Professional Documents
Culture Documents
To cite this article: Dmitri Poltavski & David Biberdorf (2014): The role of visual perception measures used in sports vision
programmes in predicting actual game performance in Division I collegiate hockey players, Journal of Sports Sciences, DOI:
10.1080/02640414.2014.951952
Taylor & Francis makes every effort to ensure the accuracy of all the information (the Content) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
Journal of Sports Sciences, 2014
http://dx.doi.org/10.1080/02640414.2014.951952
Abstract
In the growing eld of sports vision little is still known about unique attributes of visual processing in ice hockey and what
role visual processing plays in the overall athletes performance. In the present study we evaluated whether visual, perceptual
and cognitive/motor variables collected using the Nike SPARQ Sensory Training Station have signicant relevance to the
real game statistics of 38 Division I collegiate male and female hockey players. The results demonstrated that 69% of
variance in the goals made by forwards in 20112013 could be predicted by their faster reaction time to a visual stimulus,
better visual memory, better visual discrimination and a faster ability to shift focus between near and far objects.
Approximately 33% of variance in game points was signicantly related to better discrimination among competing visual
stimuli. In addition, reaction time to a visual stimulus as well as stereoptic quickness signicantly accounted for 24% of
variance in the mean duration of the players penalty time. This is one of the rst studies to show that some of the visual
skills that state-of-the-art generalised sports vision programmes are purported to target may indeed be important for hockey
players actual performance on the ice.
Keywords: sports vision, ice hockey, the Nike SPARQ Sensory Training Station, sport performance, visual perception
The role of visual perception measures used in sports On the other hand Bulson, Ciuffreda, and Hung
vision programmes in predicting actual game perfor- (2008) did not nd a signicant effect of degraded
mance in Division I collegiate hockey players. static acuity on athletic performance, while Ward
There has been considerable debate in the literature and Williams (2003) failed to report signicant
about whether generalised visual processing is better differences in performance on a dynamic visual
developed in athletes compared to non-athletes, and acuity test between elite and subelite youth soccer
in expert athletes compared to novices. Some of the players. Similarly, Milne and Lewis (1993) did not
positive ndings encompass the range of sensory, nd any differences between athletes and non-ath-
motor and perceptual aspects of basic vision and letes in either speed or span of recognition by
information processing. These include measures of evaluating the ability to recall a sequence of num-
visual resolution (dynamic visual acuity (Millslagle, bers presented tachistoscopically for 1/50 of a sec-
2000), static visual acuity (Coffey & Reichow, 1989) ond. Additionally, Classe et al. (1997) failed to
and contrast sensitivity (Kluka et al., 1995)), depth nd any differences in visual reaction time
perception (stereopsis; Laby et al., 1996), visual track- between elite and novice baseball players.
ing (vergence, pursuit, saccades and xation), visuo- Mixed ndings have also been reported in evalua-
motor integration (eyehand coordination, visual tion of results of specialised training programmes
reaction time; Hughes, Bhundell, & Waken, 1993) intended to enhance basic visual perceptual pro-
and visual information processing: visual eld (Berg cesses (e.g. visual acuity, combined saccadic/
& Killian, 1995), speed discrimination and temporal accommodative tracking and visual search) in ath-
processing (Overney, Blanke, Herzog, & Burr, 2008), letes. For example, Junyent and Sole (1995)
peripheral awareness (Zwierko, 2008) and speed of reported that specialised training of the above
recognition (Isaacs & Finch, 1983). basic visual skills improved precision shooting
Correspondence: Dmitri Poltavski, Department of Psychology, University of North Dakota, 319 Harvard St. Stop #8380, Grand Forks, ND 58202, USA.
