Content uploaded by Konstantinos Karteroliotis
Author content
All content in this area was uploaded by Konstantinos Karteroliotis
Content may be subject to copyright.
438
Stavrou, Zervas, and Karteroliotis are with the Dept. of Physical Education and Sport Sciences, Uni-
versity of Athens, Greece. Jackson is with the University of Queensland, Australia.
The Sport Psychologist, 2007, 21, 438-457
© 2007 Human Kinetics, Inc.
Flow Experience and Athletes’
Performance With Reference
to the Orthogonal Model of Flow
Nektarios A. Stavrou, Yannis Zervas,
and Konstantinos Karteroliotis
University of Athens, Greece
Susan A. Jackson
University of Queensland, Australia
The purposes of the current study were to examine (a) the differences in Flow State
Scale (FSS) subscales between the 4 experiential states of the orthogonal model
(apathy, anxiety, relaxation, and flow), (b) the relationship between challenge,
skills, and flow experience; and (c) the relationship between flow experience and
athletes’ performance. Two hundred twenty athletes volunteered to participate in
this study. Challenge of the game and skills of the athlete were measured before
and after competition. Thirty minutes after the competition, the FSS was used
to measure flow experience. In addition, subjective and objective measures of
athletes’ performance were assessed. Athletes in the flow and relaxation states
revealed the most optimal states, whereas the athletes in the apathy state showed
the least optimal state. There were positive associations between athletes’ flow
experience and their performance measures, indicating that positive emotional
states are related to elevated levels of performance. On the other hand, there were
low or no correlations between athletes’ performance and reported challenge of the
game, whereas skills of the athlete were moderately correlated with flow. Multiple-
regression analysis demonstrated significant prediction of athletes’ performance
based on flow experience during competition. Future research should examine
the relationship between flow, athletes’ performance, and additional dispositional
and state variables.
The concept of flow has been used in psychology to describe the intrinsically
rewarding experience that people can experience during an activity. According to
Csikszentmihalyi and Csikszentmihalyi (1988), the concept of flow, or optimal
experience, is obtained “when all the contents of consciousness are in harmony with
each other, and with the goals that define the person’s self. These are the subjec-
tive conditions we call pleasure, happiness, satisfaction, enjoyment” (p. 24). The
PROOF
Flow Experience and Athletes’ Performance 439
theory of flow experience has been applied in various research domains, such as
work (e.g., Csikszentmihalyi & LeFevre, 1989), leisure (e.g., Kleiber, Larson, &
Csikszentmihalyi, 1986), learning environments (e.g., Stein, Kimiecik, Daniels, &
Jackson, 1995), psychopathology (Csikszentmihalyi, 1982), sports (e.g., Jackson,
1992, 1996; Jackson, Kimiecik, Ford, & Marsh, 1998; Jackson, Thomas, Marsh, &
Smethurst, 2001), and exercise activities (e.g., Jackson & Eklund, 2002; Vlachopou-
los, Karageorghis, & Terry, 2000). Flow experience has been described as an optimal
subjective mental state marked by positive affect, centering of attention, absorp-
tion, spontaneous action, total immersion in performing an activity, perception of
control over actions and the environment, immediate and unambiguous feedback,
loss of self-consciousness, distortion of time, and perception of superior functioning
(Csikszentmihalyi, 1982; Csikszentmihalyi & Csikszentmihalyi, 1988; Jackson,
1992, 1996). An important issue associated with the generation of flow is whether
particular situational or personal characteristics facilitate the experience of flow.
The examination of flow theory in various research domains, such as work, leisure
activities, education, or sport indicates that flow experience differs in intensity and
frequency, as well as with respect to the situational or individual factors that gener-
ate flow (Csikszentmihalyi, 1975; Csikszentmihalyi & Csikszentmihalyi; Privette,
1983; Privette & Bundrick, 1991; Ravizza, 1977; Stein et al., 1995).
Orthogonal Model of Flow Theory
A central issue of flow theory is that whether or not an individual is in flow depends
on his or her perception of the existing challenges and the nature of his or her
skills rather than on the objective nature of the challenges or skills themselves
(Csikszentmihalyi, 1975, 1982). When the challenges and skills are perceived as
being in balance, the person enjoys the moment and stretches his or her capabili-
ties to learn new skills and increase self-esteem and personal complexity. Thus,
the person feels that he or she can act on these skills without feelings of boredom,
anxiety, or worry. On the other hand, when the skills outperform the challenge, there
is relaxation, whereas when the skills and challenges are below average, there is
apathy, and, finally, when the challenges outweigh the skills, there will be anxiety
(Nakamura & Csikszentmiahlyi, 2002). A person could try to reach flow state by
means of a twofold dynamic. For example, in an anxiety state, the person would
try to increase personal skills to balance the level of challenge, or, if the person
experiences relaxation, he or she would try to seek more challenging situations
(Moneta & Csikszentmihalyi, 1996). In other words, as a person masters a chal-
lenging activity, his or her skill level increases. To continue experiencing flow, the
person must try to find more complex skills, and so on, building greater complexity
in the person (Csikszentmihalyi, 1990).
Moneta and Csikszentmihalyi (1996) examined the flow experience in various
contexts, indicating that “the balance of challenges and skills has a positive effect
in some contexts and little or no effect in others” (p. 302). In other words, the chal-
lenges of the competition and the skills of the athlete are two subjective experiential
variables, which exert a dependent effect on each other or independent effect on
the quality of experience. Before or during competition, challenge and skills level
are dynamic in nature, depending on individual qualities (e.g., experience, mental
PROOF
440 Stavrou et al.
preparation, physical preparation) or situational characteristics (e.g., importance
of competition, difficult opponent). Stein and colleagues (1995) reported that in a
competitive environment the level of a person’s perceived skills is positively related
with the quality of experience, whereas in a learning environment both the person’s
perceived skills and the challenges are related to the quality of experience.
Jackson and Roberts (1992) examined the balance between challenges of the
competition and athletes’ skills during their best and worst performances. The
results revealed large differences in mean scores between challenges and skills
in athletes’ worst performances, whereas no differences were found in their best
performances. Furthermore, the mean values of challenges and skills were higher
in their best than in their worst performances.
Flow Experience and Athletic Performance
The relationship between the flow concept and sport performance is of great inter-
est for athletes, coaches, and applied sport psychology consultants. Flow has been
examined either as a phenomenon or as a concomitant of performance (Jackson
& Csikszentmihalyi, 1999; Jackson & Wrigley, 2004), with a close relationship
suggested between peak performance and flow (Jackson, 1992, 1999; Jackson &
Roberts, 1992; Jackson et al., 1998; Jackson et al., 2001; Ravizza, 1977).
