AN EXAMINATION OF ATHLETES’ SELF-EFFICACY AND

STRENGTH TRAINING EFFORT DURING AN ENTIRE

OFF-SEASON
TODD A. GILSON,1 G.F. ‘‘CISCO’’ REYES,2 AND LINDSEY E. CURNOCK1
1
Department of Kinesiology and Physical Education, Northern Illinois University, Dekalb, Illinois; and 2Department of Exercise
and Sport Science, Concordia University, Portland, Oregon

ABSTRACT INTRODUCTION

S
Gilson, TA, ‘‘Cisco’’ Reyes, GF, and Curnock, LE. An elf-efficacy, or the beliefs in one’s capabilities to
examination of athletes’ self-efficacy and strength training effort organize and execute the courses of action required to
during an entire off-season. J Strength Cond Res 26(2): 443– produce given attainments, can be a powerful force in
451, 2012—Over the past 30-plus years in which self-efficacy sport because individuals’ levels of motivation are
Downloaded from http://journals.lww.com/nsca-jscr by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC1y0abggQZXdtwnfKZBYtws= on 04/07/2022

more directly tied to what they believe to be true, rather than

(or confidence at a task) has been researched, findings have
to what is objectively true (1). Specifically, individuals’ self-
shown that in almost every domain of human functioning, self-
efficacy is theorized to be formulated from 4 sources: mastery
efficacy positively relates to effort, persistence, and other
experiences or whether an individual was successful at the
adaptive behaviors. However, in the past decade, new research
skill or task in the past; vicarious experiences or viewing
postulating that too much self-efficacy can lead to complacency another individual succeeding or failing; social persuasion or
and a subsequent downturn in behavior or performance has being told by a significant other that one has the ability to be
also experienced resurgence in the literature. Therefore, the successful; and physiological states or interpreting the body’s
purpose of this study was to test these opposing viewpoints physiological reactions as either positive or negative (13).
regarding self-efficacy and effort for Division I athletes in Because these efficacy sources and perceptions of the
a strength and conditioning domain over off-season training, environment one engages in can change from moment to
a procedure yet to be undertaken. Subjects (N = 99), from moment, self-efficacy is a state variable that fluctuates in
4 different sports (Mage = 20.0 years, SD = 1.2 years), accordance with changing perceptions. Thus, in strength and
completed self-efficacy and effort measures at 4 distinct time conditioning, an athlete’s confidence in his or her ability to
points during off-season training. In addition, strength and successfully execute a task could change based on recently
conditioning coaches also rated each subject’s effort—at each returning from an injury, seeing a teammate succeed, being
yelled at by a coach, or feeling one’s heart begin to beat faster.
time point—so that a more valid measure of this construct could
However, even though perceptions may change, the skill set of
be attained. Results were analyzed using a multilevel approach
an athlete is not in a constant state of flux during a skill;
and revealed that self-efficacy was positively, and significantly,
therefore, self-efficacy is not concerned with the number of
related to the current effort that athletes exerted in strength skills one possesses but with what an athlete believes he or she
training sessions. Consequently, practitioners are advised to can accomplish with those skills (1). Based on this, it is possible
structure strength and conditioning training sessions and that self-efficacy could be a better predictor of the amount of
the overall environment in ways that will positively impact the effort put forth to accomplish a particular skill than the actual
4 proven sources of self-efficacy. performance related to the skill itself.
Past research examining self-efficacy and a variety of outcomes
KEY WORDS confidence, hierarchical linear modeling, football,
has shown that increases in self-efficacy are a benefit to
basketball, volleyball, soccer
individuals who engage in sport (6,10,15–18,21,23,24). Specifi-
cally in strength training, social persuasion has been used in
studies to demonstrate that subjects’ performances increase
when they were unaware of the amount of weight on a bar or
intentionally informed by a significant other that the weight is
Address correspondence to Dr. Todd A. Gilson, tgilson@niu.edu. less than actually present (14,36). In addition, physiological states
26(2)/443–451 have also been altered and impending performances measured in
Journal of Strength and Conditioning Research strength training environments. In these works, the relationship
Ó 2012 National Strength and Conditioning Association between confidence and performance was again linear, because

