Psychology of Sport and Exercise 15 (2014) 521e527
Contents lists available at ScienceDirect
Psychology of Sport and Exercise
journal homepage: www.elsevier.com/locate/psychsport
Athletics and executive functioning: How athletic participation and
sport type correlate with cognitive performance
Jed Jacobson*, Leland Matthaeus 1
Whitman College, 345Boyer Ave, Walla Walla, WA 99362, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 23 June 2013
Received in revised form
18 April 2014
Accepted 21 May 2014
Available online 2 June 2014
Objectives: This study aims to further the knowledge regarding the documented link between physical
exercise and cognitive function. Specifically, we examined the relationship between the type and level of
sports in which college students participate and their executive functioning (EF).
Design: We utilized a 3-way quasi-experimental design, and grouped participants by athletic status
(athlete or non-athlete), sport type (self-paced, externally paced, or non-athlete; see Singer, 2000), and
level (high-skilled or recreational).
Methods: We evaluated EF by administering a battery of validated tests of decision making, problem
solving, and inhibition.
Results: We found that athletes scored higher on some of the EF measures than non-athletes. Furthermore, we observed that scores varied by sport type according to which subset of EF each test measured.
Self-paced athletes scored highest on an inhibition task, and externally paced athletes scored highest on
a problem-solving task.
Conclusions: Our results suggest that athletes outperform non-athletes on tests of such EF domains as
inhibition and problem solving, and that different types of athletic experience may correlate with higher
levels of particular EF domains.
© 2014 Elsevier Ltd. All rights reserved.
Keywords:
Executive functioning
Problem solving
Inhibition
Self-paced sports
Externally paced sports
Many researchers have sought to explore the ways the body and
mind influence one another. Recent studies (e.g., Keating, Castelli, &
Ayers, 2013) have established a link between exercise and cognitive
proficiency. After sessions of acute physical exercise, people tend to
score higher on cognitive tests than when they have not exercised
(Etnier & Chang, 2009). Of more significance to our study, elite
athletes appear to perform with higher proficiency on tasks testing
executive functioning (EF), a subcategory of cognitive functioning
(Vestberg, Gustafson, Maurex, Ingvar, & Petrovic, 2012). Situations
that require EF include activities that involve effortful problem
solving, inhibition, planning, or vigilance (Diamond, 2006). Executive functions are highly utilized both in goal-oriented action
under distraction and in novel response production when habitual
dominant responses are apparent (Unsworth et al., 2009).
Abbreviations: EF, executive functioning; CST, cognitive skill transfer; DM, decision making; DMA, decision-making accuracy; DMS, decision-making speed; PS,
problem solving; MPS, mental processing apeed; SP, self-paced; EP, externally
paced.
* Corresponding author. Present address: 4840 86th Ave SE, Mercer Island, WA
98040, USA. Tel.: þ1 (206) 375 3401.
E-mail addresses: jedediah.jacobson@gmail.com (J. Jacobson), matthalp@
whitman.edu (L. Matthaeus).
1
Tel.: þ1 (425) 445 4369.
http://dx.doi.org/10.1016/j.psychsport.2014.05.005
1469-0292/© 2014 Elsevier Ltd. All rights reserved.
Researchers break EF into more specific mental capacities such as
problem solving, planning, inhibition, and decision making, in order to operationally measure it (Diamond, 2006; Spreen & Strauss,
1998). Researchers have been assessing both the relationship between exercise and EF, and the relationship between sport training
and cognition, and many researchers are working to unite these
two related lines of research (Pesce, 2012). To our knowledge, no
previous researchers have tested hypotheses regarding the relationship between EF and specific types of sports; we aim to
examine this relationship in our study. We seek to illuminate the
relationship between sports and cognitive performance, which
may have implications for athletic programs and physical education. If certain sports correlate with higher cognitive ability more
than others, then physical educators, coaches, and policy-makers
may seek to emphasize certain activities in athletics and
throughout development.
The documented differences in EF performance between elite
athletes and non-athletes (e.g., Vestberg et al., 2012) may be
attributed to cognitive skill transfer (CST), the process by which
training in a cognitive task may improve performance on related
untrained cognitive tasks. Various models for how CST works have
been proposed, and, although they differ, most theorists agree that
every task consists of many skills and/or pieces of knowledge, and
522
J. Jacobson, L. Matthaeus / Psychology of Sport and Exercise 15 (2014) 521e527
that tasks that share more skills and knowledge will have the
strongest transfer effects in both the short and longterm (Taatgen,
2013). Researchers disagree on how “far” these skills can transfer;
that is, the degree of difference between two given tasks for which
transfer effects may be present. For example, chess players appear to
have improved working memory capacity with regard to arrangements of chess pieces, but their overall working memory is normal
(Chase & Simon, 1973). This finding supports the “narrow transfer”
hypothesis: the idea that individuals with expertise in a particular
field may have superior cognitive processes within that field, but not
necessarily outside of it (Furley & Memmert, 2011). Contradictorily,
video game training appears to cause improvements in scores on
laboratory reaction time tests (Green, Pouget, & Bavelier, 2010). This
finding supports the “broad transfer” hypothesis: the theory that
extensive practice of context-specific skills improves individual
components of cognition, and that these improvements are present
regardless of context (Furley & Memmert, 2011).
With regard specifically to sports, a parallel debate is ongoing.
Voss, Kramer, Basak, Prakash, and Roberts (2010) made a distinction between two paradigms for analyzing improved cognition in
athletes. Researchers using the “expert performance” approach
have investigated the idea that athletes have improved cognition
within their sport (e.g., Mann, Williams, Ward, & Janelle, 2007;
Singer, Cauraugh, Chen, Steinberg, & Frehlich, 1996). On the other
hand, proponents of the “cognitive component skills” approach
assert that athletes improve in specific cognitive skills, which are
present in non-sport contexts and can be measured in the laboratory (Voss et al., 2010). Although much research has supported this
idea with respect to skills like attention (Anzeneder & Bosel, 1998;
Pesce, Cereatti, Casella, Baldari, & Capranica, 2007), many studies
have found no benefits for athletes on related cognitive skills (e.g.,
Lum, Enns, & Pratt, 2002). Research on the transfer of EF skills from
athletics is sparse (Voss et al., 2010); we propose that differences in
EF may be present between athletes and non-athletes, and that CST
may play a role in this.