E-mail: dpoltavski@me.com
scores of elite shooters compared to baseline. In a skilled counterparts and anticipated their opponents
basketball study, Kofsky and Stareld (1989) intentions signicantly quicker than less-skilled par-
reported improvements in both visual function and ticipants suggesting that the use of advanced percep-
actual game performance following 5 weeks of gen- tual cues facilitates sport performance by means of
eral vision training (static and dynamic visual aiding in the anticipation of opponents actions and
acuity, visual reaction time, peripheral awareness, decreasing overall response time. In another meta-
eyehand coordination and visualisation/visual analysis of 20 studies Voss, Kramer, Basak, Prakash,
imagery). and Roberts (2010) also found a small-to-moderate
McLeod (1991) tested the effects of a 12-session effect size for the difference between experts versus
visual skill training programme Eyerobics in nine non-experts in multiple sports on basic cognitive
female varsity soccer players. The programme incor- measures of visual attention and processing speed.
porated basic versional, vergence, accommodative As a result state-of-the-art sports vision programmes
and visualisation skills. Compared to the control now employ integrated visual assessment systems such
group (n = 9), athletes receiving generalised visual as the Nike SPARQ Sensory Training Station,
training showed signicantly better performance fol- designed to test a broad range of basic visual and
lowing the programme on a general test of eyehand information processing skills that previously have
Downloaded by [University of North Dakota] at 07:41 25 August 2014
coordination, balance, as well as on a sport-specic been identied as important for sports. As discussed
dribble test, that involved dribbling a soccer ball above these skills include static visual acuity, dynamic
around four cones, 9 feet apart, in a gure-eight visual acuity, contrast sensitivity, distance stereopsis,
fashion. accommodativevergence facility, central eyehand
Negative ndings associated with generalised vision reaction and response speeds, peripheral eyehand
training programmes were primarily reported by the proaction (speed and precision of self-generated target
Abernethy group (Abernethy, 1986; Abernethy, changes), span of perception and stimulus discrimina-
Wann, & Parks, 1998; Abernethy & Wood, 2001; tion (a form of eyehand recognition reaction time and
Wood & Abernethy, 1997), who evaluated the effects precision).
of visual training in sport-specic performance in The system was developed to provide a custo-
tennis players. The researchers found no programme mised sensory performance prole that graphically
benets on either sport-specic perceptual tests (i.e. represents the athletes visual strengths and weak-
coincidence timing, rapid ball detection and anticipa- nesses by comparing performance (using percentile
tion) or sport-specic motor tasks (i.e. tennis fore- scores) to a database of peers within a given sport
hand drive accuracy) in the experimental group (total of 24 sports), and has been recently found a
receiving vision training compared either to the pla- reliable computer-based assessment system showing
cebo group (read tennis instructional manuals and no learning effect over multiple testing sessions
watched instructional tennis videos) or to the control (testretest reliability, Erickson et al., 2011).
group (practiced weekly motor tasks). While relative contributions (weights) of various
Erickson et al. (2011) explained the negative nd- aspects of generalised visual processing to overall
ings in basic visual skills and vision training pro- performance in a sport have been suggested for
grammes of athletes on the basis of a lack of golf, football, baseball, basketball, tennis and soccer
standardisation of many assessment techniques, (see Ciuffreda & Wang, 2004; Gardner & Sherman,
inconsistent ambient testing conditions, outdated 1995; Seiderman & Schneider, 1983), little is known
instrumentation and inappropriate assessment pro- about unique attributes of visual processing in ice
tocols (e.g. assessment of static stereopsis vs. hockey and what role visual processing plays in the
dynamic stereopsis; use of numeric stimuli in assess- overall athletes performance. Such understanding
ment of athletes perception span). In their meta- could help develop better sport-specic visual train-
analysis Ciuffreda and Wang (2004) also under- ing programmes that would emphasise specic visual
scored the importance of higher information proces- skills and elements of information processing that
sing skills (i.e. as prediction/anticipation, recall, could benet ice hockey players to a greater extent
cognitive strategy and decision-making) in athletic than the currently available one-size-ts-all visual
performance and the ability of current sport vision training protocols.