VOLUME 26 | NUMBER 2 | FEBRUARY 2012 | 443

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 Self-Efficacy and Effort

individuals performed better at 1RM tests when their arousal second game. Research has shown that when a team wins
levels were manipulated by researcher-manufactured competi- a game, the players’ self-efficacy is significantly increased, and
tion or they were ‘‘psyched up’’ before a strength test (27,37). after a loss, it significantly decreases (10). Winning the first
Research has also shown that higher levels of self-efficacy game will instill confidence, and possibly overconfidence in
have led to greater effort when engaging in a competitive sport players, whereas the losing team has the opposite effect,
and facing an obstacle outside of competition compared with lowering self-efficacy, and therefore, athletes may put forth
those with lower levels of self-efficacy (12,15). For instance, in a greater amount of effort toward the next game. As a result,
research on self-efficacy and baseball, self-efficacy was shown it may be advisable for practitioners to instill mildly negative
to significantly predict batting effort (15). In athletic settings, it expectations, to curtail overconfidence (29).
is hard to overstate the power self-efficacy has on athletes’ Although some researchers claim that the notion of
performances, because under challenging sport conditions, performance directly affecting the next performance removes
where one decision can mean the difference between success one’s cognitive processes from the equation, and thus is too
and failure, self-efficacy has been shown to be the sole barren to be classified as a theory (3), both ‘‘camps’’ of researchers
determinant of success (18). Thus, the majority of work on self- do agree that self-efficacy will positively affect behavior or
efficacy in sport (and other domains of human functioning) performance when an individual is compared with another
continues to show that self-efficacy is positively related to subject (38). Most of the research in self-efficacy is conducted in
adaptive behavior and impending performances (3,12,23). regard to performance rather than to effort. Although it is often
However, despite the seemingly insurmountable evidence hard to measure effort, there is a considerable lack of research in
put forth outlining a positive relationship between self- this area. It is important to understand the self-efficacy–effort
efficacy, behavior, and performance, the uniformity of these relationship because effort has an impact on performance and
findings is currently being challenged. In particular, new could possibly mediate this connection between self-efficacy and
results highlight the fact that individuals too high in self- performance. Additionally, future directions have been spelled
efficacy will suffer debilitating performances over time, out by many researchers to understand the effects of self-efficacy
whereas other individuals—who may unsuccessfully complete more completely. For example, researchers should examine
a task or not meet their performance expectations based on subjects in physical exertion skills, which have significance for
their current self-efficacy—will improve their subsequent the subjects (3,31). In addition, future studies should allow
performance (30–32,38). The explanation for these results is a significant amount of time to pass between measurements of
that the strong relationship between self-efficacy and perfor- self-efficacy and impending behavior (which has not always
mance occurs because of the influence of a performance on been done in past work), because measuring these constructs
self-efficacy, not because of the influence of self-efficacy on repeatedly does not allow ample time for self-efficacy to operate
a performance. In other words, the consequences one in a setting (3,31,32,38). Finally, although studies have examined
experiences will have a direct result on subsequent self-efficacy. self-efficacy and resulting behavior or performance changes for
An example of how self-efficacy may be negatively related a skill or competition, more work is needed that measures
to performance can be witnessed with a women’s soccer subjects over multiple time points and then analyzes the data for
player who strongly believes she can record a vertical jump of variations within the same person over time and also between
at least 21$ in upcoming testing. When this individual is subjects simultaneously (25,35).
tested, and if she successfully jumps at least 21$, she might not Therefore, the purpose of this study was to examine the
put forth as much as effort toward this skill in the coming relationship between self-efficacy and effort in strength training
weeks/months because of her confidence level. Therefore, for Division I athletes during the entire off-season training
when her ability is measured again, she is likely to again rate program. Although maybe not as complete as the present work,
her self-efficacy for jumping 21$, 22$, or even 23$ as very the vast amount of research in sport and exercise settings has
strong, yet may suffer a decline in performance. Conversely, if shown a positive relationship between self-efficacy and
this same women’s soccer player was initially tested in the adaptive behaviors, and it is expected that this research will
vertical jump and only scored a maximum vertical of 17 in., show similar results for athletes over time, when compared
she will likely devote more practice time and greater effort to against their previous effort performance (i.e., repeated
this skill in the future. Consequently, when tested again, her measures) and when all data are aggregated and individuals
self-efficacy for successfully jumping 21" will be lower than at are evaluated against other athletes (i.e., between subjects).
the first testing time point, because of her poor past
performance; however, her new performance will most likely METHODS
improve. This phenomenon has been found in National Experimental Approach to the Problem
Basketball Association (NBA) playoff games, for example, the Although self-efficacy and behaviors have been studied
teams that lost the first game in a playoff series were more longitudinally in sports such as baseball, hockey, and
likely to win the second (22). Although self-efficacy was not basketball (10,15,20), these works did not take advantage of
directly measured in this study, it could be one possible new multilevel modeling analyses; thus, results may differ
explanation as to why the losing team tends to win the when explored in a setting devoted to training for sport (3). As
the TM

444 Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 the TM

Journal of Strength and Conditioning Research | www.nsca-jscr.org

TABLE 1. Level 1 descriptive statistics and intercorrelations for self-efficacy and effort in strength training (n = 99).*

Subscale M (SD) 1 2 3 4 5 6 7 8

1. Self-efficacy 1 9.59 (0.34)

2. Self-efficacy 2 9.63 (0.33) 0.59†
3. Self-efficacy 3 9.64 (0.31) 0.53† 0.60†
4. Self-efficacy 4 9.66 (0.26) 0.62† 0.60† 0.76†
5. Effort 1 85.4 (1.0)% 0.39† 0.38† 0.39† 0.47†
6. Effort 2 84.8 (0.88)% 0.40† 0.46† 0.46† 0.43† 0.66†
7. Effort 3 84.8 (1.1)% 0.37† 0.45† 0.55† 0.53† 0.71† 0.73†
8. Effort 4 84.2(0.98)% 0.20‡ 0.34† 0.38† 0.38† 0.63† 0.62† 0.73†
*Numbers after each construct relate to the time point of data collection. Self-efficacy is on an 11-point scale ranging from 0 to 10.
†p # 0.01.
‡p # 0.05.

