Content uploaded by Antonis Hatzigeorgiadis
Author content
All content in this area was uploaded by Antonis Hatzigeorgiadis on Dec 07, 2018
Content may be subject to copyright.
Mechanisms underlying the self-talk–performance relationship: The effects
of motivational self-talk on self-confidence and anxiety
Antonis Hatzigeorgiadis
*
, Nikos Zourbanos, Sofia Mpoumpaki, Yannis Theodorakis
Department of Physical Education and Sport Sciences, University of Thessaly, Trikala 42100, Greece
article info
Article history:
Received 31 January 2008
Received in revised form 20 July 2008
Accepted 28 July 2008
Available online 3 August 2008
Keywords:
Self-talk functions
Self-confidence
Anxiety
Task performance
Tennis
abstract
Objectives:
The present study examined the effects of motivational self-talk on self-confidence, anxiety,
and task performance in young athletes.
Methods:
Participants were 72 tennis players. The experiment was conducted in five sessions: baseline
assessment, three training sessions, and final assessment. After the baseline assessment participants
were divided and assigned randomly into experimental and control groups. The two groups followed the
same training program with the experimental group practicing the use of self-talk. In the last session,
the final assessment took place. A forehand drive test was used to evaluate task performance, and the
Competitive Anxiety Inventory-2R was used to assess self-confidence and anxiety.
Results:
A two-way mixed model MANOVA revealed that task performance improved for the experi-
mental group (p<.01) and remained stable for the control group; self-confidence increased (p<.01) and
cognitive anxiety decreased (p<.05) for the experimental group, whereas no changes were observed for
the control group. Correlation analysis revealed that changes in task performance were moderately
related to changes in self-confidence (p<.05).
Conclusions:
The results of the study showed that self-talk can enhance self-confidence and reduce
cognitive anxiety. Furthermore, it is suggested that increases in self-confidence can be regarded as
a viable function explaining the facilitating effects of self-talk on performance.
!2008 Elsevier Ltd. All rights reserved.
Introduction
Self-talk has been central in cognitive behavioural modification
(Meichenbaum, 1977). Based on the principle that what people say
to themselves affects the way they behave (Ellis, 1976), strategies
involving mental processes have been developed to regulate
cognitions and develop or change existing thought patterns. The
use of self-talk plans to control and organize athletes’ thoughts has
been promoted as a key component for successful sport perfor-
mance, and self-talk is frequently included as an integral part of
psychological skill training (Hardy, Jones, & Gould, 1996). Accord-
ingly, sport research regarding the use and effectiveness of self-talk
has received considerable attention in recent years. Research
adopting various designs (e.g. experimental, intervention and
single-subject designs) in a variety of sports and tasks has sup-
ported the effectiveness of the self-talk strategy in facilitating
learning and improving task performance (Zinnser, Bunker, &
Williams, 2006).
Research has progressively moved towards the identification of
the functions underlying the effectiveness of self-talk, that is the
mechanisms through which self-talk affects performance (Hardy,
2006). Johnson, Hrycaiko, Johnson, and Hallas (2004) suggested
that the core of self-talk is that focusing on the desired thought
leads to the desired behaviour. In other words, ST is an instruction
to initiate or perform an action or a sequence of actions. Several
explanations have been provided regarding the facilitating effects
of self-talk on performance. Landin (1994) and Nideffer (1993)
supported an attentional interpretation of the self-talk effects.
Landin proposed that self-talk can be used to enhance attentional
focus, whereas Nideffer indicated that self-talk can be an effective
strategy for directing or redirecting attention to task relevant cues.
Finn (1985) and Zinnser et al. (2006) suggested that self-talk can
serve to regulate effort and enhance self-confidence, whereas
Hardy et al. (1996) argued that self-talk can also be effective in
controlling anxiety and triggering appropriate action.
Hardy, Gammage, and Hall (2001) in a qualitative descriptive
inquiry, based on Paivio’s (1985) conceptualisation regarding the
functions of imagery, identified two broad functions of self-talk,
cognitive and motivational. They suggested that these two general
functions can be further broken down into more specific lower
order functions. Accordingly, the motivational function comprises
*Corresponding author. Tel.: þ30 24310 47009.
E-mail address: ahatzi@pe.uth.gr (A. Hatzigeorgiadis).
Contents lists available at ScienceDirect
Psychology of Sport and Exercise
journal homepage: www.elsevier.com/locate/psychsport
1469-0292/$ – see front matter !2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.psychsport.2008.07.009
Psychology of Sport and Exercise 10 (2009) 186–192
a motivational arousal function (referring to psyching-up, relaxa-
tion, and arousal control), a motivational mastery function (refer-
ring to mental toughness, confidence and mental preparation), and
a motivational drive function (referring to regulating drive and
effort). Similarly, the cognitive function comprises a cognitive
specific function (referring to skill learning and development) and
a cognitive general function (referring to strategy and performance
enhancement). Considering Hardy et al.’s approach, Zervas, Stav-
rou, and Psychountaki (2007) developed an instrument assessing
the two broad cognitive and motivational functions. The authors
created a pool of items assessing the two dimensions (cognitive, e.g.