Examining the similarity or overlap between peak performance and flow,
Privette (1983) and Privette and Bundrick (1991) described the distinguishing and
similar characteristics of these concepts. They concluded that flow is an intrinsically
rewarding experience, whereas peak performance is characterized as a person’s
optimal functioning. In addition, Privette and Bundrick (1997) reported that peak
performance is characterized as playful, fun, and fulfilling, whereas flow is marked
by fun and enjoyment.
Based on the aforementioned, the flow concept cannot be used interchange-
ably with the terms of peak performance and peak experience, because one might
be in flow without necessarily achieving these other outcomes. When an athlete
experiences peak performance, however, he or she appears to be in flow. Csikszent-
mihalyi and LeFevre (1989) state that flow experience constitutes a combination of
characteristics that typify peak performance and peak experience, whereas Jackson
(1988; Jackson & Roberts, 1992) proposed that flow might be a precursor to, or
the psychological process underlying, peak performance.
Flow is associated with high levels of performance and positive experience.
Jackson’s qualitative content analyses of best performances (1992, 1995, 1996) and
Jackson and Roberts’s (1992) quantitative results showed that athletes’ best perfor-
mances were associated with flow characteristics. Athletes in their best performances
indicated higher flow ratings than in either their worst performances or when they
generally compete (Jackson & Roberts). Factors such as total commitment, clearly
defined goals, feedback about how well an athlete is performing, concentration
on performing the activity, task-relevant thoughts, sense of control, and feelings
of fun, confidence, and enjoyment were among the most frequent psychological
characteristics that athletes mentioned during high levels of performance (e.g.,
Gould, Eklund, & Jackson, 1992).
PROOF
Flow Experience and Athletes’ Performance 441
Kimiecik and Stein (1992) suggested that to better understand the flow experi-
ence researchers need to examine subjective states along with objective outcomes.
Jackson and colleagues (2001) attempted to examine relationships between flow
and both subjective and objective criteria. Measures of performance included
finishing position and perceived success, and associations with flow were found
with both types of measures.
To date, despite the great interest in examining the psychological issues of
athletes’ performance, sport psychologists have focused mainly on the negative
factors of athletes’ experience, such as anxiety and stress, ignoring the positive
psychological qualities underlying elevated levels of performance. Identifying the
relation between optimal psychological states and athletes’ performance might be
helpful to the development of mental training programs to help promote optimal
mental states. From a theoretical point of view, the examination of the relationship
between the orthogonal model and flow experience in sports has not been examined.
In addition, the study of flow experience and sport performance to date has been
based primarily on athletes’ subjective perceptions and interviews during high and
low levels of performance, as well as on the comparison of successful and less
successful performances (Jackson, 1992, 1995, 1999; Jackson & Roberts, 1992).
Therefore, it seems important to examine the independent relation of challenge and
skills, as well as challenge:skill ratios, with flow experience and athletes’ perfor-
mance. Moreover, the examination of the relation between flow factors and both
subjective and objective measures of performance might provide more comprehen-
sive information about the psychological qualities that underlie sport performance,
from a quantitative point of view. Thus, the purposes of the current study were to
examine (a) the differences in Flow State Scale (FSS) subscales between the four
experiential states of the orthogonal model (apathy, anxiety, relaxation, and flow);
(b) the relationship between challenge, skills, and flow experience; and (c) the
relationship between flow experience and athletes’ performance.
Method
Participants
A sample of 220 athletes (51% male, 49% female) volunteered to participate in
this study. They represented seven individual sports (80 track and field, 63 swim-
ming, 34 shooting, 13 archery, 20 cycling, and 10 canoeing and kayaking). The
athletes ranged in age from 16 to 38 years (Μ = 19.95, SD = 4.61), had competitive
experience from 2 to 22 years (Μ = 6.98, SD = 4.33), and had participated in 10–450
national games, with a mean of approximately 88 games (SD = 77.44), and their
experience in international games ranged from 0 to 125 competitions (Μ = 6.78,
SD = 16.39). Participants met the following criteria: They (a) were active athletes
in an individual sport, (b) had at least 2 years of competitive experience, and (c)
had participated in at least 10 competitions. The criterion of individual sports was
chosen, because flow might be experienced differently by team-sport athletes,
and the latter are affected by different factors such as interaction with other team
players. The criteria of minimum 2 years experience and 10 competitions were set
to ensure that the athletes were familiar with the technique of their sports. In the
PROOF
442 Stavrou et al.
beginning stages of their sports, technique is something that prevents athletes from
focusing on their performance.
Instrumentation
The FSS. In this study, the Greek version of the FSS was used. The original FSS
is a 36-item self-report scale developed by Jackson and Marsh (1996) to assess
the magnitude of flow characteristics experienced during a specific event. Stavrou
and Zervas (2004) in a series of three studies, using confirmatory factor analytic
procedures, provided acceptable factor structure for the Greek version of the FSS.
In addition, the Greek version of the FSS indicated acceptable internal consistency,
content, and concurrent validity (Stavrou & Zervas). Each of the nine subscales
consists of four items. Responses are given on a 5-point Likert-type scale from 1
(strongly disagree) to 5 (strongly agree). Flow experience is proposed to consist
of a number of characteristics (e.g., Csikszentmihalyi, 1990). These characteristics
typify the factors, or subscales, of the FSS (Jackson & Marsh), which assesses
the following qualities: challenge–skill balance, action-awareness merging, clear
goals, unambiguous feedback, concentration on task at hand, sense of control,
loss of self-consciousness, transformation of time, and autotelic experience. In
addition, a global flow score was included in the analysis to better understand the
concept of flow experience from a more holistic perspective (Jackson & Marsh).
The Cronbach’s alphas of the FSS factors of the current study ranged between .77
and .93 and are shown in Table 1.
Challenge and Skills Measures.Two 11-point Likert-type scales were administered
to measure the challenge of the competition and the perceived skill levels. This is the
typical approach to assessing levels of challenge and skill used by Csikszentmihalyi
and colleagues (e.g., Csikszentmihalyi & Csikszentmihalyi, 1988) and Jackson and
colleagues (e.g., Jackson & Marsh, 1996; Jackson & Roberts, 1992). The anchors
for the two scales used in this study ranged from 0 (not at all) to 10 (very much),
with a midpoint of 5 (medium). The two scales were “How challenging was this
event for you?” measuring the perceived challenge of the competition, and “How
skilled were you in this event?” measuring the perceived skills of the athlete.
Demographics.A questionnaire was developed to obtain demographic information
such as participants’ gender, sport, and competitive experience.