previously highlighted, a multilevel approach to understand- SD = 1.2 years), who were current Division I athletes in
ing how self-efficacy affects effort is best suited for this study football (n = 62), volleyball (n = 9), men’s soccer (n = 19), and
because the research question is concerned with both men’s basketball (n = 9). Subjects’ data showed adequate
individual changes in self-efficacy (i.e., within-person data) diversity in demographic variables collected. For example,
and how athletes deviate among each other based on self- self-reported ethnicity revealed that 61.6% of athletes were
efficacy levels (i.e., between-person data) (25,35). Therefore, Caucasian, 22.2% African American, 7.1% Hispanic, 8.1%
by using strength training, which is a common arena for most multiple ethnicities, and 1% other. In addition, the majority of
athletes preparing for their sport, a more complete picture of subjects in this study were in their freshman year of eligibility
how self-efficacy impacts effort can be obtained. (41.4%), followed by juniors, sophomores, and seniors (32.3,
19.2, and 7.1%), respectively. Finally, 2 dichotomous variables
Subjects (scholarship status and recruitment by current head sport
Approval for this study was granted by an Institutional coach) were also collected to be used in data analysis. In this
Review Board (IRB) at a Midwestern university. Complete study, 70.7% of the athletes were on either a full or partial
data were collected from 99 subjects (Mage = 20.0 years, athletic scholarship, and 48.5% of subjects were recruited by

TABLE 2. Multilevel model for effort in strength training.

Fixed effect Parameter Coefficient SE df t

Estimation of fixed effects

Intercept b00 8.53 0.08 96 109.68*
Aggregated self-efficacy b01 4.11 0.26 96 16.05*
Aggregated residual past effort b02 21.03 0.09 96 212.09*
Time b10 20.03 0.03 291 21.05
Self-efficacy b20 0.36 0.19 291 1.94†
Residual past effort b30 0.10 0.05 291 2.17‡

Random effect Parameter Variance component SD df x2

Estimation of variance components

Mean effort r0j 0.09 0.29 96 209.60*
Level 1 effect eij 0.22 0.47
*p # 0.001.
†p # 0.05.
‡p # 0.01.

VOLUME 26 | NUMBER 2 | FEBRUARY 2012 | 445

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 Self-Efficacy and Effort

associated with friend relation-

ships, significant other relation-
TABLE 3. Effort in strength training variance explained. ship(s), family, and strength and
Within-person Between-person
conditioning coaches. Responses
Predictors level (s2) level (t00) to the CSI were then averaged so
that comparisons could be made
No predictors 0.22 0.50 across all psychological measures.
Level 1 only 0.21 0.50 The final questionnaire con-
Percent of level variance 2.3
Percent of total variance 0.7 sisted of an effort report ques-
Levels 1 and 2 0.21 0.09 tionnaire (ERQ), where subjects
Percent of level variance 2.3 82.9 privately recorded how much
Percent of total variance 0.7 57.2 effort they perceived they gave
in strength training at the con-
clusion of a data collection week.
This questionnaire simply asked
the subjects to report their per-
their current head sport coach. (It should be noted that no ceived effort, ranging from 0 to 100% in increments of 10%. In
demographic variables were found to be significant and thus addition, the strength and conditioning coach, who worked with
were not included in final analyses). each sport, also completed an ERQ questionnaire for each
athlete. This way, a more valid measure of effort could be
Instrumentation calculated using both athlete and coach perceptions.
The subjects in this study completed 4 distinct questionnaires.
The first questionnaire was an 11-item self-efficacy effort Procedures
questionnaire (SEEQ). The SEEQ was constructed by Upon IRB approval, appropriate athletic staff were ap-
following appropriate guidelines put forth by past researchers proached about participation in this study and informed
(2,13). In particular, the subjects responded to the stem of, about their rights and responsibilities, should they choose to
‘‘Over the next week, how confident are you that you can participate. After all appropriate athletic staff agreed to
earn an effort score of at least . between your self-rating and participate in this study, 4 time points for data collection were
your head strength coach’s rating?’’ which was then followed set with each sport during off-season strength and condi-
by 11 effort ratings of 0–100%, stratified every 10%. The tioning training sessions (i.e., January–May). Each time point
athlete then responded to each of these 10% stratifications on was positioned approximately 1 month apart so that if any
an 11-point Likert scale, with 0 = not at all confident and changes in effort or psychological constructs presented
10 = absolutely confident. Therefore, the athletes essentially themselves, they could be observed (1).
rated their effort perceptions for intervals of at least 10%, at To facilitate ease of data collection, the lead researcher
least 20%, at least 30%, etc. Subjects’ scores were computed attended all strength and conditioning training sessions when
by averaging responses from the SEEQ. data were scheduled to be collected. All the athletes in this
The second questionnaire used in this study was a 3-item value study had completed at least 1 semester of training with their
of strength training questionnaire (VQ). The 11-point Likert current strength and conditioning coach (i.e., all were at least
scale previously used for the SEEQ was also implemented with freshman athletes), and the majority of the sample had much
this questionnaire; however, athletes’ responses were collected more experience; thus, the subjects were considered to be
through a series of questions related to the construct being previously trained. At time point #1, athletes and coaches
studied. For instance, an example item from the VQ was, ‘‘How were informed of the purpose of the study, they were
much do you currently value giving high effort in strength informed that their answers to all questionnaires would be
training sessions for attaining your personal goals in sport?’’ confidential, and they then completed the necessary IRB
Therefore, the responses to each item in the VQ could range consent form and a demographic questionnaire. Specifically,
from 0 = no value and 10 = high value. Again, subjects’ scores athletes and coaches were read a script approved by the IRB,
were calculated by averaging VQ responses. which informed them about the types of psychological
The third type of questionnaire used in this study was a 6-item constructs that were being measured, what questions this
current stress inventory measure (CSI). Again, a similar 11-point study was attempting to answer, that strength and condi-
Likert scale was used for this questionnaire that ranged from tioning coaches would also be evaluating athletic effort
0 = no stress, pleasant to 10 = high stress, unpleasant. Example displayed, and that participation was completely voluntary.
CSI items were, ‘‘Thinking about today only, what is your current Subsequently, the athletes privately completed the SEEQ,
stress level toward your academic work?’’ and ‘‘Thinking about VQ, CSI, and then returned them directly to the lead
today only, what is your current stress level toward your sport researcher, who administered all forms and questionnaires.
coach?’’ Other questions on the CSI measured stress levels Exactly 1 week after the completion of questionnaire packets
the TM