I talk to myself to give directions; motivational, e.g. I talk to myself
to increase motivation), and supported the factorial validity and the
reliability of the instrument. On concluding, they identified that
further developments should consider the role of self-talk in
regulating specific psychological aspects of performance, such as
self-confidence, mood, anxiety control and effort, as possible
functions of self-talk. Finally, Theodorakis, Hatzigeorgiadis, and
Chroni (2008) based on empirical evidence and raw data generated
through athletes’ reports further examined the functions of self-
talk. Content analysis and a series of exploratory and confirmatory
factor analyses identified five distinct functions of self-talk. In
particular, they suggested that self-talk can serve to enhance
attentional focus, increase self-confidence, regulate effort, control
cognitive and emotional reactions, and trigger automatic execution,
and provided evidence regarding the psychometric properties of
the questionnaire.
Preliminary evidence regarding the speculated effects of self-talk
has been provided through studies examining the effectiveness of
self-talk in a variety of tasks and settings, and through athletes’ post-
experimental reports. Van Raalte, Brewer, Rivera, and Petipas (1994)
asked young tennis players after the conclusion of competitive
matches to report their self-talk and how they thought their self-talk
affected their performance. Participants reported that positive self-
talk helped them concentrate and enhanced their motivation.
Landin and Hebert (1999) implemented a self-talk strategy aiming at
improving volleying skills in collegiate tennis players. Participants
reported that self-talk helped them feel more confident and direct
their attention more efficiently. Perkos, Theodorakis, and Chroni
(2002) administered a 12-week self-talk training program in young
basketball players and found that the use of self-talk improved
players’ dribbling and passing performance. In a post-experimental
short questionnaire participants indicated that the use of self-talk
improved their concentration and self-confidence. Thelwell and
Greenlees (2003) implemented a psychological skills training
program to four recreational athletes competing at a laboratory-
based triathlon task. The results revealed that participants’ perfor-
mance at the task improved across trials. Participants perceived that
self-talk increased their motivation and self-confidence and
enhanced their attentional focus. Finally, Johnson et al. (2004) using
a single-subject multiple baseline design, tested the effectiveness of
a self-talk intervention program in female football players, assessing
performance in the low drive shot over a period of three months.
Their results showed that shooting performance improvedfor two of
the three participants, whereas all three participants reported
increased self-confidence compared to baseline.
An experiment investigating the attentional function of self-talk
was conducted by Hatzigeorgiadis, Theodorakis, and Zourbanos
(2004). They assessed performance and occurrence of interfering
thoughts during performance in two experimental water-polo
tasks. Their results revealed that task performance improved and
interfering thoughts were reduced for participants using self-talk,
whereas no differences were recorded for a control group. The
authors also reported that increases in task performance were
related to decreases in interfering thoughts, and suggested that
performance enhancement could be attributed to the reduction of
interfering thoughts, even though clear inferences regarding the
causality could not be claimed.
Another line of research providing indirect evidence that self-
talk may serve several functions involves the investigation of the
effects different types of self-talk have on performance. Research on
the effectiveness of self-talk has examined and compared the
effects of different types of self-talk in experimental tasks. Theo-
dorakis,Weinberg, Natsis, Douma, and Kazakas (2000) speculated
that instructional self-talk should be more beneficial for fine tasks
(tasks placing greater demands on accuracy and precision),
whereas motivational self-talk should be more beneficial for gross
tasks (tasks placing greater emphasis on strength and endurance).
Subsequently they examined the effectiveness of motivational and
instructional self-talk in four experimental tasks, which were
characterised as fine (passing accuracy in football and serving
accuracy in badminton) or gross (3-min sit-up test and knee
extension power test). The results revealed that instructional self-
talk improved the performance for the two accuracy tasks and the
knee extension task, whereas motivational self-talk improved
performance for the knee extension task only. In a similar experi-
ment, Hatzigeorgiadis et al. (2004) tested the effectiveness of self-
talk in a precision and a power water-polo task. Instructional
self-talk improved performance for the precision task more than
motivational self-talk, whereas only motivational self-talk
improved performance for the power task. In general, even though
the evidence is not conclusive, these findings suggest that different
types of self-talk may have different effects on task performance
based on the nature of the task and the type of self-talk that is used.
Stemming from such findings, Hatzigeorgiadis, Zourbanos, and
Theodorakis (2007) suggested that if different self-talk cues have
different performance effects, different types of self-talk should
serve different functions. Zinsser et al. (2006) claimed that
instructional self-talk should be effective in enhancing attentional
focus and directing attention, whereas motivational self-talk
should be more effective in enhancing motivation, building self-
confidence and regulating effort. Two studies have examined
whether different types of self-talk serve different functions. Hat-
zigeorgiadis (2006) examined participants’ perceptions regarding
the use of instructional and motivational self-talk after imple-
menting a three-day self-talk training program in a swimming task.
According to participants’ perceptions both types of self-talk
mainly helped them to improve their attention to the task.
Furthermore, participants reported that the motivational self-talk
cue had greater impact on effort than the instructional self-talk cue.
In a similar experiment, Hatzigeorgiadis et al. (2007), in addition to
participants’ perceptions regarding the facilitative effects of self-
talk, examined self-confidence, anxiety symptoms and frequency of
interfering thoughts in the baseline (no self-talk) and experimental
trials (instructional and motivational self-talk). The results revealed
that the motivational self-talk cue was more effective in reducing
anxiety than the instructional self-talk cue. Furthermore, partici-
pants reported that the use of both cues mainly helped them
concentrate better on the task. The authors concluded that the
effectiveness of self-talk can be attributed mainly to its attention
function, at least in the case of novel tasks, but also that motiva-
tional self-talk is more effective in reducing anxiety than instruc-
tional self-talk. Subsequently they suggested that self-talk content
moderates self-talk functions, that is, different types of self-talk
may serve different functions depending on the content of the self-
talk cues. Nevertheless, it should be noted that in the above
experiments (a) students were recruited and not athletes, and (b)
within subjects differences were examined in the use of different
self-talk types, without the use of control groups.