Performance Measures
Subjective Measure of Performance.One hour before the competition, the athletes
noted the exact performance (i.e., time, distance, score, or points) they set as a goal
in the competition in which they intended to participate. Specifically, the question
given to the athletes was “What is your target for this competition? Indicate the
exact performance” (e.g., 560/600 points in archery, 7 m in long jump, 11 s in 100
m). Thirty minutes after the competition, the athletes marked the final performance
they had just achieved in the event. The records were kept either by the athletes
themselves or by the researchers. At the same time, the athletes completed an
11-point Likert bipolar scale from –5 (very low performance) to +5 (very high
PROOF
443
Table 1 Flow State Scale: Intercorrelations Among the Subscales and the Global Flow Score
Flow State Scale Subscales X
1
X
2
X
3
X
4
X
5
X
6
X
7
X
8
X
9
X
10
X1Challenge–skill balance (.81) .38** .64** .68** .46** .61** .39** –.02 .75** .82**
X2Action-awareness merging (.78) .29** .33** .36** .35** .34** .19* .28** .58**
X3Clear goals (.80) .56** .56** .59** .27** –.02 .46** .71**
X4Unambiguous feedback (.86) .45** .61** .39** –.03 .66** .78**
X5Concentration on task at hand (.86) .71** .37** –.10 .49** .71**
X6Sense of control (.88) .45** –.16 .60** .79**
X7Loss of self-consciousness (.77) –.03 .47** .63**
X8Transformation of time (.78) –.02 .13
X9Autotelic experience (.93) .81**
X10 Global flow score (.93)
*p < .01. **p < .001.
Note. Internal-consistency coefficients (Cronbach’s α) are presented in parentheses along the diagonal.
PROOF
444 Stavrou et al.
performance), on which they evaluated their performance based on the target set
before the competition.
Objective Measure of Performance. Three different formulae were used to
objectively measure athletes’ performance, with regard to the type of sport, as well
as to the type of performance measurement (time, distance, point). The athletes
were divided into three groups. Based on the suggestions of Lane, Terry, Beedie,
Curry, and Clark (2001), a different formula was applied in each group, aimed to
objectively measure athletes’ performance. The performance measures took into
account the best performance of the athlete, his or her preperformance goal, and
the final result. Researchers have indicated that using either the preperformance
goal as a criterion for estimating the relative success of performance outcome or
judging the performance outcome with athlete’s best performance is limited (Lane
& Karageorgis, 1997; Martin & Gill, 1991, 1995). Using the best performance of
the athlete and the preperformance goal provides a more comprehensive and fruit-
ful measure of athlete’s performance, because the previous best performance and
preperformance goal can serve as criteria to determine the difficulty of the athlete’s
goal, as well as the relative success of the outcome.
The first performance-measure group comprised athletes for whom perfor-
mance was measured in time. The “time sports” were track and field, swimming,
cycling, and canoeing or kayaking. The performance was measured based on the
following formula: (previous-best-performance time – result time) + (competition
goal time – result time). The time was measured in tenths of a second. The second
performance group constituted archers and shooters. The objective performance
measure for “target sports” was obtained using the following formula: (points
result – previous-best-performance points) + (points result – competition goal). The
performance was measured in points. Triple jumpers, long jumpers, pole-vaulters,
and shot-putters made up the third performance group, in which athletes’ perfor-
mance was measured in distance. The objective performance measure for “distance
sports” was measured using the following formula: (distance result – previous-
best-performance distance) + (distance result – competition goal). Then, the three
performance measures were transformed to z values to obtain a single measure of
athletes’ performance, which represented the objective performance measure.
Procedure
The athletes were recruited from various individual sports by contacting the coaches
or the athletes themselves, visiting their sport clubs. They were informed about
the purpose of the study, the assessment, and the procedure of data collection. The
athletes were asked to voluntarily participate, and they completed a consent form,
being informed about the confidentiality of the data.
The athletes completed the two scales regarding the perceived challenge of
the game and their skills 1 day and 1 hr before the competition, with reference to
how they felt at that time. Thirty minutes after the event, the athletes completed
the FSS, as well as the challenge–skill ratings, based on how they felt during the
competition. The instructions for the instruments were modified to be adequate
for immediate use after the competition. The total duration of data collection was
6 months.
PROOF
Flow Experience and Athletes’ Performance 445
Statistical Analyses
Statistical analyses of the data were divided into two phases. The first phase con-
sisted of preliminary data analysis to satisfy the assumptions of the main analyses
that were conducted (the second phase). To satisfy the assumptions for conducting
multivariate and univariate analyses of variance, data screening (univariate distri-
bution, multivariate distribution, Mahalanobis distance values, Levene’s test, Fmax
ratio values, Box’s M test) was performed before main data analysis (Tabachnick
& Fidell, 2006). In addition, Cronbach’s α coefficient was used to examine the
internal reliabilities of the FSS subscales.
In the second phase, univariate and multivariate statistical analyses were
conducted to address the main purposes of the current study. To examine whether
athletes in the four experiental states (apathy, anxiety, relaxation, flow) differed
significantly in the FSS subscales during competition, multivariate analysis of vari-
ance (MANOVA) was conducted. Follow-up univariate ANOVAs were performed
on the subscales when there were significant MANOVA effects (Scheffé test). In
addition, Bonferroni adjustment was applied to control for the inflation of Type I
error (Tabachnick & Fidell). Repeated-measure ANOVA was used to examine the
differences between the three time measures (1 day before, 1 hr before, and 1 hr
during competition) on the perceived challenge of the game and the skills of athletes
(Tabachnick & Fidell, 2006). Gender differences were examined using a t test. To
examine the relationships among the variables, Pearson’s r correlation analysis was
used. Fisher’s z transformation was performed to examine whether the value of the
correlations between challenge of the game and flow, compared with athletes’ skills
and flow, differed significantly. In addition, Fisher’s z transformation was used to
examine the difference in the value of the correlations between the two types of
performance measure (subjective, objective) and the FSS subscales (Cohen, 1988).
Standard multiple-regression analysis was conducted to examine the level of predic-
tion of the flow subscales on the two types of performance measures (objective and
subjective; Cohen, Cohen, West, & Aiken, 2003; Tabachnick & Fidell). To control
for Type I error with the inclusion of several predictor variables, a p value of .01
was used for establishing significance in these analyses.
Results
Data Screening. Univariate and multivariate distribution analyses were performed
before data analyses (Tabachnick & Fidell, 2006). Skewness and kurtosis indicated
low values of the examined variables. Examination of Mahalanobis distance values
indicated no multivariate outliers (p < .001) among the independent variables. The
equality of covariance matrices was acceptable at the univariate level (Levene’s test,
Fmax ratio values). The homogeneity of variance-covariance, however, was violated
at the multivariate level (Box’s M test). Therefore, Pillai’s trace was chosen as the
appropriate multivariate test statistic because of its robustness over test violations
(Tabachnick & Fidell). No missing data were indicated in the sample.