446 Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 the TM

Journal of Strength and Conditioning Research | www.nsca-jscr.org

by the subjects, the lead researcher returned, and all the incorporated a linear growth model using HLM because it had
athletes participating in the study were instructed to rate the (a) 3 or more waves of data; (b) an outcome whose values
perceived effort they put in strength and conditioning training change systematically over time; and (c) a sensible metric for
sessions (over the past week) by completing the ERQ. In clocking time (28).
addition, the strength and conditioning coach in charge of
each sport also gave the lead researcher an ERQ for each RESULTS
athlete within 48 hours of the completion of the week’s Table 1 presents self-efficacy and effort means, SDs, and
training. This exact process was then replicated for time intercorrelations for all 99 subjects at level 1, or the within-
points 2–4, with the only alterations being the removal of the person level. The results reveal that significant correlations
IRB consent form and demographic questionnaire. were achieved between self-efficacy and effort in strength
training at all time points, which is not surprising considering
Statistical Analyses identical findings in past research (10,15,18,32,38). At the
After data were collected, they were entered in a traditional aggregated level (i.e., level 2), the correlation between self-
software analysis program (SPSS) for preliminary analysis. To efficacy and effort was also quite strong (r = 0.57), which was
provide the most comprehensive picture of an athlete’s effort significant at the p , 0.01 level. Finally, although effort did
displayed in strength training during the weeks of inquiry, decline during the course of this study, this change was not
both the athlete’s and the strength and conditioning coach’s significant and could have occurred because the subjects tried
perceived ratings on the ERQ were used in computing the harder at the outset of this study when being evaluated by the
dependent variable of effort. Specifically, each individual was strength and conditioning coach, which was a new phe-
given a 50% weight in computing the total effort for each nomenon. Thus, when time point 4 transpired, the notion of
subject. Thus, if an athlete rated his effort in strength training being evaluated by a strength and conditioning coach—-
for the week as 90% and the strength and conditioning coach, related to effort—was less of a concern.
who directed that same sport, rated the athlete’s effort as 80%, When analyzing the unconditional means model in HLM
the value of 85% was used for that time point. v6.04 (26), it was found that the intraclass correlation
Once hard copies of data were properly collected, entered, coefficient (ICC, r) was 0.69. This simply meant that 69% of
and stored, initial descriptive analysis revealed that variables the variance in subjects’ effort was located at the between-
collected via questionnaires were normally distributed. The person level, and 31% of the variability was found at the
results showed that the most extreme skewness and kurtosis within-person level. Additionally, the random effect for the
scores were found in the SEEQ at time point 3; however, effort intercept indicated that effort did vary across
these values, of 21.4 and 2.3 were within the acceptable individuals, x2(98, N = 99) = 757.83, p , 0.001. Thus, the
range; thus, no data transformations were required. Addi- previous information revealed that HLM was an appropriate
tionally, internal consistencies all loaded above the uniformly statistic to use because significant variance was located both
accepted a = 0.7 for each questionnaire at each time point. for an individual over time and between individuals when
Finally, there were no univariate or multivariate outliers data were aggregated.
among the sample of athletes, and the relationship between To examine the hypotheses concerning self-efficacy and
self-efficacy and effort was linear across all time points, further effort in this study, the following level 1 and level 2 models
confirming the normality of the data. were built in HLM v6.04 (26):
Data were then exported and analyzed using Hierarchical
Level 1
Linear and Nonlinear Modeling (HLM) v6.04 (26). The HLM
was the appropriate statistic to use for 2 reasons: first, this study Yij ¼ p0j þ p1j ðTIMEÞ þ p2j ðSEij  SEj Þ
had various time points that were ‘‘nested’’ within individuals. þ p3j ðZEffortij  ZEffortj Þ þ eij :
For example, the central question of this study was concerned
with how self-efficacy affected effort, for both a subject over time Level 2
and comparing that subject’s aggregated effort to those of other p0j ¼ b00 þ b01 ðM SEj  M SEÞ
athletes in the study (25). Without using HLM, a researcher is þ b02 ðM ZEffortj  M ZEffortÞ þ r0j ;
left with 2 choices: to disaggregate all variables to the lowest
level of measurement (i.e., within-person measures over each p1j ¼ b10 ; p2j ¼ b20 ; p3j ¼ b30 :
period of time) or to aggregate all the within-person measures to At level 1, Yij is the effort at Time i for athlete j, p0j is the
the higher level. Both options have significant drawbacks; the level 1 intercept, p1j is the effect of TIME (i.e., 4 data
first violates the interdependence assumption required for collection points) for each athlete, p2j is the effect of self-
statistical analysis, because different measures of the same efficacy for each athlete, p3j is the effect of residualized past
person are obviously related. The second disregards any within- effort for each athlete, and eij is the error associated at each
person information, which may make up a significant portion of time point for each athlete measured. At the between-
the total explained variance and have the ability to answer the person level (level 2), all level 1 variables become outcome
research question(s) of interest (38). Second, this study also variables; thus, b00 is the average intercept for all athletes,