Researchers have employed various methods in the pursuit of
improving EF (e.g., Diamond & Lee, 2011; Kesler et al., 2013). In a
study of disabled athletes (Di Russo et al., 2010), the researchers
suggested that overall EF was lower in physically disabled populations, and that playing certain sports such as basketball, in
which the athlete interacts with his/her environment on a
constantly changing basis, may reduce this deficit by promoting
response flexibility. On the other hand, sports like swimming,
where athletes are not thinking creatively or reacting to timepressured stimuli, did not seem to benefit participants in this
manner (Di Russo et al., 2010). These results indicate that different
types of sports may differentially facilitate EF improvements.
Contrastingly, in a meta-analysis of fitness training and cognitive
function (Colcombe & Kramer, 2003), the authors suggested that
particularly aerobic fitness training is likely to have positive effects
on EF. In our study, we aim to address not only the effects of mere
aerobic exercise on cognition, but also the sport-specific mental
skills that may be related to differences in particular aspects of EF.
In a literature review of sports psychology (Singer, 2000), focused
on improving performance, the researcher asserted that sports
could be classified into two categories: self-paced (SP) and externally paced (EP). Sports like bowling, golf, and running, as well as
aspects of sports like baseball pitches and tennis serves, were classified as SP because they allow time for the athletes to prepare
themselves for critical actions and perform at a pace they control.
Sports like soccer, basketball, and volleyball require adaptability and
quick decision making in response to external cues, and are labeled
as EP. Two sub-categories of EP sports are interceptive sports, such as
racquet sports, and strategic sports, or sports which involve multiple
teammates and opponents and tactical formations (Mann et al.,
2007). The EP vs. SP sport distinctions were based partly upon
Singer's (1988) experiment, geared towards understanding how to
improve performance in self-paced athletics. Later, Singer (2000)
developed a strategy to improve performance in EP sports, highlighting attention and decision making. EP athletes, especially those
playing at a high level, may have faster and more accurate decisionmaking processes (Singer et al., 1996; Zoudji, Thon, & Debû, 2010).
We theorized that the categorical distinction between self- and
externally paced athletes would correspond to differential performances on EF tasks testing skills like inhibition and decision making.
Furthering the notion of athletics correlating with higher levels
of cognitive performance, Vestberg et al. (2012) evaluated the
relationship between EF and athletic ability in a healthy population
of elite athletes (Swedish professional soccer players) and nonathletes. They compared the participants' scores on tasks
measuring creativity, inhibition, and cognitive flexibility, finding
significant variation between two levels of athletes and a control
group of non-athletes (Vestberg et al., 2012). These results indicate
a positive correlation between EF and athletic ability. Based on this
finding, we drew a distinction in our study between high-skilled
and recreational athletes; we expected that athletes with more
expertise would score higher on EF tests. In the Swedish study, the
researchers later tracked the athletes through two seasons and
analyzed their goal and assist statistics (the most objective measure
of soccer success). They found that players who had scored higher
on the EF tasks scored and assisted more goals than those who had
not performed as well on the tasks (Vestberg et al., 2012). Therefore, high EF may predict athletic success in EP sports like soccer.
Below, we elaborate on the facets of executive functioning that
we expected to correlate most strongly with each of the classifications of athletics. These constructs include decision making (DM),
problem solving (PS), and inhibition. Decision making is “the
cognitive process of choosing between two or more alternatives,
ranging from the relatively clear cut … to the complex” (VandenBos,
2006, p. 259). It utilizes the executive functions of shifting, planning, and categorization (Brand et al., 2005). Adaptive decision
making has been closely linked with higher-level EF in military
leaders (Hannah, Balthazard, Waldman, Jennings, & Thatcher,
2013). Because EF has such a strong association with DM, we predicted that the higher EF seen in athletes (Vestberg et al., 2012)
would correlate with improvements in their DM abilities.
Problem solving is “the process by which individuals attempt to
overcome difficulties, achieve plans that move them from a starting
situation to a desired goal, or reach conclusions through the use of
higher mental functions, such as reasoning and creative thinking”
(VandenBos, 2006, p. 735). The relationship between EF and PS is
well documented (e.g., Harris, 2001; Kotsopoulos & Lee, 2012). PS
highly utilizes shifting and updating, two widely accepted foundational executive functions (Diamond, 2013; Kotsopoulos & Lee,
2012). Improvements in EF appear to correlate with improved PS
abilities in healthy college students (Wen, Butler, & Koutstaal,
2012). Because of this direct link to EF we predicted that, as with
DM, the documented EF proficiency seen in EP athletes (Vestberg
et al., 2012) would correlate with higher PS abilities.
Inhibition is “the suppression of covert responses in order to
prevent incorrect responses” (VandenBos, 2006, p. 481). Inhibition
is a stand-alone executive function according to a number of theorists (e.g., Diamond, 2006; Kotsopoulos & Lee, 2012). Multiple
types and definitions of inhibition exist (Miyake, Friedman,
Emerson, Witzki, & Howerter, 2000); in this study we focus on
the suppression of dominant or prepotent responses. Inhibition
correlates moderately with other executive functions like shifting
and updating, but the functions are separable and differentially
contribute to performance on complex executive tasks (Miyake
et al., 2000). Inhibition can be bettered through practice; in a
J. Jacobson, L. Matthaeus / Psychology of Sport and Exercise 15 (2014) 521e527
sample of children with developmental coordination disorder, a 10week soccer training intervention improved performance on an
inhibition task (Tsai, Wang, & Tseng, 2012).