The purpose of the present study was to examine whether the
use of motivational self-talk can increase self-confidence, reduce
anxiety and enhance task performance in athletes. The beneficial
A. Hatzigeorgiadis et al. / Psychology of Sport and Exercise 10 (2009) 186–192 187
effects of self-talk on task performance are well documented in the
literature (Zinsser et al., 2006). Furthermore, preliminary evidence
suggests that self-talk may serve to increase self-confidence in
athletes (Johnson et al., 2004; Landin & Hebert, 1999; Perkos et al.,
2002). Sport anxiety theory and research has provided evidence
regarding the relationship between self-confidence and anxiety, in
particular cognitive. Martens, Burton, Vealey, Bump, and Smith
(1990) supported the negative relationship between cognitive
anxiety and self-confidence and characterised self-confidence as
the relative absence of cognitive anxiety. Even though the inter-
dependence of the two constructs has been criticised on the
evidence of different relationships between the two and third
variables (Woodman & Hardy, 2003), a moderate negative rela-
tionship between the two has been consistently supported in the
literature. This led us expect that the hypothesised facilitative
effects of self-talk on self-confidence will apply for anxiety as well,
at least for cognitive. Further evidence that cognitive strategies can
reduce competitive anxiety has been provided by Maynard, Smith,
and Warwick-Evans (1995) and Maynard, Hemmings, Greenlees,
Warwick-Evans, and Stanton (1998). In these studies, cognitive
intervention programs involving positive thought control were
implemented. The results showed that the interventions were
effective in reducing competitive anxiety, in particular cognitive.
It has been suggested that motivational self-talk can have
greater impact on motivational-related outcomes, such as effort,
self-confidence, and anxiety (Zinsser et al., 2006), and preliminary
evidence seems to support this hypothesis (Hatzigeorgiadis et al.,
2007). As the primary purpose was to examine the effects of self-
talk on self-confidence and anxiety, the use of motivational self-talk
was preferred. Based on the preliminary evidence and the above
assumptions, it was hypothesised that the use of motivational self-
talk (a) will enhance performance, (b) will increase self-confidence
and reduce cognitive anxiety, whereas no predictions were made
regarding somatic anxiety.
Method
Participants
Participants were 72 (36 males and 36 females) competitive
young tennis players (mean age 13.47"1.78). They were recruited
from three tennis clubs situated in the midlands of Greece. Partic-
ipants had been training systematically for 4.10 ("2.32) years and
had been competing for 2.05 ("1.95) years. All players had regional
age-group rankings and their competitive experience involved
regional and national competitions at junior level. All participants
completed the experimental procedures.
Task and instruments
Performance
Forehand drive performance was evaluated through the Broer–
Miller Forehand Drive test (as described by Barrow, McGee, &
Tritschler, 1989). Participants were standing at the baseline of the
court. The opposite half of the court was divided into zones cor-
responding to a scoring-system (two, four, six, and eight points),
with balls landing close to the baseline counting for eight points
and balls landing close to the net counting for two points. Following
the description of the test, a rope was placed over the net, at
a height of 1.22 m. Participants were hitting balls coming from
a ball machine (Lobster Elite Freedom). The score of participants
was the total points gained out of 10 strokes. Balls travelling over
the rope were scored half their original value.
Self-confidence and anxiety
The Competitive State Anxiety Inventory-2 revised (CSAI-2R;
Cox, Martens, & Russell, 2003) was used to assess self-confidence
and anxiety. The scale comprises 17 items assessing cognitive
anxiety (five items), somatic anxiety (seven items) and self-confi-
dence (five items). Responses were given on a four-point Likert
scale from 0 (not at all) to 3 (very much so). Cronbach’s alpha in this
study ranged from .74 to .90.
Procedures
Participation was voluntary and athletes were informed that they
could withdraw at anytime if they wanted to. All participants agreed
to participate and parental consent was obtained for all athletes. The
research ethics committee of the researchers’ institution granted
ethical approval for the conduct of the study. The experiment was
completed in five sessions: baseline trial (session 1), training
(sessions 2–4), and experimental trial (session 5).
Session 1
Participants were initially informed that for the following five
sessions they were going to participate in a program aiming to
assess their tennis abilities. The aim was to raise participants’
anxiety to levels comparable to a sport competition, so that the
hypothesised impact of self-talk could be detected. Towards this
direction, Murray and Janelle’s (2003) recommendations were
adopted. In particular, participants were informed that the whole
procedure was going to be recorded, that the results were to be
made public to the club, that performance of individuals was going
to be compared, and that awards (tennis goods) would be given to
the top three players. Subsequently, the procedures regarding the
evaluation were explained, and participants were allowed to ask
questions with regard to these procedures. Finally, the stressful
instructions were repeated and before the beginning of the first trial
participants completed a one-item manipulation check regarding
the stressfulness of the situation (from 1 not at all stressful, to 10
very stressful). Participants performed three sets of 10 drives. The
first set was used for purposes of familiarization (not assessed),
whereas the two following sets were assessed. Upon completion of
the third set participants completed the CSAI-2R (the instruction
was to indicate how they felt during the execution of the task).