VOLUME 26 | NUMBER 2 | FEBRUARY 2012 | 447

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 Self-Efficacy and Effort

b01 is the effect of mean self-efficacy of all athletes, b02 is the past performance (i.e., the exact effort he or she earned at the
effect of mean residualized past effort of all athletes, r0j is previous time point and not the residual version of that effort)
the level 2 error associated with the intercept, b10 is the was analyzed to see if it predicted one’s future self-efficacy. The
effect of linear practice or TIME, b20 is the effect of self- results revealed that this relationship was nonsignificant t(293) =
efficacy, and b30 is the effect of residualized past effort. 20.244, p . 0.05, thus rebuking previous work (31,32) and
Finally, because the advantage of HLM is that level 1 and further validating the contention that past behavior will not be
level 2 models are analyzed simultaneously, the complete the direct cause of future self-efficacy or current behavior (1).
model tested was Second, self-efficacy was again examined as the dependent
variable, and scores from subjects’ VQ and CSI were imputed
EFFORT ¼ b00 þ b01 ðM SEj  M SEÞ into a new model to determine if these psychological variables
affected self-efficacy, which then affect effort. The findings
þ b02 ðM ZEffortj  M ZEffortÞ interestingly showed that at level 2 (i.e., aggregated data), only
þ b10 ðTIMEÞ þ b20 ðSEij  SEj Þ a subject’s value of strength training predicted greater self-
efficacy scores when compared with other athletes t(96) = 5.581,
þ b30 ðZEffortij  ZEffortj Þþ r0j þ eij : p , 0.01. At level 1 (or within-person data analysis), the changes
in value had no effect on a subject’s self-efficacy; however, stress
Findings from this model are presented in Table 2. The
did. As one might expect, as stress levels declined for the same
results show that when data for each athlete were aggregated
athlete over time, their self-efficacy increased, t(291) = 21.94,
and compared with those of other subjects, self-efficacy
p = 0.05. Besides an actual reduction of stress associated with
predicted greater effort in strength training, t(96) = 16.05, p ,
subjects’ lives, this decline in stress levels for some athletes
0.001. At level 1 (or comparing each athlete to their previous
during the course of this study may also be attributed to
psychological scores and effort), the effect of time had no
increased familiarity with the measures completed; however, it
significant impact on a subject’s effort in strength training,
should be noted that when individuals have no incentive to
simply demonstrating that effort did not meaningfully
manipulate responses, self-report measures are an accurate way
change over the course of this study, absent of other
to assess psychological constructs (1). Thus, the decrease in
variables. In contrast, self-efficacy and effort again showed
stress levels for some athletes would most likely be related to
a positive and significant relationship, t(291) = 1.94, p = 0.05
perceived stress levels and not manipulated levels of stress
at the within-person level. Thus, the more each athlete
caused by this study.
became confident in his or her ability to display high effort,
the more effort that athlete subsequent displayed, in both the DISCUSSION
self-report ERQ and the ERQ completed by the strength and
The purpose of this study was to determine how self-efficacy
conditioning coach. It is important to note that this
affected athletes’ effort in strength training over time, because of
relationship held true when past performance was residual-
the recent debate regarding this matter (3,9,13,30–32,38). The
ized, a procedure that is recommended by researchers (3,4,9).
results confirmed our hypothesis, in that self-efficacy was
Finally, no other psychological or demographic variables
positively related to effort in strength training sessions, at both
collected were significant in the prediction of effort at level
within- and between-person levels of analysis. As stated in the
1 or level 2.
Introduction of this article, this result is not surprising, given
Another advantage of HLM v6.04 (26) is the ability to
the vast amount of research in sport settings confirming the
account for how much of the variance in subjects’ effort can be role of self-efficacy in producing adaptive behaviors (23).
explained by the measured constructs. As shown in Table 3, Although previous studies have examined this relationship in
57.9% of the variance in effort could be explained with all a strength training or a weightlifting environment (14,27,36),
significant predictors (e.g., TIME, self-efficacy, and residualized this was the first study to explore the relationship between self-
past performance). Level 1 variables accounted for only 2.3% of efficacy and highly skilled athletes over the course of a training
within-person variance or 0.7% of the total difference in effort cycle. Thus, because of these methodological advancements,
(after multiplying by 1 2 ICC). The explained variance in effort this study can now support research findings in other sports
at the between-person level was 82.9% (or 57.2% of the total (e.g., crew, hockey, tennis, etc.), which showed that positive
variance) with all variables entered. In summary, the relation- changes in self-efficacy will result in increased performance or
ship between self-efficacy and effort was positive and more adaptive behaviors (8–10,15–18,21,23).
significant at both levels of analysis, supporting well- This study also examined the self-efficacy from dual
documented findings that if self-efficacy levels improve for perspectives of the athletes’ and the strength and conditioning
a subject, his or her impending effort in strength training will coaches’ perceptions of effort. Although this methodology of
also show a corresponding and positive change. including coaches’ evaluations of athletes’ effort could be
Although not the main hypotheses of the study, 2 subsequent considered adding a confounding variable to this study, in
HLM v6.04 models were run to fully examine how self-efficacy fact, the inclusion of coaches’ ratings in this study prevented
was manifested over time (26). In the first, a subject’s level 1, raw the athletes from rating themselves as high in strength
the TM