We hypothesized that athletes would outperform non-athletes
on all EF tasks. We further hypothesized that EP athletes would
score highest on the DMA, DMS, and PS tasks, followed by their SP
and non-athlete counterparts, because EP sports require the athletes
to constantly be adapting to the conditions of the contest and to act
quickly and strategically to outperform the opposing side. Because
of the inhibition required in the highly-regimented practice of tasks
requiring intense focus despite internal distractors such as fatigue
and external distractors such as cheering spectators, we hypothesized that SP athletes would score highest on the inhibition task,
followed by EP athletes and non-athletes. Lastly, we hypothesized
that high-skilled athletes would score higher on all EF tasks than
recreational athletes because they practice more often and perform
at a higher level. We did not expect significant inter-group variation
on the vocabulary task because we recruited students from one
school with constant admission standards, although admissions
standards may be more lenient for incoming varsity athletes.
Methods
523
Design
We utilized a between-groups 3-way quasi-experimental
design. Participants completed four psychometric tests and one
self-report questionnaire. We chose the two EF tests based on their
demonstrated close correspondence to EF, the MPS test because of
its influence on our measures of EF, and the vocabulary test because
vocabulary correlates strongly with overall intelligence. We chose a
quasi-experimental design in which we did not assign participants
to groups, but rather grouped them by preexisting characteristics.
We tested hypotheses regarding differences between athletes and
non-athletes, between self-paced athletes, externally paced athletes, and non-athletes, and between high-skilled and recreational
athletes, in their scores on the EF tasks, the MPS task, and the vocabulary task. We grouped participants as athletes or non-athletes
based on their self-reported sports participation, where participating in sports once or more per week qualified an individual as an
athlete. We grouped participants by sport type (EP or SP) based
upon Singer's (2000) article, and we grouped participants by sport
level (high-skilled or recreational). The high-skilled athlete group
included college varsity, semi-professional, and professional athletes. The statistical calculations in this study were performed in
SPSS.
Participants
Measures
We sampled from a population of students at a small undergraduate college in the western United States. There were 54 participants in total, with ages ranging from 18 to 24 (M ¼ 20.13;
SD ¼ 1.30); participants were predominantly female (57.41%), and
predominantly White (81.48%). Further participant demographics
are given in Table 1. Participants were recruited via the college's email
listservs. Potential participants received an email inviting them to
participate in a study on extracurricular activities and cognitive skills.
The sampling method was quick, convenient, and permitted us to
reach a large number of possible participants. Exclusion criteria
included: a) Self-reported diagnosis of an attention disorder by a
psychologist or psychiatrist; b) Head injury that caused a loss of
consciousness lasting more than one to 2 min; and c) Being under the
influence of a mind-altering substance at the time of testing.
Portions of a few participants' data were excluded because of
either misunderstanding of the instructions or experimenter error
(e.g., Tower Test scores when the experimenter mistakenly omitted
an item; Coding Test scores when the participant ignored the instructions to complete all items in order). This may have led to
some type of immeasurable selection bias, but if so, it was likely not
significant. In adherence to ethical guidelines, participants were
informed that they could choose not to complete any part of the
study and that individuals' data would be kept confidential. The
Whitman College Institutional Review Board approved this study.
Participants were asked not to share information from debriefing.
Table 1
Demographic statistics of the sample.
Athletes (N ¼ 39)
Age
Sex
Race
GPA
M
SD
Male
Female
Caucasian
Latino
Asian
M
SD
Externally
paced (N ¼ 22)
Self-paced
(N ¼ 17)
20.05
1.17
14
8
19
1
2
3.46
.34
20.18
1.51
3
14
15
0
2
3.64
.24
Non-athletes
(N ¼ 15)
To measure EF, we used two validated psychological tests:
namely the DeliseKaplan Executive Function System (D-KEFS)
Tower Test and ColoreWord Interference Test (Delis, Kaplan, &
Kramer, 2001a). To evaluate discriminant validity we administered a vocabulary test in order to establish a baseline of intelligence, and a coding exercise to measure MPS. Lastly we
administered a questionnaire designed to measure participant
demographics as well as place participants into groups. To control
for potential differences between researchers, we ran inter-rater
reliability analyses for each of the tests (see Table 2). Due to time
and equipment constraints, we were only able to analyze interrater reliability for 10 participants' data. Despite the small sample,
the analyses indicated very strong inter-rater reliability for each
test, and we have no concerns about scores being influenced by the
experimenters. For the D-KEFS tests, we videotaped the first ten
participants as they performed the tests and the second scorer
scored the tests independently while watching the video. For the
vocabulary test and the coding test, we simply scored each test
separately and compared scores afterward for the first ten
participants.
D-KEFS tower
The D-KEFS Tower Test (Delis et al., 2001a) consists of nine items
that require the participant to move circular disks on and off pegs to
create a tower that matches a given model. The disks differ in size;
the participant must avoid placing larger disks on top of smaller
ones, and he/she may only touch one disk at a time. The participant
Total
(N ¼ 54)
Table 2
Inter-rater reliability analysis.
20.20
1.27
6
9
10
1
4
3.51
.32
20.13
1.30
23
31
44
2
8
3.53
.31
Inhibition
Decision Making Accuracy
Decision Making Speed
Mental Processing Speed
Vocabulary
k
Std. Error
t
p
1.000
.674
.158
.888
.887
.000
.148
.114
.104
.111
6.58
7.50
2.36
8.23
6.09
<.001**
<.001**
.018*
<.001**
<.001**
*Significant at a ¼ .05, **Significant at a ¼ .009 (Sidak correction for multiple
testing), df ¼ 9.
524
J. Jacobson, L. Matthaeus / Psychology of Sport and Exercise 15 (2014) 521e527
is instructed to construct the tower in the fewest moves possible.
Later items feature more disks and tend to be more challenging. We
obtained three scores from this test: the Total Achievement Score,
the Time-per-Move Ratio, and the Move Accuracy Ratio. The Total
Achievement Score measures overall performance in terms of
building the correct tower in the fewest possible moves, and
demonstrates participants' spatial planning, rule learning, and set
creation and shifting (Delis et al., 2001a, p. 195). Because this task
required the participants to use these skills to solve a complex
puzzle, we employed the age-scaled Total Achievement Score as
our measure of problem solving (PS).