Sessions 2–4
Upon completion of the baseline trial, participants were divided
into two groups that were randomly assigned as experimental and
control. Participants were placed in the groups so that no baseline
differences would emerge between the two groups in the variables
of interest (performance, self-confidence, and anxiety). The two
groups followed similar training protocols for the next three
training sessions. In addition, all participants received additional
training as part of their participation to the program. For
the training phase, the backhand drive was used, so that partici-
pants did not practice the stroke that was to be evaluated in the
final assessment. The use of the backhand drive aimed to minimize
possible performance increases due to practicing the stroke, and to
isolate to the highest possible degree the effects of self-talk on task
performance. Participants in the experimental group were intro-
duced to the use of self-talk and were informed that they were
going to use this strategy for their training. The instructor explained
and showed them how to use self-talk. To prevent the appearance
of a Hawthorne effect, participants in the control group spent the
same time receiving a short lecture on tactical aspects of the shot.
All participants performed four sets of eight drives, with a 1 min
interval in-between. Participants in the experimental group used
one self-talk cue for each set. These cue words wereprovided by the
instructor and were both instructional (e.g. shoulder, low, deep)
and motivational (e.g. go, I can, strong). The rational for using
instructional and motivational cues for the training phase was to
have participants practicing, understanding and learning how to
A. Hatzigeorgiadis et al. / Psychology of Sport and Exercise 10 (2009) 186–192188
use the self-talk technique thoroughly, and not just to practice self-
talk cues to use in the experimental measure. Participants were
instructed that they could repeat the cue aloud or in their head
without verbalising it, according to their preference. After the
completion of each set, participants were asked to indicate on a 10-
point scale how frequently they use the instructed cues (1 ¼not at
all, 10 ¼all the time). For the training sessions, balls were thrown
by the coach (not the ball machine) to allow participants to
concentrate better on the practice of self-talk (adjust the timing
and get acquainted with the use of self-talk without the time
pressure put by the ball machine), which was a new strategy. The
same procedures were followed for the two sessions that followed.
The training took place in groups of four or five athletes. To prevent
contact between participants from different groups, the two groups
were scheduled to train different hours of the day.
Session 5
In the fifth session, the test of the first session was repeated.
Participants were reminded of the stressful instructions. With
regard to the awards, to keep all participants involvement high
irrespective of their scores in the initial assessment, it was
announced that the awards would be given to athletes showing
greater improvement compared to the initial test. Despite that
there would be less room for improvement for participants with
higher scores, this maneuver would help participants with lower
scores to sustain interest in the assessment. In addition, to maintain
stress levels, an individual was introduced to the participants as
a member of the tennis federation who had come to watch their
test. Subsequently, the one-item manipulation check regarding the
stressfulness of the situation was administered. As in the baseline
assessment, all participants performed three sets of 10 drives, with
the first set used for familiarization. Participants in the experi-
mental group were asked to choose and state a motivational cue of
their preference that they would use. Upon completion of the two
sets participants completed the CSAI-2R. Following this, a manip-
ulation check protocol was administered. Participants in the
experimental group were asked (a) to indicate on a 10-point scale
the degree to which they used the cue they selected (1¼not at all,
10 ¼all the time), (b) to report whether they used any other cue, (c)
if so, what this cue was, and (d) if so, the degree to which they used
this other cue (1 ¼not at all, 10 ¼all the time). Participants in the
control group were informed that athletes frequently say things to
themselves while performing and were asked to indicate (a)
whether they purposely used with consistency any form of self-talk
during the execution of the task, (b) if so, what was that, and (c) if
so, to what degree (1 ¼not at all, 10 ¼all the time).
After the conclusion of the experimental procedures partici-
pants were explained the purpose of the study. In addition, they
were debriefed in relation to the stressful instructions and in the
presence of the so-called member of the federation, and were
thanked for their participation.
Results
Manipulation check
Stress condition
The first manipulation check involved participants’ perceptions
regarding the stressfulness of the situation. Examination of the
mean scores revealed that participants perceived the situation as
moderately stressful in both the initial and the final assessments
(mean scores 5.32 "1.87 and 5.26 "2.19, respectively).
Self-talk in training – Experimental group
The second manipulation check involved the use of self-talk
during the training sessions by participants in the experimental
group. Examination of the mean scores revealed that athletes made
adequate use of self-talk during the three training sessions
(7.78 "1.19, 7.74 "1.10, and 7.96 "1.07 for the three sessions,
respectively).
Self-talk in final trial – Experimental and control groups
The final manipulation check involved the use of self-talk during
the final trial by participants in the experimental and control
groups. Regarding the experimental group it was revealed that two
of the participants reported not using the selected cue consistently
(scored 2 out of 10). Regarding the control group it was revealed
that six participants reported consistent use of some form of self-
talk. In particular, two of the athletes reported using the cue ‘‘let’s
go’’, one athlete reported using the cue ‘‘strong’’, one athlete
reported using the cue ‘‘I can’’, one athlete reported using the cue
‘‘focus’’, and finally one athlete reported using the cue ‘‘baseline’’.
To ensure the integrity of the experimental manipulations the two
athletes from the experimental group and the six athletes from the
control group were excluded from subsequent analyses. After
removing these participants examination of the means showed
that participants in the experimental group reported consistently
using the cue they selected (7.79 "1.55) and no other self-talk
(1.26 "1.58), whereas participants in the control group reported
not using consistently any form of self-talk (1.67 "1.63).