448 Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 the TM

Journal of Strength and Conditioning Research | www.nsca-jscr.org

training effort efficacy and then simply reporting that they flexibility (through specific joint mobility testing), and even
displayed maximum effort to corroborate their previous cardiovascular endurance and speed-endurance (through spe-
efficacy beliefs. Therefore, the importance of self-efficacy was cific team conditioning tests) at the conclusion of their training
not just isolated to athletes’ perceptions, because the positive cycle.
relationship between self-efficacy and effort perceptions held Additionally, researchers could also examine the training
true when using strength and conditioning coaches’ effort logs of athletes during data collection time points. Although
perceptions for athletes to comprise 50% of a subject’s psychological measures are reliable for respondents when
complete effort, a novel step in research for this domain. there is no incentive to embellish answers (1), corroborating
In addition to discovering a positive relationship between self-efficacy levels with intensity percentages during a specific
self-efficacy and effort for collegiate athletes in a strength and portion of the training cycle would aid researchers and
conditioning setting, this study also helped clarify the role of practitioners with additional information athletes hold re-
past performance (or in this case, past effort) on future self- garding their perceptions of past workouts. To date and to our
efficacy levels. In particular, previous researchers have found knowledge, this link between these psychological and
that a subject’s past performance will be the direct cause of his physiological aspects for competitive athletes over the course
or her future self-efficacy levels (31,32). However, in these of a training cycle has not been investigated.
studies, researchers had either examined in self-efficacy via In addition to crossdisciplinary work that incorporates the
computer simulations, thereby only measuring self-efficacy at aforementioned psychological and physiological measures of
purely cognitive levels (3,31,32), or had drawn conclusions athletes, future studies should explore how self-efficacy
regarding physical performances over time that were devoid directly affects the strength training performance (not just
of self-efficacy measurements altogether (22). By incorpo- effort) of athletes, from a multilevel modeling approach.
rating a real-world task (i.e., not manipulating the environ- Although it is hypothesized that self-efficacy relates more
ment for subjects), results contradicted these previous strongly to behavior—when compared with performance
assertions and showed that past effort was nonsignificant (1)—this proposed link would greatly aid the strength and
and thus did not directly affect an athlete’s impending self- conditioning professionals’ ability to motivate athletes and
efficacy for strength and conditioning effort. This finding solidify the importance of this psychological variable when
supports the notion that past behavior will not be the direct training for sport. However, for this relationship between
cause of current behavior, as instead, behavior is cognitively confidence and effort to be fully understood, future research
‘‘filtered’’ through current efficacy states to produce beliefs should also place equal emphasis on the recruitment of
about present capabilities (1,13). All of the aforementioned female athletes during the design of potential studies. Besides
findings are important for practitioners because this study has the sheer number of female athletes and female sports at the
shown that self-efficacy levels result in greater perceptions of collegiate level of competition, these sports also potential
effort from athletes and strength and conditioning coaches, revenue sources for highly skilled teams; thus, the aspects of
and furthermore, subjects’ past effort (either high or low) will performance for all athletes should not be overlooked.
not impact future levels of self-efficacy or effort.
One limitation of this study was its exclusive collection of PRACTICAL APPLICATIONS
psychological data. As strength and conditioning coaches will Because self-efficacy was shown to be more influential in effort
attest to, training for sport does not exist in a vacuum where only than were stress levels, value, or demographic variables (19),
mental factors affect behavior and performance. Future coaches should work to increase the confidence levels of their
endeavors in this subject could observe the possible connection athletes. First and foremost, coaches should structure the
between self-efficacy and physiological measures of perfor- environment to allow for mastery experiences (i.e., configuring
mance. Professionals rely on quantitative data to track their the task in a manner that lends itself to success). Researchers
athletes’ physical developments throughout the training year. studying sports ranging from baseball to wrestling have found
Most strength and conditioning professionals assess, then that successful past performances increased subjects’ confi-
reasses, the physiological profile of their athletes throughout dence, which thereby increased their future performance
their training year. Therefore, it would be interesting to note the (15,18,23,34). To aid in mastery skill development, complex
self-efficacy of athletes between assessment and reassessment of skills can be broken down into manageable parts to better
athletic measures in their training year. For example, the self- facilitate mastery; then, performance aids can be gradually
efficacy of athletes could be assessed before their off-season removed when athletes become more comfortable with the
training cycle, which would coincide with an assessment of their skill set needed to accomplish a task (8,13). An example of this
current athletic profile. The athletes would progress through technique can be employed with athletes struggling in 1RM
their designed training cycle where their self-efficacy could be power clean testing. At the appropriate times during the
periodically noted. Finally, athletes’ perceptions of their effort training cycle, coaches can have athletes gain confidence by
would be correlated to their reassessed athletic profile such as performing clean pulls, clean high pulls, and drop cleans to
muscular strength (via 1RM testing), muscular power (through ‘‘learn how the weight feels’’ and master the correct technique.
measured jumps, medicine ball throws, and timed jumps), This alteration of the environment will produce athletes who