For the purposes of our study, we divided the concept of DM into
two domains: Decision-Making Accuracy (DMA; one's ability to
make desirable decisions among alternatives during goal-oriented
activity) and Decision-Making Speed (DMS; one's ability to
quickly make goal-oriented choices, regardless of accuracy). The
Move Accuracy Ratio provides a measure of how often participants
are able to make correct decisions among alternatives during goaloriented activity. Age-scaled Move Accuracy Ratio scores provided
us with our measure of decision-making accuracy (DMA). The
Time-per-Move Ratio measures how quickly the participant arrives
at a decision for his/her next move; these ratios provided our
decision-making speed (DMS) scores.
The D-KEFS Tower Test has demonstrated strong internal consistency, construct validity, and testeretest reliability, as well as
moderate correlations between the sub-scores of the test (Delis,
Kaplan, & Kramer, 2001b, p. 40). Inter-rater reliability scores
(given by Cohen's Kappa; see Table 2) in our study were as follows:
k ¼ .767, t ¼ 6.25, p < .001 (Total Achievement Score); k ¼ .674,
t ¼ 7.50, p < .001 (Move Accuracy Ratio); and k ¼ .158 t ¼ 2.36
p < .018 (Time-per-move Ratio). The D-KEFS Tower is a modified
version of a long-standing Tower of Hanoi test, also known as the
Tower of London. It has demonstrated construct validity in the
detection of brain damage (particularly frontal lobe dysfunction) as
well as measuring important areas of higher-level executive functions (Delis et al., 2001b).
D-KEFS ColoreWord Interference
The D-KEFS Color-Word Interference Test (Delis et al., 2001a), a
modified Stroop Test, consists of four items (participants were only
asked to complete three), which require the participant to name ink
colors and read color names. The first condition, color naming,
contains 50 colored boxes (some red, some green, and some blue).
The participants named the colors one by one as fast as possible.
The second condition, word reading, has 50 names of the same
three colors all written in black ink, and the participant must read
these. The third condition, inhibition, contains 50 color names
written in different colored ink, and requires the participant to
inhibit the more salient response of reading the word, and instead
name the dissonant ink color. From this test we obtained the inhibition score by subtracting the scaled scores from the first condition from the scaled scores on the third. This score shows verbal
inhibition controlling for color naming speed and participant age
(Delis et al., 2001a).
The authors reported high internal consistency and testeretest
reliability scores on the ColoreWord Interference test (Delis et al.,
2001b, p. 40). The inter-rater reliability statistics in our study
were k ¼ 1.00, t ¼ 6.58, p < .001. The test's documented ability to
detect brain damage and evaluate higher-level EF, much like the
Tower, indicates strong construct validity (Delis et al., 2001b).
Coding test
We also measured a related construct that works in tandem
with EF but is not considered a subcategory of EF: mental processing speed (MPS). MPS is the speed of cognitive functioning and
is used as a measure of the efficiency of a variety of cognitive
abilities, including EF (Baudouin, Clarys, Vanneste, & Isingrini,
2009). As such, it was a prime candidate for a covariate to partial
out in our planned analyses. We controlled for MPS to ensure that
our EF variables, not other mental factors, were accounting for
potential observed EF differences. Researchers (e.g., Baudouin et al.,
2009) frequently use coding exercises to measure MPS. In these
exercises, the participant learns pairs of corresponding numbers
and symbols and uses this information to fill in blanks.
The coding test we used is an equivalent version of the Digit
Symbol Substitution Test (DSST, Wechsler, 1997), a highly reliable
and frequently used test of processing speed (Baudouin et al., 2009;
Dickinson & Gold, 2008). Participants were shown a code table with
pairs of digits (0e9) and symbols, and rows of empty boxes with a
digit from 0 to 9 in the top corner of the box. The task requires the
participant to use the code table to write the symbol associated
with each digit in the box. The participants had two min to write as
many symbols as possible in the empty boxes below each corresponding digit. The digits were randomized using an online
random number generator to ensure that no patterns or redundancies could sway the results. The MPS score was given by the
number of symbols correctly coded by a participant in this exercise.
Our inter-rater reliability analysis showed k ¼ .888, t ¼ 8.23,
p < .001.
Vocabulary test
Because EF measures correlate strongly with tests of overall
intelligence, such as the WAIS-III (Davis, Pierson & Finch, 2011), we
attempted to statistically control for overall intelligence while
evaluating EF. We employed a vocabulary test as a brief intelligence
measure because in recent studies (e.g., Smith, Smith, Taylor, &
Hobby, 2005) researchers have found that vocabulary tests correlate strongly with every section of standardized intelligence tests.
In a correlational analysis of 243 7e17 year old learning-disabled
students' performances on standardized vocabulary tests and IQ
tests (Smith et al., 2005), scores on the vocabulary test correlated
highly with Full Scale IQ (r ¼ .74; p < .01).
The vocabulary test consists of 15 items, each ranked equally on
the scoring scale. The words were selected from a list Graduate
Records Examinations (GRE) vocabulary words; selected to represent different parts of speech, different semantic categories and
different affective valences (e.g., turbid, a goad, to preclude). The
participants were not under time constraints and were asked to
define every word to the best of their ability. We calculated the total
score by adding together the fifteen item scores. Scoring criteria are
based on the possible definitions listed in the American Heritage
Dictionary (Morris, 1976). This test resembles those used in the
WAIS-IV (Wechsler, 2008). Each word is scored on a two-point scale;
a correct definition received two points, a partially correct definition
received one point, and an incorrect definition received zero points.
Participants' vocabulary scores were the total number of points
scored. Our inter-rater reliability was k ¼ 887, t ¼ 6.09, p < .001.