Baseline differences
Baseline differences were subsequently examined to ensure that
there were no significant differences between the two groups in
performance, self-confidence, cognitive anxiety and somatic
anxiety. This test was performed to secure the meaningfulness of
the repeated measures that would follow. Furthermore, because
boys and girls were distributed to the two groups, gender was also
considered. A two-way MANOVA revealed no significant effect for
group, F(4, 57) ¼.14, p¼.97, gender, F(4, 57) ¼.89, p¼.47, and their
interaction, F(4, 57) ¼.57, p¼.69.
Main analysis – Repeated measures MANOVA
Descriptive statistics for performance, self-confidence, cognitive
and somatic anxiety, for the total sample are presented inTable 1.A
two-way (group by trial) mixed model MANOVA was calculated to
test for differences in performance, self-confidence, cognitive and
somatic anxiety over trials between the two groups. The assump-
tions of normality and homogeneity of variance and covariance
were met. The analysis revealed a significant multivariate group by
trial interaction, with large effects size, F(4, 59) ¼7.08, p<.001,
h
2
¼.32, observed power ¼.99. Examination of the univariate
effects showed significant interaction effects for performance, F(1,
62) ¼19.46, p<.001,
h
2
¼.24, observed power ¼.99, self-confi-
dence, F(1, 62) ¼5.06, p¼.028,
h
2
¼.08, observed power ¼.60, and
cognitive anxiety, F(1, 62) ¼4.96, p¼.030,
h
2
¼.07, observed
power ¼.59, and a non-significant interaction for somatic anxiety,
F(1, 62) ¼1.77, p¼.19. Examination of the pairwise comparisons
and the means revealed that for the experimental group task
Table 1
Descriptive statistics for the total sample
Initial assessment Final assessment
MSD MSD
Performance 37.26 12.65 40.35 13.44
Self-confidence 1.59 .72 1.74 .71
Cognitive anxiety 1.20 .69 1.12 .76
Somatic anxiety .86 .57 .75 .63
A. Hatzigeorgiadis et al. / Psychology of Sport and Exercise 10 (2009) 186–192 189
performance improved (p<.001), self-confidence increased
(p¼.002) and cognitive anxiety decreased (p¼.031), whereas
a decrease that approached significance was revealed for somatic
anxiety (p¼.062). In contrast, no significant differences were
revealed for any of the variables for the control group. The inter-
action pattern and the mean scores for the two groups are dis-
played in Fig. 1 (performance), Fig. 2 (self-confidence), Fig. 3
(cognitive anxiety), and Fig. 4 (somatic anxiety).
Correlations
Finally, the degree to which changes in task performance were
related to changes in self-confidence and anxiety were examined
through Pearson’s correlations. For the purposes of this particular
analysis scores reflecting the changes between baseline and
experimental assessments were calculated by subtracting scores in
the baseline assessment from scores in the experimental assess-
ment (positive scores indicating increases). The analysis, which
involved the whole sample, revealed a positive moderate rela-
tionship between changes in task performance and changes in self-
confidence (r¼.29, p¼.020), whereas no relationships were
identified between changes in task performance and changes in
cognitive anxiety (r¼$.01, p¼.996) and between changes in task
performance and changes in somatic anxiety (r¼$.06, p¼.668).
Discussion
The primary purpose of the present study was to explore the
effects of motivational self-talk on self-confidence, anxiety, and
task performance. Furthermore, the degree to which changes in
task performance were related to changes in self-confidence and
anxiety was tested. Overall, it was found that self-talkhad a positive
effect on task performance, increased self-confidence, reduced
cognitive anxiety, and that changes in task performance were
related to changes in self-confidence.
Before proceeding to the hypotheses testing, the experimental
conditions that were sought were evaluated. The firstobjective was
to create an environment that would reasonably raise stressful
perceptions, so that effects of self-talk on anxiety and self-confi-
dence could be examined. The recommendation of Murray and
Janelle (2003) for creating stressful conditions were adopted, which
have proved effective in previous research (e.g. Murray & Janelle,
2007; Wilson, Smith, Chattington, Ford, & Marple-Horvat, 2006).
The results showed that participants perceived the situation as
moderately stressful. Moreover, examination of the anxiety scores,
in particular cognitive, for the baseline assessment showed
moderate levels of anxiety intensity, which resemble anxiety levels
reported in field studies with young athletes (e.g. Hall & Kerr, 1997).
Considering that participants were athletes with experiences of
competitive stress, the moderate levels of stress that were induced
were deemed satisfactory.
The second objective was to familiarise participants in the
experimental group with the use of self-talk. Results regarding the
use of self-talk in the training session and the final assessment
showed that participants made adequate use of the self-talk
strategy. Finally, to ensure the integrity of the experimental
conditions with regard to the use of self-talk, participants’ reports
in the final assessment were examined. Six participants from the
control group reported systematic use of cues that could be
described as instructional or motivational self-talk. Furthermore,
two participants from the experimental group reported not using
the cue they selected. With regard to the control group there is little
that can be done and that mainly relates to employing participants
with no previous experience in the use of mental strategies and
keeping participants unaware of the experimental conditions. To
43.62 ± 14.23
37.09 ± 13.84 36.65 ± 11.63
37.45 ± 11.38
35
37
39
41
43
45
initial final
Trial
Performance
experimental
control
Fig. 1. Performance scores in the initial and final trials for the experimental and
control groups.