VOLUME 26 | NUMBER 2 | FEBRUARY 2012 | 449

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 Self-Efficacy and Effort

are more comfortable with the heavier weight, or technique feeling fine and more time devoted to changing their
required, and will facilitate higher confidence during future perceptions—by perhaps reminding them that they have
training sessions and 1RM testing (1). been tired before and still produced high levels of effort when
A second way to build confidence is through vicarious necessary. It should be noted however, that this last source
experiences (or social models), wherein individuals observe of self-efficacy is generally regarded as the least effective,
other people, similar to themselves, succeeding. Studies have in altering confidence levels of athletes because people
again confirmed the link between vicarious experience and generally find the other sources of self-efficacy more
increased confidence (11); however, the benefits are diagnostic of their personal capabilities (1). Thus, coaches
contingent on the modeler. Specifically, modelers who have should focus on the previous 3 sources—especially mastery
the same ability or a slightly higher ability provide the most experiences—to improve the confidence (and resulting effort)
useful feedback in terms of judging one’s own capabilities of their athletes in strength and conditioning.
(13). Therefore, athletes who need to increase their self-
efficacy levels in strength and conditioning should seek out
others who have slightly higher performances or better REFERENCES
technique as a benchmark to measure progress. For instance, 1. Bandura, A. Self-Efficacy: The Exercise of Control. New York, NY:
Freeman, 1997.
if strength and conditioning coaches find specific athletes
2. Bandura, A. Adolescent development from an agentic perspecctive.
lacking in effort production, they could pair these athletes In: Self-Efficacy Beliefs of Adolescents. F. Pajares, and T. Urdan, eds.
with others who give slightly more effort. By following this Greenwich, CT: Information Age Publishing, 2006. pp. 1–43.
protocol, athletes who give subpar effort will observe the 3. Bandura, A and Locke, EA. Negative self-efficacy and goal effects
modeler become more proficient at a skill or task and receive revisited. J Appl Psychol 88: 87–99, 2003.
praise, which will trigger thoughts of, ‘‘If he or she can do 4. Bandura, A and Wood, RE. Effect of perceived controllability and
this, I can too.’’ performance standards on self-regulation of complex decision-
making. J Pers Soc Psychol 56: 805–814, 1989.
A third way that confidence can be improved is through
5. Chase, MA. Competition plans and performance routines. In:
verbal (or social) persuasion or when significant others Encyclopedia of International Sports Studies. R. Bartlett, C. Gratton, and
attempt to verbally persuade individuals that they have the C. Rolf, eds. London, United Kingdom: Taylor and Francis, 2006.
skills or ability necessary to be successful at a task (33). pp. 290–292.
Although coaches are already no doubt proficient with this 6. Elston, TL and Martin-Ginis, KA. The effects of self-set versus
assigned goals on exercisers’ self-efficacy for an unfamiliar task.
suggestion, caution must be exercised by practitioners for 2
J Sport Exerc Psychol 26: 500–504, 2004.
reasons. First, the debilitating effects of verbal persuasion are
7. Feltz, DL. Gender differences in the causal elements of self-efficacy on
stronger than the enhancing effects (1). Thus, a coach can a high-avoidance motor task. J Sport Exerc Psychol 10: 151–166, 1988.
more quickly create doubt in an athlete by suggesting that the 8. Feltz, DL. Self-confidence and performance. In: Learning,
athlete lacks the ability to put in effort than he or she can Remembering, Believing: Enhancing Human Performance. D. Druckman
create enhanced beliefs through positive suggestions. Second, and R.A. Bjork, eds. Washington, DC: National Academy Press,
1994. pp. 173–206.
the athlete must trust the source of the verbal persuasion for
9. Feltz, DL, Chow, GM, and Hepler, TJ. Path analysis of self-
this method to produce increased confidence (11). For
efficacy and diving performance revisited. J Sport Exerc Psychol
example, if a teammate conveys that there is 235 lbs on the 30: 401–411, 2008.
bar (during a 3RM bench press) and demands effort in 10. Feltz, DL and Lirgg, CD. Perceived team and player efficacy in
competing the set, but there is actually 245 lbs and the athlete hockey. J Appl Psychol 83: 557–564, 1998.
later finds out, this athlete will be less likely to respond to this 11. Feltz, DL and Lirgg, CD. Self-efficacy beliefs of athletes, teams, and
type of motivation—with increased effort levels—from the coaches. In: Handbook of Sport Psychology. R.N. Singer,
H.A. Hausenblas, and C.M. Janelle, eds. New York, NY: John Wiley
teammate in the future. & Sons, Inc., 2001. pp. 340–361.
Finally, somatic or emotional states (i.e., how one’s body 12. Feltz, DL and Magyar, MT. Self-efficacy and adolescents in sport and
feels) can alter athletes’ confidence when interpretations of physical activity. In: Self-Efficacy Beliefs of Adolescents. F. Pajares and
these physiological signs are linked to an impending T. Urdan, eds. Greenwich, CT: Information Age Publishing, 2006.
performance (13). Specifically, perceptions such as tiredness, pp. 161–179.
heart pounding, sweaty palms, fatigue, and even how a recent 13. Feltz, DL, Short, SE, and Sullivan, PJ. Self-Efficacy in Sport.
Champaign, IL: Human Kinetics, 2008.
injury feels can positively or negatively affect confidence and
14. Fitzimmons, PA, Landers, DM, Thomas, JR, and van der Mars, H.
effort in a skill or task (1,7). Even though coaches will have Does self-efficacy predict performance in experienced weightlifters?
a difficult time trying to convince an athlete that he or she is Res Q Exerc Sport 62: 424–431, 1991.
not nervous before a big game, if the athlete’s perceptions of 15. George, TR. Self-confidence and baseball performance: A causal
nervousness can be changed from fear and anxiety to being examination of self-efficacy theory. J Sport Exerc Psychol 16: 381–
399, 1994.
prepared and ready, positive changes in self-efficacy should
result (5,13). For instance, if athletes come into a training 16. Greenlees, IA, Bradley, A, Holder, T, and Thelwell, RC. The impact
of opponents’ non-verbal behaviour on first impressions and
session feeling tired and lethargic, coaches should spend less outcome expectancies of table tennis players. Psychol Sport Exerc
time trying to convince these athletes that they are actually 6: 103–115, 2005.
the TM

