. E. (1993). Anxiety and performance in track and field athletes: a comparison of the invertedU hypothesis with zone of optimal function theory. Personality and Individual Differences, 14, 163–171.
Rice, K. G., Ashby, J. S., & Slaney, R. B. (1998). Self-esteem as a mediator between perfectionism and depression: a
structural equations analysis. Journal of Counseling Psychology, 45, 304–314.
Saboonchi, F. (2000). Perfectionism. Conceptual, emotional, psychopathological, and health-related implications.
Doctoral Dissertation, Stockholm University, Stockholm.
Saboonchi, F., & Lundh, L. G. (1997). Perfectionism, self-consciousness, and anxiety. Personality and Individual Differences, 22, 921–928.
Sinden, L. M. (1999). Music performance anxiety: Contributions of perfectionism, coping style, self-efficacy, and selfesteem. Doctoral Dissertation, Arizona State University, 1999 [Dissertation Abstracts International Section A (p.60)].
Sorotzkin, B. (1985). The quest for perfection: avoiding guilt or avoiding shame? Psychotherapy, 22, 564–571.
Stumpf, H., & Parker, W. D. (2000). A hierarchical structural analysis of perfectionism and its relation to other personality characteristics. Personality and Individual Differences, 28, 837–852.
Terry-Short, L. A., Owens, R. G., Slade, P. D., & Dewey, M. E. (1995). Positive and negative perfectionism. Personality and Individual Differences, 18, 663–668.
Turner, P. E., & Raglin, J. S. (1996). Variability in precompetition anxiety and performance in college track and field
athletes. Medicine and Science in Sports and Exercise, 28, 378–385.
Williams, J. M., & Leffingwell, T. R. (1996). Cognitive strategies in sport and exercise psychology. In J. L. Van Raalte, &
B. W. Brewer (Eds.), Exploring sport and exercise psychology. Washington, DC: American Psychological Association.
Wilson, G. S., Raglin, J. S., & Harger, G. J. (2000). A comparison of the STAI and CSAI-2 in five-day recalls of precompetition anxiety in collegiate track and field athletes. Scandinavian Journal of Medicine and Science in Sports, 10, 51–54.