Demographic questionnaire
The questionnaire was designed for the dual purposes of sorting
the participants into groups depending on their athletic level/type
and obtaining some basic demographic information regarding age,
race, and sex. We also asked a number of questions to capture
potential confounding variables such as time of day, the participants' sleep, and the participants' caffeine consumption.
Procedure
The participants were tested in a laboratory. They were
informed that they would be performing cognitive tests for a study
J. Jacobson, L. Matthaeus / Psychology of Sport and Exercise 15 (2014) 521e527
on extracurricular activities and EF, and that the study would take
approximately 30 min. Tests were administered in the following
order for all participants: 1) D-KEFS ColoreWord Interference; 2)
coding exercise; 3) D-KEFS Tower; 4) vocabulary test; 5) demographic questionnaire. To mitigate potential fatigue/boredom
effects, participants were given a 1-min break between the coding
exercise and the Tower test. For the same reasons, the vocabulary
test, which allows participants to relax and take their time, was
administered last. After testing, the participants were offered an
incentive (baked goods and/or fruit) and debriefed about the study.
Results
We removed two outliers from the data. Two non-athletes
scored extremely high on inhibition, to the extent that their data
were statistical outliers within their own group and within the
entire sample. These participants may have misunderstood the
instructions, and differentially weighted speed and accuracy on
condition 1 vs. condition 3. This could cause a normal inhibition
score to appear very high when compared with a very low word
reading score. In our hypothesis testing, we set alpha levels at .05
for all tests.
We used independent-samples t tests to test the hypothesis that
athletes and non-athletes would differ on the EF tasks. Some of our
hypotheses received support. As shown in Table 3, scores on the
inhibition task show significant mean variation between athletes
and non-athletes. Athletes also outscored non-athletes on our
measure of PS, the Tower Test Total Achievement Score. Effect sizes
of both the inhibition finding and the PS finding were large. Differences between the mean Time-per-move (DMS) and Move Accuracy ratios (DMA) for athletes and non-athletes were not
statistically significant. Mean scores on the MPS and vocabulary
tasks showed no significant intergroup variation. Table 4 shows
between-subjects variation for all athletes according to sport level
and sport type. We found no significant variation in EF scores, MPS
scores, or vocabulary scores by sport level, sport type, or the
interaction between these variables, in the absence of our control
group of non-athletes.
As shown in Table 5, hypotheses regarding intergroup differences on the inhibition and PS tasks between SP athletes, EP athletes, and non-athletes, received support. We ran planned contrast
analyses for each EF variable. All contrast values are listed above the
mean and standard deviation values of each variable and reflect our
predictions expressed in the hypotheses. With contrasts included,
mean scores on inhibition varied significantly and consistently with
our hypotheses. As predicted, SP athletes scored the highest on
Table 3
Independent samples t test analysis of athlete and non-athlete scores.
Inhibition
Problem Solving
Decision Making Accuracy
Decision Making Speed
Mental Processing Speed
Vocabulary
M
SD
M
SD
M
SD
M
SD
M
SD
M
SD
Athletes
Nonathletes
t
p
d
r
11.26
1.55
10.97
2.28
9.58
2.43
10.21
1.56
86.99
11.84
6.35
2.86
9.69
1.65
9.47
1.60
8.53
2.97
9.87
2.90
84.67
8.52
6.53
3.80
3.10
.002**
.98
.44
2.33
.012**
.76
.36
1.32
.096
.39
.19
.56
.290
.15
.07
.69
.490
.23
.11
.20
.845
.05
.03
Table 4
Between-subjects analysis of variance for sport level, sport type, and interaction
therein for athlete participants.
Independent variable
Dependent variable
F
Sig.
Partial s2
Problem Solving
<.01 .951 <.01
Decision-Making Speed
.49 .489 <.01
Decision-Making Accuracy
.50 .486
.01
Inhibition
.01 .923
.01
Mental Processing Speed
.04 .834 <.01
Vocabulary
3.51 .067
.07
Sport Type (Self-Paced or
Problem Solving
1.71 .200
.01
Externally Paced)
Decision-Making Speed
.05 .819
.04
Decision-Making Accuracy
.53 .472
.01
Inhibition
.28 .602 <.01
Mental Processing Speed
1.88 .177
.04
Vocabulary
.79 .378
.02
Sport Level*
Problem Solving
.04 .837 <.01
Sport Type (Interaction) Decision-Making Speed
<.01 .956 <.01
Decision-Making Accuracy 1.92 .175
.03
Inhibition
.11 .743 <.01
Mental Processing Speed
.11 .741 <.01
Vocabulary
2.04 .160
.04
Sport Level
(High-Skilled or
Recreational)
df1 ¼ 1; df2 ¼ 38 (Inhibition, PS, DMA, and DMS analyses are one-tailed tests.
MPS and vocabulary analyses are two-tailed tests).
inhibition, followed by EP athletes and non-athletes, respectively,
with a large effect size. On the PS task analysis with contrasts, EP
athletes scored highest, followed by SP athletes and non-athletes
respectively, again with a large effect size. We found no significant differences in DMA, DMS, vocabulary or MPS scores.
In a study like this one with many test variables, multiplicity of
tests inherently poses a threat to validity. However, we utilized the
Sidak correction for multiple testing and adjusted our alpha levels
to reduce the likelihood of false positives. Nearly all of our findings
had p-values significant at Sidak-corrected alpha levels of .013 or
.009, indicating that significant differences were present despite
the potential for error. The majority of our findings also had large
effect sizes.
Discussion
We found substantial support for many of our hypotheses. We
predicted that athletes would outscore non-athletes on all EF
measures. We expected SP athletes to outscore EP athletes and nonathletes on the inhibition task, and EP athletes to outscore SP
athletes and non-athletes on the PS, DMS, and DMA tasks. All of
these hypotheses received substantial support, except for those
involving DMS and DMA scores. Potential reasons for this are
Table 5
Planned contrast analysis of externally paced, self-paced, and non-athlete scores
(one-tailed).