1.89 ± .63
1.59 ± .70
1.59 ± .78
1.59 ± .75
1,5
1,6
1,7
1,8
1,9
2
initial final
Trial
Self-confidence
experimental
control
Fig. 2. Self-confidence scores in the initial and final trials for the experimental and
control groups.
0.89 ± .65
1.25 ± .72 1.38 ± .80
1.15 ± .68
0,8
1
1,2
1,4
1,6
1,8
initial final
Trial
Cognitive anxiety
experimental
control
Fig. 3. Cognitive anxiety scores in the initial and final trials for the experimental and
control groups.
A. Hatzigeorgiadis et al. / Psychology of Sport and Exercise 10 (2009) 186–192190
prevent the promotion of self-talk, participants’ self-talk during the
baseline assessment was not assessed. Furthermore, participants in
the two groups trained and completed the final assessment sepa-
rately. The vast majority of participants in the experimental group
made adequate use of the self-talk strategy which suggests that
short training may have been beneficial for the consistent use of the
technique. Furthermore, allowing participants to choose among
appropriate cues may have also facilitated the use of self-talk.
In accordance with previous findings, the use of self-talk
improved task performance. Research has generally supported the
beneficial effects of self-talk on learning and task performance in
various settings (novice athletes, Perkos et al., 2002; highly skilled
athletes, Landin & Hebert, 1999; learned skills, Harvey, Van Raalte,
& Brewer, 2002; new skills, Hatzigeorgiadis et al., 2004), and sports
(ski, Rushall, Hall, Roux, Sasseville, & Rushall, 1988; sprints, Mallet
& Hanrahan, 1997; tennis, Landin & Hebert, 1999; basketball, The-
odorakis, Chroni, Laparidis, Bebetsos, & Douma, 2001). The results
of the present study confirm that self-talk is an effective strategy for
enhancing task performance. Although motivational self-talk has
been primarily recommended for gross tasks requiring strength
and endurance (Theodorakis et al., 2000), the present results
suggest that it can also be effective for tasks requiring precision,
a finding previously reported by Hatzigeorgiadis et al. (2004). They
suggested that different types of self-talk may have different effects
on task performance based on the nature of the task and the type of
self-talk that is used. That self-talk also improved self-confidence
may explain why task performance in this precision task improved.
This possibility is further discussed below.
Post-experimental reports from studies examining the effective-
ness of self-talk have given indications that self-talk may increase self-
confidence. Landin and Hebert (1999), in a study with skilled female
tennis players, used a single-item measure to assess their self-confi-
dence in accomplishing a task, before and after the implementation of
a self-talk treatment program. The study involved five players, thus
statistical analyses were not conducted; however, an increase in
players’ self-confidence was recorded. In a study following the same
research design with female football players, Johnson et al. (2004)
reported similar findings. Finally, Perkos et al. (2002) in a study with
young basketball players used a single-item measure to assess
perceived effectiveness of self-talk in relation to athletes’ self-confi-
dence, after implementing a 12-week intervention program. Partici-
pants reported that the use of self-talk helped them feel more
confident. Nevertheless, so far this hypothesis had not been tested
experimentally. The present study examined changes in self-
confidence after the implementation of a self-talk training program.
The results showed that self-confidence of participants using motiva-
tional self-talk increased, whereas that of control participants
remained unchanged,thus providing empirical evidence for theeffects
of motivational self-talk on self-confidence.
The effectiveness of cognitive strategies in reducing anxiety in
athletes has been supported by Maynard et al. (1995, 1998) in two
intervention studies where positive thought control training was
used. With specific regard to the impact of self-talk on anxiety,
preliminary evidence has been provided by Hatzigeorgiadis, Zour-
banos, and Theodorakis (2007) in a novel experimental motor task
with students. Their findings showed that the use of self-talk
resulted in reductions of cognitive anxiety. Furthermore, comparing
the effects of an instructional and an anxiety control cue on anxiety
symptoms showed that the anxiety related cue resulted in greater
reduction of cognitive anxiety than the instructional cue. The
authors supported the specificity of self-talk effects in relation to the
selected cues. Nevertheless, no control group was employed in that
study. The above evidence and the theoretical links between self-
confidence and anxiety along with the respective empirical evidence
(Martens et al., 1990), led us to hypothesise that if self-talk increases
self-confidence, then anxiety should be reduced. The results of the
study supported this hypothesis, in particular for cognitive anxiety,
as intensity of symptoms was reduced for the experimental group
but not for the control group. Regarding somatic anxiety the same
pattern was revealed; however the reduction was not significant.
Further analyses were conducted to test the relationship
between changes in task performance and changes in self-confi-
dence and anxiety. The results revealed that changes in self-
confidence were moderately positively related to changes in task
performance. In contrast, no relationships were found between
changes in cognitive and somatic anxiety and changes in task
performance. Research on the relationship between self-confidence
and performance has provided consistent results indicating that
self-confidence and performance are positively related, and this
relationship is moderate in size. A meta-analysis by Craft, Magyar,
Becker, and Feltz (2003) showed an effect of .36, and a similar meta-
analysis by Woodman and Hardy (2003) revealed an effect of .24. In
contrast, examination of the relationship between anxiety and
performance has provided equivocal results, suggesting that
cognitive and somatic anxiety can have either positive, negative or
no relationships with sport performance. Characteristic of this
inconsistency are the results of the two aforementioned meta-
analyses. Craft et al. (2003) reported an effect of .13 between
cognitive anxiety and performance and an effect of .09 between
somatic anxiety and performance, whereas Woodman and Hardy
(2003) reported an effect of $.10 for the relationship between
cognitive anxiety and performance (the effect between somatic
anxiety and performance was not tested in that study). The results
of the present study are in line with the findings regarding the self-
confidence – performance relationship. Given that the use of
motivational self-talk enhanced task performance and increased
self-confidence and considering that increases in self-confidence
were related to increase in task performance, it could be speculated
that increases in self-confidence may be a viable mechanism
explaining the facilitating effects of self-talk on task performance.