programmes to specically target these skills. In the present study we evaluated whether visual,
Studies examining decision-making ability in rela- perceptual and cognitive/motor variables collected
tion to athletic performance within specic sports using the Nike SPARQ Sensory Training Station
also provide evidence of better decision-making in have a signicant association with actual athletic
experts compared to novices. In their meta-analysis performance of elite ice hockey players. We hypothe-
of 42 studies Mann, Williams, Ward, and Janelle sised that a number of dynamic components of
(2007) concluded that experts were more accurate visual perception and visuomotor control would be
in their decision-making relative to their lesser more important for prediction of game statistics than
Visual perception in hockey 3
varies sinusoidally in brightness from the centre to Dynamic visual acuity (target capture). Dynamic
the edge. Participants are instructed to swipe the visual acuity (DVA) generally is dened as the ability
screen of the iPod touch in the direction of the circle of the visual system to resolve detail when there is
with the pattern. Contrast sensitivity is measured relative movement between the target and the obser-
binocularly at two spatial frequencies, 6 and 18 ver (Erickson et al., 2011). The method of DVA
cycles per degree (cpd), using a staircase reversal assessment used in the Nike SST, however, does
algorithm. Final threshold contrast sensitivity is not conform to the traditional method involving a
measured between 10% and 1.0% (1.0 to 2.0 log moving target. Its construct validity is yet to be
units) contrast at 6 cpd and between 32% and 2.5% determined. The developers of this test refer to it
(0.5 to 1.6 log units) contrast at 18 cpd. as target capture. On this test the participant is
instructed to xate a central white dot until a yellow-
Depth perception (stereopsis at far). For the measure of green Landolt ring (dominant wavelength about
distance depth perception the participant wears a 555 nm at maximum saturation possible on the dis-
pair of liquid crystal goggles (NVIDIA 3D Vision, play) appears briey in one of the four corners of the
Santa Clara, CA), connected via wireless link to the screen. The size of the Landolt ring is automatically
computer, and faces the Nike Sensory Station 22- set by the computer at 0.1 log unit above the thresh-
Downloaded by [University of North Dakota] at 07:41 25 August 2014
inch display. The liquid crystal shutter system cre- old determined with the Static Visual Acuity assess-
ates simulated depth in one of four black rings pre- ment, and the angular distance along the diagonal
sented on a white background, such that one ring from the xation dot to the centre of the Landolt
appears to oat in front of the screen. The size and ring is 6.1. Since there is a reduction in visual acuity
arrangement of the rings are identical to those of the the farther the stimulus is away from the fovea, indi-
circles used in Contrast Sensitivity. The width of the viduals with visual acuity of 20/50 or better would
lines dening each ring is 12 mm, subtending 0.14. need to saccade from the xation dot to the Landolt
Participants are instructed to swipe the screen of the ring to correctly discriminate the direction of the
iPod touch in the direction of the oating ring and gap. This quick saccade coupled with a need to
are encouraged to respond as quickly as possible. quickly (within milliseconds) identify the target (the
Threshold stereopsis is measured between 237 and direction of the gap) is thought to assess DVA
12 arc seconds using a staircase reversal algorithm (Erickson et al., 2011).
similar to that described previously. In addition, The participant is instructed to move the eyes
response time for the rst two stimulus presentations from the centre xation dot to the Landolt ring
at the participants threshold is recorded, and an that would briey appear at one of the four random
average response time for the testing is automatically corners of the screen and to try to correctly discri-
calculated by the software. Next, the participant is minate the direction of the gap by swiping the screen
instructed to turn 90 to the right and turn the head of the iPod touch. The duration of the Landolt ring
to left in order to view the screen in a way that would presentation starts at 500 ms and is progressively
test stereopsis while viewing over the left shoulder. shortened after a correct response. The threshold
The procedure is then repeated. Following this, the stimulus exposure duration is determined using a
participant is instructed to turn 180 and turn the staircase reversal algorithm.
head to the right in order to view the screen in a way The four remaining Nike SST tests are performed
that would test stereopsis while viewing over the within arms reach of the instrument and utilises
right shoulder. The procedure is again repeated. the high denition 42-inch touch screen monitor.
These tests include Perception Span, EyeHand
Nearfar quickness. The participant is instructed to Coordination, Go/No-Go and Hand Reaction Time.
hold the iPod touch at 16 inches (40 cm) from the
eyes, with the top edge positioned just below the Perception span. The standing participant is positioned
bottom of the far screen. In alternating style, a 20/ within arms length of the Nike Sensory Stations 42-
80 equivalent black Landolt ring is presented in a inch touch-sensitive display, with the centre of the
box on the hand-held screen, and a black Landolt screen at about eye level. Automated instructions
ring 0.1 log unit above the threshold determined direct the participant to focus on a shrinking white
with the Visual Clarity assessment is presented on dot in the centre of a grid pattern composed of up to
the far screen. The participant is instructed to swipe 30 circles. When the dot disappears, a pattern of
the screen of the iPod touch in the perceived direc- yellow-green dots (same colour parameters as above)
tion of the gap in the ring presented on each display. ashes simultaneously for 100 ms within the grid.