Externally Self- Nont
paced
paced athletes
Inhibition
*Significant at a ¼ .05, **Significant at a ¼ .013 (Sidak correction for multiple
testing), df ¼ 52 (Inhibition, PS, DMA, and DMS analyses are one-tailed tests. MPS
and vocabulary analyses are two-tailed tests).
525
Contrast
M
SD
Problem Solving Contrast
M
SD
Decision Making Contrast
Accuracy
M
SD
Decision
Contrast
Making Speed M
SD
0
11.14
1.61
1
11.43
2.27
1
10.05
2.54
1
10.29
1.27
1
11.41
1.50
0
10.41
2.24
0
9.00
2.24
0
10.12
1.90
1
9.69
1.65
1
9.47
1.60
1
8.53
2.97
1
9.87
2.90
p
d
r
2.94 .003** .80 .37
2.77 .004** .76 .35
1.51 .045*
.41 .20
.61 .273
.17 .08
*Significant at a ¼ .05, **Significant at a ¼ .013 (Sidak correction for multiple
testing), df ¼ 52.
526
J. Jacobson, L. Matthaeus / Psychology of Sport and Exercise 15 (2014) 521e527
discussed below. We also hypothesized that high-skilled athletes
would outperform recreational athletes on all EF tests; these hypotheses did not receive support.
Our findings suggest the presence of a substantial link between
athletics and subcategories of executive functioning (EF). The first
between-groups analysis (see Table 3) suggests that athletes outperformed non-athletes in the problem solving task as well as the
inhibition task. There were not significant differences in vocabulary
scores (a proxy for overall intelligence), or mental processing speed
scores (a variable highly correlated with EF), between athletes and
non-athletes. This suggests that the groups did not differ in overall
intelligence or mental processing capacity. Although our results do
not permit us to infer a causal relationship between the variables,
we speculate that the observed differences in EF may result, at least
in part, from athletic participation. The athletes in this study
participated in sports that require the use of goal-oriented planning
and suppression of inappropriate responses (Singer, 2000). Mere
non-sport exercise has been shown to benefit EF throughout
development (Davis et al., 2011) and in adult populations
(Colcombe & Kramer, 2003). Therefore, the physical exercise in
sports, plus habitual practice of EF skills in training and competition, may lead to even further improvements. The field is lacking in
research on the effect of sport participation on EF throughout
development (Diamond & Lee, 2011); further research should
address this.
It is also possible that individuals who naturally (without
training) develop strong EF skills are more likely to become and
remain athletes. The findings of Vestberg et al. (2012), suggest that
EF competence predicts athletic success. People tend to get more
enjoyment from activities at which they excel; this may serve to
partially explain our findings. More likely still is an interaction
between these two theoretical phenomena: high-EF individuals
become athletes more often, and their EF subsequently further
improves with training, in a reinforcing cycle.
Our analysis of the relationship between sport type and EF
performance provided further support for our hypotheses. We
found that EP athletes scored higher on the PS task than SP athletes
and non-athletes. In EP sports, like soccer and tennis, athletes make
time-pressured decisions in response to external cues during goaloriented activity (Singer, 2001). In the Tower test, participants
solved each individual problem by planning and shifting in
response to the configuration of the pieces while maximizing efficiency (Delis et al., 2001a). We speculate that our findings can be
explained in part by a broad transfer from sport-related EF training,
similar to the way athletes have improved attention (Anzeneder &
Bosel, 1998). For example, EP athletes have experience with quickly
determining the positions of people and objects around them, and
subsequently using their bodies to manipulate the movements of
these people and objects. A parallel process occurs in the Tower
Test, where the participants were under time pressure to alter the
arrangement of the pieces to achieve the desired outcome.
Assuming that broad transfer of EF skills is possible, these mental
challenges appear similar enough that transfer effects could be
present in our study.
In our study, SP athletes scored highest on the inhibition task,
compared to EP athletes and non-athletes. SP sports, such as
running and swimming, allow the athlete time to plan each critical
movement and require high levels of focus and discipline (Singer,
2001). The ColoreWord Interference Test requires the suppression
of a dominant responsedreading the worddin favor of a more
preferable response: naming the ink color (Delis et al., 2001a). An
SP athlete trains and competes in activities where his primary
mental challenge is to suppress external and internal distractors to
maximize performance (Singer, 1988). For example, a cross-country
runner must suppress his internal pain and exhaustion in order to
remain focused and continue running efficiently. Although, to our
knowledge, no evidence exists of an effect of sport practice on EF,
we speculate that the mental aspects of athletic training, through
cognitive skill transfer, may improve laboratory performance on EF
tasks. Importantly, these skill-specific findings indicate that the
positive effects of purely aerobic exercise (Colcombe & Kramer,
2003), do not account for the entirety of the EF improvements
observed in athletes; some other mechanism appears to be
contributing as well.
Our hypotheses involving the EF performance of high-skilled
and recreational athletes did not receive support. Inconsistent
with findings by Vestberg et al. (2012), high-skilled athletes in our
study did not score higher on any of the EF tasks than recreational
athletes. This may be related to the possibility that in our study, the
high-skilled and recreational athletes may not have differed substantially in expertise and time devoted to their sports. We grouped
them according to varsity or non-varsity status, which is a fairly
limited indicator of sport level. There is also evidence that athletes
of different levels may not differ significantly in some aspects of
non-sport-specific cognition (Memmert, Simons, & Grimme, 2009).
Our study has a number of limitations. These limitations, listed
below along with suggestions for how they may be avoided in
further research, weaken the statistical power of our study to some
extent. However, our most noteworthy findings, regarding the
sport-type-based differences in EF scores, were significant with
99.5 percent confidence intervals and large effect sizes. First, two of
the variables, decision-making speed and decision-making accuracy, did not yield any significant results. The measures for these
variables were the Tower Time-per-move Ratio and Move Accuracy
Ratio, respectively. These two scores may be misleading, because
poor overall performance on the Tower test could actually lead to
desirable scores on either of these measures. Participants who were
wildly inaccurate with their moves sometimes had low Time-permove Ratios (indicating strong DMS), because they would
attempt move after move with little planning. The more incorrect
decisions they made, the faster they worked. On the other hand,
many participants failed to complete items, but did not attempt
many moves. This would lead to a strong Move Accuracy Ratio
because despite not completing the items, their numbers of
attempted moves were similar to or even less than the minimum
number of moves required to complete the items. Future researchers should avoid using these scores when seeking to evaluate
higher-level EF, as they are more useful in the detection of brain
damage.