450 Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 the TM

Journal of Strength and Conditioning Research | www.nsca-jscr.org

17. Greenleas, IA, Buscombe, R, Thelwell, RC, Holder, T, and Rimmer, M. 27. Rhea, MR, Landers, DM, Alvar, BA, and Arent, SM. The effects of
Perceptions of opponents in tennis: The impact of opponents’ clothing competition and the presence of an audience on weight lifting
and body language on impression formation and outcome expect- performance. J Strength Cond Res 17: 303–306, 2003.
ations. J Sport Exerc Psychol 27: 39–52, 2005. 28. Singer, JD and Willett, JB. Applied Longitudinal Data Analysis.
18. Kane, TD, Marks, MA, Zaccaro, SJ, and Blair, V. Self-efficacy, New York, NY: Oxford University Press, 2003.
personal goals, and wrestlers’ self-regulation. J Sport Exerc Psychol 29. Stone, DN. Overconfidence in initial self-efficacy judgments: Effects
18: 36–48, 1996. on decision processes and performance. Organ Behav Hum Decis
19. Locke, KD and Sadler, P. Self-efficacy, values, and complementarity Process 59: 452–474, 1994.
in dyadic interactions: Integrating interpersonal and social-cognitive 30. Vancouver, JB and Kendall, LN. When self-efficacy negatively relates
theory. Pers Soc Psychol Bull 33: 94–109, 2007. to motivation and performance in a learning context. J Appl Psychol
91: 1146–1153, 2006.
20. MacLean, D and Sullivan, P. A season long case study investigation
of collective efficacy in male intercollegiate basketball. Athl Insight 31. Vancouver, JB, Thompson, CM, Tischner, EC, and Putka, DJ. Two
Online J Sport Psychol 5: 2003. http://www.athleticinsight.com/ studies examining the negative effect of self-efficacy on performance.
Vol5Iss3/CollegeBasketballCaseStudy.htm. J Appl Psychol 87: 506–516, 2002.

21. Magyar, TM, Feltz, DL, and Simpson, IP. Individual and crew level 32. Vancouver, JB, Thompson, CM, and Williams, AA. The changing
signs in the relationships among self-efficacy, personal goals, and
determinants of collective efficacy in rowing. J Sport Exerc Psychol
performance. J Appl Psychol 86: 605–620, 2001.
26: 136–153, 2004.
33. Vargas-Tonsing, TM. An exploratory examination of the effects of
22. Mizruchi, MS. Urgency, motivation, and group performance: The coaches’ pre-game speeches on athletes’ perceptions of self-efficacy
effect of prior success on current success among professional and emotion. J Sport Behav 32: 92–111, 2009.
basketball teams. Social Psychol Q 54: 181–189, 1991.
34. Vargas-Tonsing, TM, Myers, ND, and Feltz, DL. Coaches’ and
23. Moritz, SE, Feltz, DL, Fahrbach, KR, and Mack, DE. The relation of athletes’ perceptions of efficacy enhancing techniques. Sport Psychol
self-efficacy measures to sport performance: A meta-analytic review. 18: 397–414, 2004.
Res Q Exerc Sport 71: 280–294, 2000.
35. Vealey, RS and Chase, MA. Self-confidence in sport. In: Advances in
24. Myers, ND, Feltz, DL, and Short, SE. Collective efficacy and team Sport Psychology. T.S. Horn, ed. Champaign, IL: Human Kinetics,
performance: A longitudinal study of collegiate football teams. Group 2008. pp. 65–97.
Dynamics: Theory Res Pract 8: 126–138, 2004. 36. Wells, C, Collins, D, and Hale, H. The self-efficacy-performance link
25. Pfeiffer, KA, Dowda, M, Dishman, RK, Sirard, JR, and Pate, RR. in maximum strength performance. J Sports Sci 11: 167–175, 1993.
Cardiorespiratory fitness in girls—Change from middle to high 37. Wilkes, RL and Summers, JJ. Cognitions, mediating variables and
school. Med Sci Sports Exerc 39: 2234–2247, 2007. strength performance. J Sport Psych 6: 351–359, 1984.
26. Raudenbush, SW, Bryk, AS, and Congdon, R. Hierarchical Linear and 38. Yeo, GB and Neal, A. An examination of the dynamic relationship
Nonlinear Modeling [software]. Lincolnwood, IL: Scientific Software between self-efficacy and performance across levels of analysis and
International, 2007. levels of specificity. J Appl Psychol 91: 1088–1101, 2006.

VOLUME 26 | NUMBER 2 | FEBRUARY 2012 | 451

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.