Each participant continually switches focus between The participant then touches the screen to recreate
the far and near screens for 30 s, trying to correctly the pattern of dots. If the participant achieves a pas-
identify as many rings as possible. The number of sing score (greater than or equal to 75% correct), the
correct responses determines the score. grid pattern increases in size with an increasing
Visual perception in hockey 5
number of dots. The rst two levels have six circles in two concentric circles, 11.4 cm and 3.2 cm in dia-
the grid pattern with 2 and 3 dots, the next ve levels meter. Automated instructions direct the participant
have 18 circles with 3 to 7 dots, and the last four to place the ngertips of the dominant hand on the
levels has 30 circles with 7 to 10 dots. Each circle is inner circle of the annulus on that side of the screen,
19 mm in diameter, and the largest grid pattern is with no portion of the hand extending across the
18 cm in diameter. The grids and dot patterns are boundary line marked on the screen. If the hand is
preset by the computer to maintain standardisation. aligned correctly, this control annulus changes col-
The overall score for this assessment is based on the our to yellow-green (same colour parameters as
cumulative number of correct responses; missed above). The participant is instructed to centre the
responses and extra guesses are subtracted from the body in front of the opposite test annulus and focus
cumulative score. If the participant does not achieve a attention on the centre of that annulus. After a ran-
passing score on a level, that level is repeated until domised delay of 2, 3 or 4 s, the test annulus turns
two consecutive failures after which the assessment is yellow-green, and the participant moves the hand to
terminated. The maximum score possible on this touch its inner circle as quickly as possible. Five
assessment is 64. trials are conducted per participant to calculate aver-
age reaction and response times. Reaction time is
Downloaded by [University of North Dakota] at 07:41 25 August 2014
Eyehand coordination (peripheral eyehand response). measured as the elapsed time between onset of the
Participants stand in front of the 42-inch touch-sen- test annulus and release of the control annulus. After
sitive display screen holding their arms at shoulder ve trials, the computer calculates the averages and
height within easy reach of a grid of circles presented standard deviations for the reaction and response
on the display. The grid consists of 8 columns times. If any single measure differs from the mean
(68.6 cm) and 6 rows (44.5 cm) of equally spaced by more than two standard deviations in either direc-
circles, with each circle 48 mm in diameter. During tion, another trial is conducted to replace the out-
the assessment, a yellow-green dot (same colour lying measure for that trial. The software is
parameters of above) appears within one circle of programmed such that no more than two trials are
the grid. Automated instructions direct the partici- repeated for any participant.
pant to touch the dot as quickly as possible using
either hand. As soon as the dot is touched, a subse-
Procedure
quent dot is presented. The score recorded is the
total time to touch all 96 presented dots. All testing was carried out over the summer after the
completion of the 20112012 season and before the
Go/No-Go. The position of the participant and the beginning of the 20122013 season. Since several of
grid pattern remains the same as that used for Eye the players tested were new recruits and did not have
Hand Coordination but the dot stimulus presented any 20112012 statistics, while some other players
is either yellow-green (same parameters as above) or missed a number of games in the 20112012 season
red (dominant wavelength about 620 nm at maxi- due to injury, we aggregated data over two regular
mum saturation possible on the display). If the dot is seasons to achieve greater accuracy of dependent
yellow-green, the participant is instructed to touch it measures and greater statistical power for the sam-
as before. But if the dot is red, the participant is ple. Following the testing (during the 20122013
instructed not to touch it. Both the red and yellow- season) none of the evaluated players underwent
green dots appear at random locations for only any visual therapy (including sports vision), per-
450 ms, with no time gap between dot presentations. formed any supplementary training (beyond conven-
If a yellow-green is not touched within this time, no tional protocols) or altered their visual correction.
point is awarded for that presentation; if a red dot is Upon arrival at the testing location (a local opto-
touched, a point is subtracted from the overall score. metric clinic) informed consent was obtained from
Again, participants are encouraged to touch as many each participant followed by administration of a Z-
yellow-green dots as possible. Ninety-six total dots View Aberrometer & Autorefractor (Ophthonix,
(64 yellow-green, 32 red) are presented, and the Vista, CA) over the participants habitual playing
overall score is calculated as the cumulative number refraction to determine what, if any, refractive error
of yellow-green dots touched minus any red dots or residual error there might be for each eye under
touched. non-cyclopleged conditions. If contact lenses were
worn, the test was repeated without contact lenses
Hand reaction time (central eyehand reaction and and the lenses were replaced on the participants
response time). For this test participants remain stand- eyes after the test was completed. The refractive
ing at arms length from the 42-inch touch-sensitive outcome (uncorrected refraction or contact lens
display. Two annular patterns appear on the screen over-refraction) was then recorded for each eye
with centres 30.5 cm apart; each annulus consists of along with the Aberration Index.