Furthermore, the inexact understanding of cognitive skill
transfer (CST) causes some ambiguity. Research is inconclusive on
whether sport-related cognitive skills can transfer beyond the
sport-specific context (Pesce, 2012). Although much research supports the existence of CST (e.g., Taatgen, 2013), it remains unclear
exactly how and on what type of timeline this phenomenon
operates. This causes multiple problems. First, in our study, we only
counted as athletes those participants who currently participated
in sports at the time of testing. This meant that former varsity,
youth, or high school athletes who may have quit very recently
were considered non-athletes, when they may have actually been
more similar to the participants in the athlete group. On the other
hand, out-of-season varsity athletes, who may have not played
their sport for multiple months, were still counted as athletes.
Additionally, one might argue that the apparent differences in EF
between athletes and non-athletes may result from the likelihood
that athletes are more likely to have recently exercised. As Etnier
and Chang (2009) found, recent physical exercise improves laboratory EF performance. This effect may have been present in our
results comparing athletes and non-athletes; however, it cannot
account for our findings regarding the differences between athletes
J. Jacobson, L. Matthaeus / Psychology of Sport and Exercise 15 (2014) 521e527
of different sport types. In our study, we found no significant
relationship between time of day of testing and scores on any of the
tests. Further research is necessary to specify the nature of CST as it
applies to EF. A longitudinal study that tracks individuals' EF as they
alter their athletic and exercise habits over time could likely provide useful information to clear up this uncertainty.
Because executive functions are so highly utilized in various
activities, potential confounds exist. Not only athletes, but also
musicians, debaters, and even video gamers all utilize mental sets,
inhibit distractions, and make rapid decisions in goal-oriented activity. According to our CST-based theory, participation in activities
like these may similarly correlate with higher EF. We were unable
to control for non-athletic extracurricular activities in our study, so
some of the observed variation may have resulted from non-sports
related transfer effects. Further research may focus on differences
in EF between those who participate in these other activities and
those who do not (perhaps controlling for athletic participation).
Our study is somewhat limited in depth and scope. Although our
sample size (N ¼ 54) was large enough to yield significant results
and large effect sizes, a study with a larger N could have stronger
statistical power. Furthermore, our sample was not representative
of the population at large. We sampled from a small liberal arts
college, which contains mostly affluent, Caucasian, highfunctioning young adults. In order to generalize these findings
beyond this type of population, we would need to collect data from
a less restricted sample. We suggest further research on different
populations with a larger sample. Finally, because our study was
quasi-experimental, we were unable to determine causality and
could only speculate about the reasons for our results. We suggest a
future pretesteposttest design study in which participants are
randomly assigned to SP athlete, EP athlete, and non-athlete
groups. If EF subcategories improve through athletic training and
competition in a manner consistent with our findings, we would
have strong evidence that certain types of athletics indeed differentially improve EF.
References
Anzeneder, C. P., & Bosel, R. (1998). Modulation of the spatial extent of the attentional focus in high-level volleyball players. European Journal of Cognitive Psychology, 10(3), 247e267.
Baudouin, A., Clarys, D., Vanneste, S., & Isingrini, M. (2009). Executive functioning
and processing speed in age-related differences in memory: contribution of a
coding task. Brain and Cognition, 71(3), 240e245.
Brand, M., Fujiwara, E., Borsutzky, S., Kalbe, E., Kessler, J., & Markowitsch, H. J.
(2005). Decision-making deficits of Korsakoff patients in a new gambling task
with explicit rules: associations with executive functions. Neuropsychology,
19(3), 267e277.
Chase, W. G., & Simon, H. A. (1973). Perception in chess. Cognitive Psychology, 4,
55e61.
Colcombe, S., & Kramer, A. F. (2003). Fitness effects on the cognitive function of
older adults: a meta-analytic study. Psychological Science, 14(2), 125e130.
Davis, A. S., Pierson, E. E., & Finch, W. (2011). A canonical correlation analysis of
intelligence and executive functioning. Applied Neuropsychology, 18(1), 61e68.
Davis, C. L., Tomporowski, P. D., McDowell, J. E., Austin, B. P., Miller, P. H.,
Yanasak, N. E., et al. (2011). Exercise improves executive function and
achievement and alters brain activation in overweight children: a randomized,
controlled trial. Health Psychology, 30(1), 91e98.
Delis, D. C., Kaplan, E., & Kramer, J. H. (2001a). DeliseKaplan executive function
system: Examiner's manual. Texas: The Psychological Corporation.
Delis, D. C., Kaplan, E., & Kramer, J. H. (2001b). DeliseKaplan executive function
system: Technical manual. Texas: The Psychological Corporation.
Di Russo, F., Bultrini, A., Brunelli, S., Delussu, A., Polidori, L., Taddei, F., et al. (2010).
Benefits of sports participation for executive function in disabled athletes.
Journal of Neurotrauma, 27(12), 2309e2319.
Diamond, A. (2006). The early development of executive functions. In E. Bialystok, &
F. M. Craik (Eds.), Lifespan cognition: Mechanisms of change (pp. 70e95). New
York, NY: Oxford University Press.
Diamond, A. (2013). Executive functions. Annual Review of Psychology, 64, 135e168.
Diamond, A., & Lee, K. (2011). Interventions shown to aid executive function
development in children 4e12 years old. Science, 333, 959e964.