Gender Differences. No significant differences were found between males and
females in age (t = .98, n.s.), and competitive experience (t = 1.16, n.s.). The FSS
PROOF
446 Stavrou et al.
subscales were examined for gender differences. No significant differences, Pillai’s
trace = .046, F(2, 218) = 1.117, n.s.) were found between males and females in
flow experience.
Correlational Analysis. Significant intercorrelations were found among the flow
state scale subscales (Table 1). The factors of the FSS indicated low- to high-value
intercorrelations, ranging from .27 to .75, except for the factor transformation of
time. Transformation of time did not correlate significantly with the other FSS
subscales. The mean value of the FSS subscale intercorrelations (excluding time
transformation) was .48 (rmean = .48) with a shared variance proportion of 23%.
Six of the scales (challenge–skill balance, clear goals, unambiguous feedback,
concentration on task at hand, sense of control, and autotelic experience), however,
revealed a mean correlation of .59 (shared variance 35%).
Significant positive correlations were found among the FSS subscales (except
for time transformation) and athletes’ performance (subjective, objective). These
correlations are presented in Table 2. The value of the correlations between the two
types of performance measures and FSS subscales were compared using Fisher’s z
transformation. The results indicated that subjective measure of performance was
significantly more highly correlated with challenge-skill balance, loss of self-con-
sciousness, autotelic experience, and global flow experience, when compared to
the objective measure of performance. Regarding the rest of the FSS factors there
were no significant differences, but there was a trend toward higher correlations
with subjective compared with objective measures of performance. Summarizing
the results of correlations between FSS and the two types of performance measures,
the FSS factors indicated a mean value correlation of .33 and .42 with objective
and subjective measures of performance, respectively.
The correlations between challenge of the game and athlete skills with the
FSS factors and measures of performance (objective and subjective) are presented
in Table 2. An important issue of the correlational analysis was whether the value
of the correlations of athletes’ skills with flow experience was higher than those
between challenge of the competition and flow experience (Cohen, 1988; Meng,
Rosenthal, & Rubin, 1992; Steiger, 1980). Fisher’s z transformation was applied to
examine whether there were significant differences in the value of the correlations
of the two dimensions of the orthogonal model (challenge–skills) of flow experi-
ence and the FSS factors. The results indicated significantly higher correlations
between athletes’ skills and FSS subscales compared than between challenge of
the game and FSS subscales (Table 2).
In addition, Fisher’s z transformation showed a significant increase in the value
of correlations between athletes’ skills and flow experience as the time to competi-
tion approached (p < .05; 1 day before: rmean = .22, 1 hr before: rmean = .28, during
competition: rmean = .34). On the other hand, the correlations between perceived
challenge and flow experience remained rather stable (1 day before: rmean = .15, 1
hr before: rmean = .14, during competition: rmean = .18; Table 2).
Repeated-measure ANOVAs were conducted to examine the changes in per-
ceived challenge of the game and skills of the athletes among the three time measures
(1 day before, 1 hr before, and 1 hr during competition). The results indicated no
significant differences either for perceived challenge of the game, F(2, 218) = .168,
n.s., or athletes’ skills, F(2, 218) = 1.260, n.s.
PROOF
447
Table 2 Correlations Among Athletes’ Performance, Challenge of the Competition, Skills of the Athlete, and
Flow State Scale (FSS) Subscales
Challenge of the
Competition Skills of the Athlete
FSS subscale A B C A B C
Objective measure
of performance
Subjective measure
of performance
Challenge–skill
balance .29** .36** .35** .34** .41** .49**†.46** .63**‡
Action–awareness
merging –.02 –.01 –.00 .12 .20*†.25**†.24** .28**
Clear goals .28** .28** .32** .30** .36** .47**†.25** .31**
Unambiguous
feedback .13 .19* .18* .30**†.35**†.45**†.45** .49**
Concentration on task
at hand .10 .02 .11 .22** .28**†.33**†.29** .36**
Sense of control .16 .10 .15 .21* .28**†.33**†.41** .47**
Loss of self-
consciousness .05 .03 .06 .12 .17*†.25**†.19* .32**‡
Transformation of
time .03 .02 .08 –.11 –.15 –.08 –.02 –.03
Autotelic experience .21* .21* .30** .17 .23** .29** .52** .71**‡
Global flow score .21* .20* .26** .27* .35*†.46**†.48** .61**‡
rmean .15 .14 .18 .22 .28 .34 .33 .42
Subjective
performance .06 .10 .14 .02 .11 .18*
Objective
performance .16 .17* .14 .15 .25** .21*
Note. A = 1 day before the competition; B = 1 hr before the competition; C = During competition
*p < .01. **p < .001.
a Significantly higher correlation between “skills of the athlete” and FSS subscales than the “challenge of the competition” and FSS subscales (Fisher’s z transforma-
tion). b Significantly higher correlation between “subjective measure of performance” and FSS subscales compared to the “objective measure of performance” and
FSS subscales (Fisher’s z transformation).
PROOF
448 Stavrou et al.
Orthogonal Model of Flow Theory. Using median splits, athletes were divided
into low and high groups, with regard to the scales “challenge of situation” and
“skills of athlete.” Thus, based on the orthogonal model of flow, the athletes made
up four states: apathy (low challenge–low skills), anxiety (high challenge–low
skills), relaxation (low challenge–high skills), and flow (high challenge–high
skills). MANOVA results indicated significant differences between the four states
(Pillai’s trace = .355, F(3, 216) = 3.353, p < .001). Follow-up ANOVAs (Scheffé
test) on each dependent factor, applying Bonferroni adjustment, indicated significant
differences in the following factors: challenge–skill balance, F(3, 216) = 21.950,
p < .001, η2
p = .234; action–awareness merging F(3, 216) = 4.877, p < .01, η2
p =
.063; clear goals, F(3, 216) = 18.897, p < .001, η2
p =.208; unambiguous feedback,
F(3, 216) = 12.064, p < .001, η2
p = .144; concentration on task at hand, F(3, 216)
= 6.425, p < .001, η2
p = .082); sense of control, F(3, 216) = 7.267, p < .001, η2
p =
.092; loss of self-consciousness, F(3, 216) = 4.853, p < .01, η2
p = .063; and autotelic
experience, F(3, 216) = 8.653, p < .001, η2
p = .107. The flow and relaxation states
showed higher mean values in these FSS factors than the apathy and anxiety states,
which indicated lower values in flow experience (Table 3). Table 3 also provides
the differences between the four quadrants. Specifically, although flow state did
not differ significantly from the relaxation quadrant (except for challenge–skill
balance) there was a trend toward higher mean values in flow experience subscales.