6 D. Poltavski & D. Biberdorf
The athletes then completed the Nike SST assess- Statistical analyses. Collected measures of visual pro-
ment, which took approximately 30 min. cessing were used to predict individual players per-
formance statistics using a series of linear multiple
regression analyses. To determine which variables
Predictors. Eleven Nike SST variables were used to pre- should be included into the prediction equation for
dict performance measures. They were obtained from each dependent measure, we rst obtained indivi-
the nine Nike SST tests described above. Each test was dual Pearsons r correlation coefcients between
associated with one corresponding variable except depth SST variables and dependent measures. Variables
perception (the test generated two variables: depth per- with signicant bivariate correlations were then
ception threshold (arc seconds), depth perception response included into the regression models to determine
time (ms)) and hand reaction time (comprised: average the relative contributions of specic visual/percep-
reaction time and average response time (ms)). The tual/motor skills in explaining variance on perfor-
average reaction and response times were signicantly mance measures.
correlated (r = 0.43, p < 0.01). To avoid the issue of We determined the maximum allowable number
multiple collinearity, we created a calculated variable of predictors in our model based on the number of
average motor time, which was derived by subtracting available data points for each dependent measure.
the average reaction time from the average response
Downloaded by [University of North Dakota] at 07:41 25 August 2014
Table I. Means and standard deviations for dependent variables as a function of player position.
Per cent goals 0.10 (0.05) 0.080.12 0.06 (0.03) 0.040.08 3.17** 1.0
Average number of points per game 0.53 (0.56) 0.300.80 0.27 (0.13) 0.190.35 2.16* 0.75
Average number of penalty minutes per game 0.64 (0.51) 0.450.89 0.70 (0.61) 0.341.07 0.31 0.11
a possible multicollinearity problem, and a CI > 30 Figure 1 for details). These variables were then
suggests a serious multiple collinearity problem. entered into a regression model using a stepwise
Finally, using G-Power 3.1 (Faul, Erdfelder, backward elimination procedure.
Buchner, & Lang, 2009) we performed post hoc ana- The results showed that the nal model contained
lyses of achieved statistical power of each model based all four of the entered variables. The linear model was
on the observed effect sizes (Cohens f2) and = 0.05 to able to explain the variability in the criterion signi-
determine the sensitivity of the model to Type II error cantly better than chance (F(4,16) = 8.93, p < 0.01)
(failure to reject the null hypothesis when it is false). and accounted for 69% of the total variance in the
goal percentage. Regression coefcients for all of the
variables were statistically signicant at = 0.05.
Results These results are presented in Table III. The post
Per cent goals hoc test of the achieved power of the model was
determined on the basis of the obtained effect size
Four Nike SST variables had signicant bivariate (f2). With explained variance of 0.69 and residual
correlations with goal percentage (see Table II and
Table II. Signicant bivariate correlations (Pearson r) of Nike SST measures with measures of athletic performance in Division I hockey
Downloaded by [University of North Dakota] at 07:41 25 August 2014
players.
Pearson r
Game point average Per cent goals Average penalty minutes per
NST measure Mean (s) 95% CI (n = 24) (n = 24) game (n = 38)
.15 .15
Per cent goals
.10 .10
.05 .05
.00 .00
0 10 20 30 40 50 0 10 20 30 40 50 60
Go/No-GO score Preceptual span
.15 .15
Per cent goals
.10 .10
r = 0.45
.05 .05
.00 .00
Figure 1. Bivariate correlations between total goal percentage and variables entered into the multiple regression model: the total score on
the Go/No-Go trials, the nearfar quickness score, perception span and the average reaction time to a visual stimulus.
8 D. Poltavski & D. Biberdorf
Table III. Regression analyses summary for individual Nike SST variables predicting performance measures in Division I hockey players.