Even though self-talk reduced anxiety, inparticular cognitive, these
changes were not related to changes in task performance.
Limitations of the study should be addressed, in particular with
regard to the mediating role of self-confidence in the self-talk – task
performance relationship that was discussed. Measures of anxiety
and self-confidence were obtained after the conclusion of the task,
therefore it is possible that participants’ responses could have been
influenced by their performance. Furthermore, the analyses that
were performed were independent for self-confidence, anxiety,
0.61 ± .45
0.83 ±.57
0.90 ± .76
0.90 ± .59
0,50
0,60
0,70
0,80
0,90
1,00
initial final
Trial
Somatic anxiety
experimental
control
Fig. 4. Somatic anxiety scores in the initial and final trials for the experimental and
control groups.
A. Hatzigeorgiadis et al. / Psychology of Sport and Exercise 10 (2009) 186–192 191
and task performance. Self-confidence and performance have
a reciprocal relationship, so it is possible that either increases in
self-confidence due to self-talk raised task performance, or
increases in task performance due to self-talk raised self-confi-
dence (or even both in a reciprocal manner). The timing of anxiety
and self-confidence measures is a common problem in sport
anxiety research. Measures administered before or after an event
can shed limited light into what happens during task performance.
Still, the fact that participants were not aware of their exact
performance and whether they were improving or not, seems to
strengthen the possibility of the self-confidence mediation. Overall,
the results did show that motivational self-talk increased self-
confidence and reduced cognitive anxiety, so the mediation
hypothesis is a viable explanation regarding the facilitating effects
of self-talk on task performance. Nonetheless, the present findings
cannot support the mediation, but rather suggest that self-confi-
dence is a likely mechanism through which self-talk facilitates task
performance, and challenge further research with appropriate
designs to support the mediational role of self-confidence. To
further explore the mediation hypothesis future research could
employ laboratory tasks where anxiety can be assessed at the time
of the performance using a combination of physiological and
psychological measures. Furthermore, because self-talk is said to
operate through several functions, future research should test
simultaneously multiple functions and identify how these may
interact in raising task performance.
Despite the above issues, the present study offers valuable
evidence regarding the role of self-talk. The results provided further
support regarding the effectiveness of self-talk on task performance.
Hatzigeorgiadis et al. (2004) postulated an attentional interpreta-
tion of the facilitating effects of self-talk on task performance and
suggested that further research should look for other likely functions
of self-talk. The present findings suggest two more possible func-
tions through which motivational self-talk mayoperate, increases in
self-confidence and reduction in anxiety intensity symptoms. That
the effects of self-talk on anxiety were not related to task perfor-
mance seems to indicate that self-talk can be used to reduce anxiety,
however, whether this reduction will relate to increases in task
performance is probably moderated by otherfactors. Understanding
the functions through which self-talk operates will facilitate the
development of effective self-talk plans and towards this direction
the present study provided valuable evidence.
References
Barrow, H., McGee, R., & Tritschler, K. (1989). Practical measurement in physical
education and sport. Philadelphia: Lea & Febiger.
Cox, R. H., Martens, M. P., & Russel, W. D. (2003). Measuring anxiety in athletics: the
revised Competitive State Anxiety Inventory-2. Journal of Sport and Exercise
Psychology, 25, 519–533.
Craft, L. L., Magyar, T. M., Becker, B. J., & Feltz, D. L. (2003). The relationship between
the Competitive State Anxiety Inventory-2 and sport performance: a meta-
analysis. Journal of Sport and Exercise Psychology, 25, 44–65.
Ellis, A. (1976). Reason and emotion in psychotherapy. New York: Lyle Stuart.
Finn, J. A. (1985). Competitive excellence: it’s a matter of mind and body. The
Physician and Sportmedicine, 13, 61–75.
Hall, H. K., & Kerr, A. W. (1997). Motivational antecedents of precompetitive anxiety
in youth sport. The Sport Psychologist, 11, 24–42.
Hardy, J. (2006). Speaking clearly: a critical review of the self-talk literature.
Psychology of Sport and Exercise, 7, 81–97.
Hardy, J., Gammage, K., & Hall, C. R. (2001). A descriptive study of athletes self-talk.
The Sport Psychologist, 15, 306–318.
Hardy, L., Jones, G., & Gould, D. (1996). Understanding psychological preparation for
sport: Theory and practice of elite performers. Chichester, UK: Jones Wiley & Sons.
Harvey, D. T., Van Raalte, J., & Brewer, B. W. (2002). Relationship between self-talk
and golf performance. International Sports Journal, 6, 84–91.
Hatzigeorgiadis, A. (2006). Instructional and motivational self-talk: an investigation
on perceived self-talk functions. Hellenic Journal of Psychology, 3(2), 164–175.