On the other hand, flow state showed significantly higher values than apathy state
(except for action–awareness merging and transformation of time) and anxiety state
(except for concentration on task at hand and transformation of time). In addition,
the relaxation state revealed significantly higher mean values than the apathy state
on six of the nine FSS factors (challenge–skill balance, clear goals, unambiguous
feedback, concentration on task at hand, loss of self-consciousness, and autotelic
experience).
Table 3 Scores on the Flow State Scale Subscales Based on the
Orthogonal Model of Flow Theory, M (SD)
Flow State Scale subscale Apathy–1 Anxiety–2 Relaxation–3 Flow–4
Challenge–skill balancea,b,c,d 11.71 (2.78) 12.78 (3.16) 13.91 (3.91) 15.94 (2.73)
Action–awareness mergingb12.00 (3.02) 11.58 (3.06) 13.53 (3.70) 13.49 (3.32)
Clear goalsa,b,d 14.31 (2.83) 15.50 (2.06) 16.68 (2.83) 17.56 (2.48)
Unambiguous feedbacka,b,d 11.43 (3.01) 11.75 (3.66) 13.53 (4.30) 14.71 (3.07)
Concentration on task at handa,d 13.66 (3.58) 13.98 (3.48) 15.70 (3.73) 15.87 (3.01)
Sense of controla,b 13.00 (2.99) 13.60 (3.29) 14.77 (3.98) 15.51 (3.12)
Loss of self-consciousnessa,d 12.46 (3.96) 13.25 (4.11) 14.96 (3.54) 14.37 (3.69)
Transformation of time 11.35 (3.18) 11.83 (3.51) 10.96 (3.75) 11.44 (3.81)
Autotelic experiencea,b,d 11.48 (4.32) 12.45 (5.39) 12.92 (5.07) 15.44 (4.16)
a Group 4 significantly higher than Group 1. b Group 4 significantly higher than Group 2. c Group 4 significantly
higher than Group 3. dGroup 3 significantly higher than Group 1.
PROOF
Flow Experience and Athletes’ Performance 449
Performance Measures
No significant differences were found between the four states (apathy, anxiety,
relaxation, and flow) on either the subjective measure of performance F(3, 216) =
1.300, n.s., or the objective measure of performance F(3, 216) = 2.769, n.s. Sig-
nificant positive correlations were found between the FSS factors and performance
measures, both subjective and objective (Table 2).
Prediction-of-Performance Measures
Standard multiple-regression analyses (Tabachnick & Fidell, 2006) were performed
to examine the contribution of each of the FSS factors in the prediction of athletes’
performance. Two standard multiple-regression analyses were conducted in which
the FSS subscales served as the predictor variables, with the objective and subjective
measures of performance as the dependent variable in the two separate regression
analyses. Table 4 displays the unstandardized regression coefficients (B), standard
error (SE B), standardized regression coefficients (β) t values, and the level of
significance (p; Cohen et al., 2003; Tabachnick & Fidell). There was no evidence
of multicollinearity among the independent variables. Regarding the subjective
Table 4 Standard Multiple-Regression Analysis on Subjective and
Objective Performance Measures
Dependent variable
B SE B βt
Subjective performance measure
challenge–skill balance .23 .07 .29 3.33**
action–awareness merging .06 .05 .07 1.23
clear goals .16 .07 .17 2.38
unambiguous feedback –.02 .05 –.02 –0.29
concentration on task at hand .01 .06 .01 0.16
sense of control .06 .07 .08 0.94
loss of self-consciousness –.04 .04 –.06 –1.06
transformation of time –.02 .04 –.02 –0.39
autotelic experience .30 .05 .53 6.61**
Objective performance measure
challenge–skill balance .25 .30 .09 0.83
action–awareness merging .25 .20 .09 1.27
clear goals .45 .29 .13 1.54
unambiguous feedback .43 .23 .16 1.84
concentration on task at hand –.07 .24 –.03 –0.30
sense of control .46 .28 .16 1.61
loss of self-consciousness –.35 .17 –.14 –2.01
transformation of time –.00 .17 .00 –0.01
autotelic experience .77 .20 .37 3.82**
**p < .001.
PROOF
450 Stavrou et al.
measure of performance, the R for regression was significantly different from zero,
F(9, 210) = 27.403, p < .001, and the values of R2, as well as adjusted R2, were
.54 and .52, respectively. The significant predictors were autotelic experience (β =
.53) and challenge–skill balance (β = .29). The second regression analysis, using
the FSS factors as the independent factors and objective measure of performance
as the dependent variable, was also significant, F(9, 210) = 11.319, p < .001. The
value of R2 was .33, and the total explained variance for the objective measure of
performance was 30%. The significant predictor was autotelic experience (β = .37),
which predicted 33% (30% adjusted) of the variability of the objective measure
of performance.
Discussion
Examining Athletes’ Flow Experience
The results of this study indicated that the quality of experience between the four
quadrants of the orthogonal model are differentiated. Specifically, the athletes in
the flow and relaxation states showed the highest FSS factor scores, the athletes in
the apathy state showed the lowest values, and the scores in the anxiety state were
between flow and apathy. In general, the results showed that the athletes in flow and
relaxation states experienced significantly higher flow characteristics, as assessed
by the FSS, than athletes in the apathy state. In addition, significant differences
were revealed between the athletes on apathy and anxiety, compared with flow
state, with the latter having significantly higher positive experience. Furthermore,
athletes in the relaxation state indicated significantly more positive experience than
those in the apathy and anxiety states. This finding suggests that it is the athletes’
skill level that is more important than the perceived challenge of the situation for
getting into flow in a competitive environment. This result extends the research in
flow theory, showing that the perception of skillfulness is essential to experience
positive mental states, whereas challenge might have more of a secondary role in
athletes’ flow experience.
The importance of athletes’ skills on the quality of their experience is further
supported by the significantly higher correlations between the FSS factors and
perceived skills, suggesting a close relationship, than the correlations between the
FSS factors and perceived challenges. On the other hand, the lack of relationship
between the challenge ratings and the flow subscales suggests that this variable
might not be as relevant to flow experiences as one’s perception of his or her skills
in a competitive environment. Providing support to this notion, Stein and colleagues
(1995) reported that challenge was related to the quality of experience only in a
recreational-athletic sample, not in a competitive one.
Regarding the time-to-competition measures of athletes’ skills and challenge
of the situation, the value of the correlations between the skills and the FSS sub-
scales increased as time for competition neared, whereas the correlations between
challenge of the situation and flow remained stable across the three time measures.
The correlational-analysis results suggest that, during competition, a skillful athlete
will feel that he or she can meet the challenge of the game, have clear goals, and
receive immediate and unambiguous feedback about how well he or she performs.