Notes: Adjusted R2 = 0.61 (n = 22, p = 0.01); f 2 = 2.23. B = Unstandardised coefcient B; SEB = Standard Error of B; = Standard Error of .
*p < 0.05.
Mean points per game Go/No-Go total score 0.026 0.009 0.29**
Average number of penalty minutes per game Depth perception; mean reaction time (ms) 0.000 0.000 0.323*
Average reaction time over ve trials (ms) 0.004 0.002 0.351*
0 10 20 30 40 50
Go/No-GO score
Average number of points per game
Figure 2. Bivariate correlations between the average number of
Only one variable the total score on the Go/No-Go game points and variables entered into the regression model: the
trials showed a signicant bivariate correlation with total score on the Go/No-Go trials.
the criterion (average number of points per game,
see Table II and Figure 2). When entered into a
regression model it accounted for 33% of the var- Average number of penalty minutes per game
iance in the average number of game points Depth perception mean response time (dynamic
(R2 = 0.33), which was statistically signicant (F stereopsis) and average reaction time were the
(1,20) = 9.63, p < 0.01) . The obtained power of the only two variables that had signicant correlations
model based on f2 of 0.49 (large according to with the average number of penalty minutes per
Cohens (1988) guidelines) was 0.90, suggesting game (see Table II and Figure 3). These variables
that the model was well powered to detect the true were then entered into a backward, stepwise regres-
effect of the predictor variable. sion model.
The results, thus, suggested that a greater number The results showed that both variables were
of game points to a signicant extent are related to retained in the nal model and each variable by itself
better discrimination among competing visual sti- was signicant at = 0.05 in explaining variability in
muli (greater number of hits on the Go/No-Go the criterion. The model with two predictors
test) and the ability to inhibit non-target responses accounted for the total of 24% of variance in the
(false alarms on the Go/No-Go test). average number of penalty minutes per game,
Visual perception in hockey 9
Average number of penalty minutes per game 1.00 our original hypothesis, static visual processing and
r = 0.37 nearpoint visual skills (i.e. static visual acuity (SVA),
contrast sensitivity and stereopsis) were not signi-
.80
cantly related to indexes of on-the-ice performance.
At the same time more dynamic components of visual
.60 perception and visuomotor control such as stimulus
discrimination, nearfar quickness and dynamic
.40 stereopsis predicted signicant amounts of variance
in our dependent measures. Somewhat contrary to
the original hypothesis dynamic visual acuity (DVA)
.20
and eyehand coordination were not signicantly
associated with either goal scoring or average number
.00 of game points.
280.00 300.00 320.00 340.00 360.00 380.00 The lack of a signicant relationship between
Average reaction time (ms) offense statistics and static measures of visual per-
ception observed in the present study with ice
Average number of penalty minutes per game
1.00
Downloaded by [University of North Dakota] at 07:41 25 August 2014
eye movements in response to a fast-moving target hockey ensures that preparatory and attentional fac-
seems to be one of the key factors inuencing ones tors assist in the production of critical movement
DVA (Kohmura, Aoki, Honda, Yoshigi, & decisions, which in turn affects performance.
Sakuraba, 2008). Since the Nike SST test of target Reaction time and response time (movement
capture, purported to assess DVA, did not involve time) are considered to be the classic measurements
either a moving target or head turning, it may have of the efciency and effectiveness of an individual'
measured something other than dynamic visual capacity to perform sport skills (Magill, 2006).
acuity. Together, reaction time plus movement time is
An alternative explanation (provided that target equal to visual motor response time (VMRT).
capture does represent DVA) is that DVA VMRT has been identied as a key performance
decreases with increased target velocities and indicator of prociency in many ball sports
decreased target size (Hoffman, Rouse, & Ryan, (Erickson, 2007). Ciuffreda (2011) reports that
1981), producing an increased physiological demand VMRT in the retinal periphery can be reduced with
on the observer (e.g. resolving power of the eye, training by up to 20 ms, which could confer to the
oculomotor abilities, peripheral awareness and psy- athlete potentially signicant benets on the eld.