Hatzigeorgiadis, A., Theodorakis, Y., & Zourbanos, N. (2004). Self-talk in the
swimming pool: the effects of ST on thought content and performance on
water-polo tasks. Journal of Applied Sport Psychology, 16, 138–150.
Hatzigeorgiadis, A., Zourbanos, N., & Theodorakis, Y. (2007). The moderating
effects of self-tal k content on se lf-talk functions. J ournal of App lied Sport
Psychology, 19, 240–251.
Johnson, J. J. M., Hrycaiko, D. W., Johnson, G. V., & Hallas, J. M. (2004). Self-talk and
female youth soccer performance. The Sport Psychologist, 18, 44–59.
Landin, D. (1994). The role of verbal cues in skill learning. Quest, 46, 299–313.
Landin, D., & Hebert, E. P. (1999). The influence of ST on the performance of skilled
female tennis players. Journal of Applied Sport Psychology, 11, 263–282.
Mallett, C. J., & Hanrahan, S. J. (1997). Race modelling: an effective cognitive strategy
for the 100 m sprinter? The Sport Psychologist, 11, 72–85.
Martens, R., Burton, D., Vealey, R. S., Bump, L. A., & Smith, D. E. (1990). Development
and validation of the Competitive State Anxiety Inventory-2. In R. Martens,
R. S. Vealey, & D. Burton (Eds.), Competitive anxiety in sport (pp. 117–190).
Champaign, IL: Human Kinetics.
Maynard, I. W., Hemmings, B., Greenlees, I. A., Warwick-Evans, L., & Stanton, N.
(1998). Stress management in sport: a comparison of unimodal and multimodal
interventions. Anxiety, Stress, and Coping, 11, 225–246.
Maynard, I. W., Smith, M. J., & Warwick-Evans, L. (1995). The effects of a cognitive
intervention strategy on competitive state anxiety and performance in
semi-professional soccer players. Journal of Sport and Exercise Psychology, 17,
428–446.
Meichenbaum, D. (1977). Cognitive-behaviour modification: An integrative approach.
New York: Plenum Press.
Murray, N. P., & Janelle, C. M. (2003). Anxiety and performance: a visual search
examination of the processing efficiency theory. Journal of Sport and Exercise
Psychology, 25, 171–187.
Murray, N. P., & Janelle, C. M. (2007). Event-related potential evidence for the
processing efficiency theory. Journal of Sports Sciences, 25, 161–171.
Nideffer, R. (1993). Attention control training. In R. N. Singer, M. Murphey, &
I. K. Tennant (Eds.), Handbook of research on sport psychology (pp.127–170). New
York: Macmillan.
Paivio, A. (1985). Cognitive and motivational functions of imagery in human
performance. Canadian Journal of Applied Sport Sciences, 10, 22–28.
Perkos, S., Theodorakis, Y., & Chroni, S. (2002). Enhancing performance and skill
acquisition in novice basketball players with instructional self-talk. The Sport
Psychologist, 16, 368–383.
Rushall, B. S., Hall, M., Roux, L., Sasseville, J., & Rushall, A. S. (1988). Effects of three
types of thought content instructions on skiing performance. The Sport
Psychologist, 2, 283–297.
Thelwell, R. C., & Greenlees, I. A. (2003). Developing competitive endurance
performance using mental skills training. The Sport Psychologist, 17, 318–337.
Theodorakis, Y., Chroni, S., Laparidis, K., Bebetsos, V., & Douma, I. (2001). Self-talk in
a basketball shooting task. Perceptual and Motor Skills, 92, 309–315.
Theodorakis, Y., Hatzigeorgiadis, A., & Chroni, S. (2008). Self-Talk: it works, but
how? Development and preliminary validation of the functions of Self-Talk
Questionnaire. Measurement in Physical Education and Exercise Science, 12,
10–30.
Theodorakis, Y., Weinberg, R., Natsis, P., Douma, E., & Kazakas, P. (2000). The effects
of motivational versus instructional self-talk on improving motor performance.
The Sport Psychologist, 14, 253–272.
Van Raalte, J. L., Brewer, B. W., Rivera, P. M., & Petipas, A. J. (1994). The relationship
between observable self-talk and competitive junior tennis players’ perfor-
mances. Journal of Sport and Exercise Psychology, 16, 400–415.
Wilson, M., Smith, N. C., Chattington, M., Ford, M., & Marple-Horvat, D. E. (2006).
The role of effort in moderating the anxiety–performance relationship: testing
prediction of processing efficiency theory in simulated rally driving. Journal of
Sports Sciences, 24, 1223–1233.
Woodman, T., & Hardy, L. (2003). The relative impact of cognitive anxiety and self-
confidence upon sport performance: a meta-analysis. Journal of Sports Sciences,
21, 443–457.
Zervas, Y., Stavrou, N. A., & Psychountaki, M. (2007). Development and validating of
the self-talk questionnaire (S-TQ) for sports. Journal of Applied Sport Psychology,
19, 142–159.
Zinnser, N., Bunker, L., & Williams, J. M. (2006). Cognitive techniques for building
confidence and enhancing performance. In J. M. Williams (Ed.), Applied sport
psychology: Personal growth to peak performance (5th ed.) (pp. 349–381).
New York, NY: McGraw-Hill Companies, Inc, Higher Education.
A. Hatzigeorgiadis et al. / Psychology of Sport and Exercise 10 (2009) 186–192192