In addition, elevated skills seem to have a positive relation to other qualities of
PROOF
Flow Experience and Athletes’ Performance 451
experience, such as concentration and sense of control over the activity. Further-
more, the moderate positive correlations between challenge of competition with
challenge–skill balance, autotelic experience, and clear goals suggest that a chal-
lenging competition will help athletes have an intrinsically enjoyable experience
and set clear goals. The remaining flow dimensions did not indicate significant
correlations with challenge of the game.
The value of the correlations of the challenge–skills ratings with the FSS factors
ranged from low to medium, indicating that the quality of experience is affected not
by the level of challenge and skills per se but might depend on the relative balance
or imbalance between the two scales. Moneta and Csikszentmihalyi (1996) reported
that the effect of challenges became positive for higher values of skills, whereas
the negative effect of challenges became smaller as skills increased. In this study,
athletes’ skills revealed significant higher positive correlations—compared with the
challenge of the competition—with the challenge–skill-balance dimension, indicat-
ing that in a competitive environment the athletes who estimate that they have the
abilities to manage the demands of the game might be more likely to experience a
balance of challenge and skills, even when the challenge is fairly high.
Based on the real-time-measure results of this study, the perceived skills of the
athlete might be the primary factor for him or her to experience positive or optimal
mental states, whereas perceived challenge seems to have a facilitative role only
when athletes’ skills are sufficiently high. From a practical point of view, coaches,
athletes, and sport psychology consultants should focus on how skillful an athlete
feels during a game, as well as on factors related to perceived skillfulness, such as
self-confidence and perceived competence. Psychological-preparation programs
could include positive thinking, self-talk, and goal-setting interventions to increase
the level of athletes’ perceived skills.
Challenge–Skills Ratings
According to the results of the present study, the level of perceived challenge of the
competition and athletes’ skills did not change across the three times of assessment
(1 day before, 1 hr before, and 1 hr during competition). This suggests that how an
athlete estimates his or her skills for the upcoming performance, as well as how
challenging the competition is perceived to be, seems to remain stable close to,
just before, and during the competition.
The level of perceived challenge of the game refers to a cognitive evaluation,
which is based on, or modulated by, athletes’ expectations or goals, as well as
their perceived ability to manage competition demands. On the other hand, how
skillful an athlete feels about an upcoming performance is based on an evaluation
of his or her efficiency and skills, especially in relation to the level of difficulty of
the competition. An unanticipated event, such as an athlete becoming injured, or
a difficult opponent not participating at the competition could be associated with
change in perceived challenges and skills. A challenging competition will become
either frightening (if athletes feel that they cannot manage the demands of the
situation because of an unexpected injury) or boring (if perceived skills outweigh
the difficulty of the opponents). In relation to the four experiential states of the
orthogonal model of flow theory, this finding suggests that an athlete’s experience
(i.e., flow, relaxation, boredom, or anxiety), might depend on how she or he feels
PROOF
452 Stavrou et al.
at a certain time before the event (e.g., 1 day before the competition) and not just
directly before the competition.
Little or no correlation was found between challenge of the competition and
athletes’ skills with performance measures, both objective and subjective. The lack
of correlations between challenge and skills with athletes’ performance indicated
that these ratings did not have a direct and close relationship with the level of
performance, suggesting that other psychological or environmental characteristics
might be mediated in these relationships. Taking into consideration the independent
relationship, as well as the ratio between challenge and skills, it might be that situ-
ational challenge and skills, as well as the balance between them, formulate athletes’
flow experience, which in turn might affect the level of their performance.
Correlations of FSS Subscales
The FSS subscales indicated a wide range of intercorrelations. Low (r = .27) to
high (r = .75) intercorrelations were found among most of the FSS factors. The
exception was transformation of time, which had nonsignificant correlations with
the other factors. With regard to the value of the correlations, some of the factors
revealed closer relationships than others. Specifically, challenge–skill balance,
clear goals, unambiguous feedback, concentration on task at hand, sense of control,
and autotelic experience indicated the higher interfactor correlations, suggesting
a close relationship among these factors in this sample (shared variance 35%). If
these results can be generalized, then being in flow might be associated with high
values in all six of these FSS subscales.
The high positive correlation between challenge–skill balance and autotelic
experience indicates that enjoyment of the activity is related to whether an athlete
feels fairly skillful to meet the demands of the competition. That is, the subjective
ratio or balance between challenge of the situation and athletes’ skills might be
critical to the autotelic-experience dimension. The high positive correlations among
autotelic experience, sense of control, and unambiguous feedback suggest that the
enjoyment of the competition is associated with an athletes having a strong sense
of control over actions, so as to have the desired result, and clear feedback of how
well he or she performs. In addition, close relationships were found between chal-
lenge–skill balance, clear goals, and unambiguous feedback. These three dimen-
sions can be thought of as “setting the stage for flow” (Jackson, 1996). In other
words, perhaps these dimensions modulate the rest of flow-experience qualities and
represent the preconditions to get into flow, as has been suggested by Nakamura
and Csikszentmihalyi (2002).
The high correlation between challenge–skill balance and sense of control
indicated that the perceived balance between the challenge of the game and the
efficiency of athletes to meet the demands of the task is related to a sense of control
over their efforts and performance. This will help athletes concentrate on the task
at hand, eliminating irrelevant cues or errors, which also finds support in the high
positive correlation between concentration and sense of control. Furthermore, chal-
lenge–skill balance, sense of control, and autotelic experience were closely related
to the feedback given during the activity. This means that athletes’ sense of control,
concentration on the performing activity, and perceptions about skills to manage
the demands of the situation would provide them the opportunity to have a strong
PROOF
Flow Experience and Athletes’ Performance 453
and immediate sense of performance quality. In other words, if athletes feel that
they can manage the demands of the competition and enjoy the activity, this will
be associated with having clear information about their performance.
On the other hand, action–awareness merging and loss of self-consciousness
showed low to moderate positive correlations (r mean = .36, range .27–.47) with
the rest of the FSS factors. The highest correlation among this set of intercorrela-
tions was between loss of self-consciousness and autotelic experience, suggesting
that when athletes are free of worry about others’ evaluations, this is linked with
enjoying the situation. The relatively lower intercorrelations with these two dimen-
sions indicates that action–awareness merging and loss of self-consciousness were
more independent from the set of more strongly related flow dimensions in this
sample of athletes. Further research is needed to examine how these two factors are
experienced when an athlete is in flow state and whether the experience depends on
particular characteristics of the sample, such as competitive experience (amateur vs.
high level), type of sport (sports and physical activities), and level of participation
(recreational to national level).