Dickinson, D., & Gold, J. M. (2008). Processing speed and the Digit Symbol Substitution Test in schizophrenia. In J. DeLuca, & J. H. Kalmar (Eds.), Information
527
processing speed in clinical populations (pp. 125e151). Philadelphia: Taylor &
Francis.
Etnier, J. L., & Chang, Y. (2009). The effect of physical activity on executive function:
a brief commentary on definitions, measurement issues, and the current state
of the literature. Journal of Sport & Exercise Psychology, 31(4), 469e483.
Furley, P., & Memmert, D. (2011). Studying cognitive adaptations in the field of
sport: broad or narrow transfer? Perceptual and Motor Skills, 113(2), 481e488.
Green, C. S., Pouget, A., & Bavelier, D. (2010). Improved probabilistic inference as a
general learning mechanism with action video games. Current Biology, 20,
1573e1579.
Hannah, S. T., Balthazard, P. A., Waldman, D. A., Jennings, P. L., & Thatcher, R. W.
(2013). The psychological and neurological bases of leader self-complexity and
effects on adaptive decision-making. Journal of Applied Psychology, 98(3),
393e411.
Harris, M. E. (2001). A cognitive-behavioral approach to teaching social problem
solving to elementary-age children with mental disabilities. Tampa, FL: University
of South Florida.
Keating, X. D., Castelli, D., & Ayers, S. (2013). Association of weekly strength exercise
frequency and academic performance among students at a large university in
the United States. Journal of Strength & Conditioning Research, 27(7), 1988e1993.
Kesler, S., Hosseini, H., Heckler, C., Janelsins, M., Palesh, O., Mustian, K., et al. (2013).
Cognitive training for improving executive function in chemotherapy-treated
breast cancer survivors. Clinical Breast Cancer, 13(4), 299e306.
Kotsopoulos, D., & Lee, J. (2012). A naturalistic study of executive function and
mathematical problem-solving. The Journal of Mathematical Behavior, 31(2),
196e208.
Lum, J., Enns, J. T., & Pratt, J. (2002). Visual orienting in college athletes: explorations
of athlete type and gender. Research Quarterly in Exercise and Sport, 73(2),
156e167.
Mann, D. T., Williams, A. M., Ward, P., & Janelle, C. M. (2007). Perceptual-cognitive
expertise in sport: a meta-analysis. Journal of Sport and Exercise Psychology,
29(4), 457e478.
Memmert, D., Simons, D. J., & Grimme, T. (2009). The relationship between visual
attention and expertise in sports. Psychology of Sport and Exercise, 10(1),
146e151.
Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., & Howerter, A. (2000). The
unity and diversity of executive functions and their contributions to complex
“frontal lobe” tasks: a latent variable analysis. Cognitive Psychology, 41(1),
49e100.
Morris, W. (Ed.). (1976). The American heritage dictionary. Boston, MA: Houghton
Mifflin Company.
Pesce, C. (2012). Shifting the focus from quatitative to qualitative exercise characteristics in exercise and cognition research. Journal of Sport & Exercise Psychology, 34(6), 766e786.
Pesce, C., Cereatti, L., Casella, R., Baldari, C., & Capranica, L. (2007). Preservation of
visual attention in older expert orienteers at rest and under physical effort.
Journal of Sport and Exercise Psychology, 29(1), 78e99.
Singer, R. N. (1988). Strategies and metastrategies in learning and performing selfpaced athletic skills. The Sport Psychologist, 2(1), 49e68.
Singer, R. N. (2000). Performance and human factors: considerations about cognition and attention for self-paced and externally-paced events. Ergonomics,
43(10), 1661e1680.
Singer, R. N., Cauraugh, J. H., Chen, D., Steinberg, G. M., & Frehlich, S. G. (1996).
Visual search, anticipation, and reactive comparisons between highly-skilled
and beginning tennis players. Journal of Applied Sport Psychology, 8(1), 9e26.
Smith, B. L., Smith, T. D., Taylor, L., & Hobby, M. (2005). Relationship between intelligence and vocabulary. Perceptual and Motor Skills, 100(1), 101e108.
Spreen, O., & Strauss, E. (1998). A compendium of neuropsychological tests: Administration, norms, and commentary (2nd ed.). New York, NY: Oxford University
Press.
Taatgen, N. A. (2013). The nature and transfer of cognitive skills. Psychological Review, 120(3), 439e471.
Tsai, C., Wang, C., & Tseng, Y. (2012). Effects of exercise intervention on event-related
potential and task performance indices of attention networks in children with
developmental coordination disorder. Brain and Cognition, 79(1), 12e22.
Unsworth, N., Miller, J. D., Lakey, C. E., Young, D. L., Meeks, J., Campbell, W., et al.
(2009). Exploring the relations among executive functions, fluid intelligence,
and personality. Journal of Individual Differences, 30(4), 194e200.
VandenBos, G. R. (2006). APA dictionary of psychology. Washington, D.C.: American
Psychological Association.
Vestberg, T., Gustafson, R., Maurex, L., Ingvar, M., & Petrovic, P. (2012). Executive
functions predict the success of top-soccer players. PLoS ONE, 7(4).
Voss, M. W., Kramer, A. F., Basak, C., Prakash, R., & Roberts, B. (2010). Are expert
athletes experts in the cognitive laboratory? A meta-analytic review of cognition and sport expertise. Applied Cognitive Psychology, 24(6), 812e826.
Wechsler, D. (1997). Wechsler adult intelligence scale (3rd ed.). San Antonio, TX: The
Psychological Corporation.
Wechsler, D. (2008). Wechsler adult intelligence scale (4th ed.). San Antonio, TX:
Pearson.
Wen, M., Butler, L. T., & Koutstaal, W. (2012). Improving insight and non-insight
problem solving with brief interventions. British Journal of Psychology, 104(1),
97e118.
Zoudji, B., Thon, B., & Debû, B. (2010). Efficiency of the mnemonic system of expert
soccer players under overload of the working memory in a simulated decisionmaking task. Psychology of Sport and Exercise, 11(1), 18e26.