The lack of correlations that occurred between the FSS subscales and trans-
formation of time indicated that the alteration of the sense-of-time factor was not
relevant to the sport-flow experience of the athletes in this sample. Low associa-
tions have consistently been found with time transformation in athlete samples
(e.g., Jackson & Marsh, 1996; Jackson et al., 1998, 2001). From a practical point
of view, the occurrence and the positive evaluation of transformation of time
might depend on specific sport requirements. For example, whereas in some sports
(e.g., long jump, triple jump, shot put) sense of time is not important, it seems to
be essential in sports such as running, cycling, swimming, archery, and shoot-
ing, constituting a checkpoint of athletes’ performance. Cyclists, swimmers, and
long-distance runners should cover the intermediate distances in specific times,
using these times as a mark of their performance or using their tempo as a way to
get feedback, preserving themselves from disorientation of time. In addition, the
sense of loss of time requires, or presupposes, a high level of implementation and
good technical performance, which might only be experienced by elite athletes in
favorable conditions.
Performance Measures and Flow Experience
In comparison with the other flow dimensions, autotelic experience and chal-
lenge–skill balance indicated higher positive correlations with athletes’ perfor-
mance, which is in agreement with Jackson and colleagues’ (1998, 2001) research
findings. Specifically, the results of the current study indicated that high levels of
performance are enjoyable in nature. In high levels of performance, athletes enjoy
the moment, providing support to the notion that best performance shares common
characteristics with flow experience. In addition, a crucial issue for high levels of
performance is whether an athlete participating in a challenging activity feels that
he or she can manage the requirements of the game, achieving the desired result.
The positive correlation that occurred between unambiguous feedback and perfor-
mance indicates that during high levels of performance athletes have a strong sense
of how well they perform. This gives them the opportunity to recognize errors and
make corrections. Feedback about correct performance motivates athletes to keep
PROOF
454 Stavrou et al.
trying, whereas error-related information works better to facilitate skill acquisi-
tion (Magill, 2004). In addition, the positive correlation between sense of control
and performance suggests that an athlete who feels in control also performs at a
high level.
Regression analysis indicated that more than half the variability of subjective
performance was predicted by two FSS factors, namely, autotelic experience and
challenge–skill balance. Regression analyses showed that autotelic experience was
the most significant predictor of athletes’ performance. Thus, if an athlete enjoys
the activity, this is associated with achieving high levels of performance. In the case
of the subjective measure of performance, challenge–skill balance was the next
most significant predictor of performance. In other words, the balance between the
perceived demands of the competition and perceived skill level seems important for
athletes’ self-reported measures of performance. Relative to the regression analysis
of the objective measure of performance, autotelic experience predicted almost a
third of the variability of athletes’ performance. The pattern of results across the
two measures of performance demonstrated that autotelic experience is the most
significant predictor of the flow dimensions for the athletes in this sample. In other
words, if an athlete really enjoys his or her experience, this might be helpful in
the upcoming performance. In qualitative research with athletes, Jackson (1996)
found that autotelic experience was the most highly endorsed factor when athletes
were describing their flow experiences.
The current study provided interesting and useful findings for flow theory,
which is a well-accepted area in psychology, although there are certain limitations
that should not be overlooked. First, we examined flow experience from a quantita-
tive perspective. Trying to quantify athletes’ flow experience has certain limitations,
because it cannot portray the subjective nature of flow phenomenon. In addition, the
FSS has the same limitations as all self-report instruments that quantify athletes’
experience. Although self-report instruments provide substantial information, they
are limited in the extent to which they can tap into the subjective nature of athletes’
experience during sport participation. As Csikszentmihalyi (1992) and Jackson and
Marsh (1996) have mentioned, the content of flow cannot be perfectly assessed by
a score on a questionnaire. Researchers should take into consideration the complex
nature of flow experience. Using various instruments and research methods will be
helpful to capture, understand, and interpret the experience of flow from an athlete’s
perspective. For example, the experience sampling method (Csiksentmihalyi &
Csikszentmihalyi, 1988) could be useful in providing crucial information in the
assessment and understanding of flow experience as it occurs during a sport activ-
ity. Finally, the sample of the present study consisted only of athletes of individual
sports. Using team-sport athletes, flow experience might be different. In addition,
because participants of this study were competitive athletes, the results might not
be generalizable to participants in recreational or leisure activities.
Future research is needed to examine flow experience in a variety of sport
activities (i.e., individual vs. team sports), levels of athlete (amateur vs. high
level), environments (i.e., leisure, everyday activities vs. competitive sports), and
the temporal relationships between competition and the formulation of challenge
and skills. This will aid understanding of when these perceptions are formed and
how stable they are over time. A measure of athletes’ skills and challenge of the
PROOF
Flow Experience and Athletes’ Performance 455
competition 1 week or more before competition will provide fruitful information
regarding the dimensions of the orthogonal model of flow.
In addition, the perception of competition challenge and athletes’ skills, as well
as the relative balance between challenge and skills, in different sport contexts needs
further examination. Moreover, interindividual and intraindividual differences need
to be examined to understand potential influences on challenge and skills.
The current study examined flow only in one competition. To better capture
flow experience, it would be interesting to examine how an athlete reacts in various
competitive conditions (e.g., two or three competitions). For this purpose, qualitative
research, using interviews, would be useful for enlightening the content of competi-
tion flow experiences. A qualitative approach can provide important information
regarding the subjective nature of flow experience and how athletes formulate
flow in a sport setting. Furthermore, qualitative methods (e.g., interpretive phe-
nomenological analysis, in-depth interviews) can overcome quantitative restraints,
providing information, interpretation, and understanding of flow phenomena from
an athlete’s point of view (Sparkes & Partington, 2003).
The current study had the advantage of examining flow in real-time measure,
before and during competition. Although no causality can be implied from this
study, it does seem from this, and other flow research, that the experience of flow
can have an important positive influence on performance. Athletes’ skills might
be the critical factor for attaining flow in competitive sport, because athletes in
relaxation and flow states revealed the most positive experiential characteristics.
Challenge and skills seem to be formulated before competition and remain stable
close to the game. The results of the present study provide important information
to coaches and sport psychology consultants, and might be of use in formulating
psychological preparation programs that will foster the experience of flow, and, in
turn, facilitate athlete’s performance.
References
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale,
NJ: Lawrence Erlbaum.
Cohen, J., Cohen, P., West, S.G., & Aiken, L.S. (2003). Applied multiple regression/correla-
tion analysis for the behavioral sciences (3rd ed.). Mahwah, NJ: Lawrence Erlbaum.
Csikszentmihalyi, M. (1975). Beyond boredom and anxiety. San Francisco: Jossey-Bass.
Csikszentmihalyi, M. (1982). Towards a psychology of optimal experience. In L. Wheeler
(Ed.), Annual review of personality and social psychology (Vol. 3, pp. 13-36). Beverly
Hills, CA: Sage.
Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. New York:
Harper & Row.
Csikszentmihalyi, M. (1992). A response to the Kimiecik & Stein and Jackson papers.
Journal of Applied Sport Psychology, 4, 181-183.
Csikszentmihalyi, M., & Csikszentmihalyi, I. (1988). Optimal experience: Psychological
studies of flow in consciousness. New York: Cambridge University Press.
Csikszentmihalyi, M., & LeFevre, J. (1989). Optimal experience in work and leisure. Journal
of Personality and Social Psychology, 56, 815-822.
Gould, D., Eklund, R.C., & Jackson, S.A. (1992). 1988 US Olympic wrestling excellence
I: Mental preparation, precompetitive cognition, and affect. The Sport Psychologist,
6, 358-382.
PROOF
456 Stavrou et al.
Jackson, S.A. (1988). Positive performance states of athletes: Toward a conceptual under-
standing of peak performance. Unpublished master’s thesis, University of Illinois at
Urbana–Champaign.
Jackson, S.A. (1992). Athletes in flow: A qualitative investigation of flow states in elite
figure skaters. Journal of Applied Sport Psychology, 4, 161-180.
Jackson, S.A. (1995). Factors influencing the occurrence of flow state in elite athletes.
Journal of Applied Sport Psychology, 7, 135-163.
Jackson, S.A. (1996). Toward a conceptual understanding of the flow experience in elite
athletes. Research Quarterly for Exercise and Sport, 67, 76-90.
Jackson, S.A. (1999). Joy, fun, and flow state in sport. In Y.L. Hanin (Ed.), Emotions in sport
(pp. 135-155). Champaign, IL: Human Kinetics.
Jackson, S.A., & Csikszentmihalyi, M. (1999). Flow in sports: The keys to optimal experi-
ences and performances. Champaign, IL: Human Kinetics.
Jackson, S.A., & Eklund, R.C. (2002). Assessing flow in physical activity: The Flow State
Scale-2 and the Dispositional Flow Scale-2. Journal of Sport & Exercise Psychology,
24, 133-150.
Jackson, S.A., Kimiecik, J.C., Ford, S., & Marsh, H.W. (1998). Psychological correlates of
flow in sport. Journal of Sport & Exercise Psychology, 20, 358-278.
Jackson, S.A., & Marsh, H.W. (1996). Development and validation of a scale to measure
optimal experience: The Flow State Scale. Journal of Sport & Exercise Psychology,
18, 17-35.
Jackson, S.A., & Roberts, G.C. (1992). Positive performance states of athletes: Toward a
conceptual understanding of peak performance. The Sport Psychologist, 6, 156-171.
Jackson, S.A., Thomas, P.R., Marsh, H.W., & Smethurst, C.J. (2001). Relationship between
flow, self-concept, psychological skill, and performance. Journal of Applied Sport
Psychology, 13, 129-135.
Kleiber, D., Larson, R.G., & Csikszentmihalyi, M. (1986). The experience of leisure in
adolescents. Journal of Leisure Research, 18, 169-176.
Kimiecik, J.C., & Stein, G.L. (1992). Examining flow experience in sport contexts: Con-
ceptual issues and methodological concerns. Journal of Applied Sport Psychology, 4,
144-160.
Lane, A.M., & Karageorgis, C.I. (1997). Discriminant effectiveness of goal confidence and
goal difficulty in predicting goal attainment in junior high school cross-country runners.
Perceptual and Motor Skills, 84, 747-752.
Lane, A.M., Terry, P.C., Beedie, C.J., Curry, D.A., & Clark, N. (2001). Mood and perfor-
mance: Test of a conceptual model with a focus on depressed mood. Psychology of
Sport and Exercise, 2, 157-172.
Magill, R.A. (2004). Motor learning and control: Concepts and applications. New York:
McGraw-Hill.
Martin, J.J., & Gill, D.L. (1991). The relationships among competitive sport orientation,
sport-confidence, self-efficacy, anxiety, and sport performance. Journal of Sport &
Exercise Psychology, 13, 149-159.
Martin, J.J., & Gill, D.L. (1995). The relationships among competitive orientations and
self-efficacy to goal importance, thoughts, and performance in high school distance
runners. Journal of Applied Sport Psychology, 7, 50-62.
Meng, X.R., Rosenthal, R., & Rubin, D. (1992). Comparing correlated correlation coef-
ficients. Psychological Bulletin, 111, 172-175.
Moneta, G.B., & Csikszentmihalyi, M. (1996). The effect of perceived challenges and skills
on the quality of subjective experience. Journal of Personality, 64, 275-310.
Nakamura, J., & Csikszentmihalyi, M. (2002). The concept of flow. In C.R. Snyder &
S.J. Lopez (Eds.), Handbook of positive psychology (pp. 89-105). New York: Oxford
University Press.
PROOF
Flow Experience and Athletes’ Performance 457
Privette, G. (1983). Peak experience, peak performance, and flow: A comparative analysis
of positive human experiences. Journal of Personality and Social Psychology, 45,
1361-1368.
Privette, G., & Bundrick, C.M. (1991). Peak experience, peak performance, and flow:
Personal descriptions and theoretical constructs. Journal of Social Behavior and Per-
sonality, 6, 169-188.
Privette, G., & Bundrick, C.M. (1997). Psychological processes of peak, average, and failing
performance in sport. International Journal of Sport Psychology, 28, 323-334.
Ravizza, K. (1977). Peak experience in sport. Journal of Humanistic Psychology, 17, 35-
40.
Sparkes, A.C., & Partington, S. (2003). Narrative practice and its potential contribution to
sport psychology: The example of flow. The Sport Psychologist, 17, 292-317.
Stavrou, N.Α., & Zervas, Y. (2004). Confirmatory factor analysis of the Flow State Scale in
sports. International Journal of Sport and Exercise Psychology, 2, 161-181.
Steiger, J.H. (1980). Tests for comparing elements of a correlation matrix. Psychological
Bulletin, 87, 245-251.
Stein, T.K., Kimiecik, J.C., Daniels, J., & Jackson, S.A. (1995). Psychological antecedents
of flow in recreational sport. Personality and Social Psychological Bulletin, 21, 125-
135.
Tabachnick, B.G., & Fidell, L.S. (2006). Using multivariate statistics (5th ed.). New York:
Pearson Education.
Vlachopoulos, S.P., Karageorghis, C.I., & Terry, P.C. (2000). Hierarchical confirmatory factor
analysis of the Flow State Scale in exercise. Journal of Sports Sciences, 18, 815-823.
Manuscript submitted: October 4, 2006
Revision received: July 12, 2007
PROOF