chological abilities to interpret what is seen). It is Similar improvements around 20 ms in choice reac-
Downloaded by [University of North Dakota] at 07:41 25 August 2014
thus possible that at certain puck speeds DVA ceases tion time (a measure of the time from the arrival of a
to be a reliable visual cue facilitating decision-mak- suddenly presented stimulus (visual reaction time)
ing and offensive play. until the beginning of the actual action (motor
Similarly, the measure of eyehand coordination response)) have been recently reported by Schwab
in the present study involved rapid touching of lit-up and Memmert (2012) in youth male eld hockey
dots on the touch-sensitive display using either hand players following a 6-week generalised visual training
depending on the target location. This procedure programme (DynamicEye SportsVision Training
may have not adequately simulated eyehand coor- Program). The authors, however, did not report
dination tasks involved in ice hockey where (except any improvement following training on a multiple
for goaltenders) both hands typically work in syn- object-tracking task thought to be important for the
chronicity (wielding a hockey stick) when interacting sport.
with the target (i.e. the puck). There are many situations in sport that require the
On the other hand other general visuomotor vari- athletes to make a specic and appropriate motor
ables in combination with perceptual, attentional response to a certain visual stimuli (stimulus discri-
and cognitive parameters showed a much stronger mination). Therefore, both the speed and the accu-
relationship to goal scoring and average number of racy of linking visual to neuromuscular processing
game points. Specically, faster simple motor reac- were associated by Erickson (2007) as evidence of
tion time in combination with a greater perception the integrity of the visual motor control system. In
span, better visual discrimination (decision-making) the present study the ability to quickly recognise and
and a faster ability to shift focus between near and far accurately respond to visual stimuli (total score on
objects (dynamic visual attention) predicted 69% of the Go/No-Go measure) was an important variable
the variability in the goals made by forwards in in predicting individual goal percentages and the
20112013. Additionally, visual attention and mean number of game points.
motor control as measured by the total score on the A signicant relationship between the ability to
Go/No-Go trials were shown to predict the average quickly shift attentional focus from near to far
number of game points. Regression analyses of indi- objects and the athletic performance has previously
vidual Nike SST variables demonstrated that about been demonstrated in volleyball players (Di Russo
33% of variability in accumulated game points could et al., 2003). Ciuffreda and Wang (2004) further
be explained by better decision-making as measured contended that specialised visual attentional training
by the ability to rapidly discriminate among compet- emphasising dynamic shifting of ones visual atten-
ing visual stimuli and to inhibit non-target motor tional focus can signicantly contribute to the overall
responses. improvement of the athletes performance and
The above ndings are consistent with previous should, thus, be incorporated into any sports vision
studies that suggested that the ability to quickly read training programme. Our ndings are, indeed, in
offensive and defensive play patterns is critical in ice line with this recommendation as the nearfar quick-
hockey (Martell & Vickers, 2004). The ability to ness scores were also predictive of the goal
quickly read and react in ice hockey has been dened percentage.
as the perceptual ability to selectively attend to key The results further suggest that a decrease in sim-
components of the game and rapidly execute the ple reaction time alone is not enough to facilitate
correct decision. According to Martell and Vickers performance on the ice: improvements in simple
(2004), this perception action relationship in ice reaction times need to be accompanied by
Visual perception in hockey 11
corresponding improvements in higher level infor- steps in the direction of designing sport-specic
mation processing/decision-making. Otherwise visual training programmes that may supplement
rapid motor responses in some cases may result in currently existing conventional training protocols,
faster non-target responses, which may lead to a which may result in additional benets to players
greater number of penalty minutes. In the present and improved performance in the eld. A corollary
study faster simple reaction time to a visual stimulus of the present study may be a study of performance
as well as faster stereopsis of a 3-D target in 24% of improvements in hockey players following a sport
the cases predicted greater mean duration of the vision therapy that specically emphasises the
players penalty time. Certain personality character- aspects of visual perception and visuomotor control
istics of athletes may help explain these ndings. described above.
Although in the present study we did not measure
impulsivity, Edman, Schalling, and Levander (1983)
found signicantly shorter simple reaction times on a Conclusion
choice reaction time task in more impulsive partici- This is one of the rst studies to show that some of
pants, who also made signicantly more errors than the visual skills that state-of-the-art generalised
less impulsive participants. Logan, Schachar, and sports vision programmes are purported to target
Downloaded by [University of North Dakota] at 07:41 25 August 2014
Tannock (1997) further reported problems with may indeed be important for hockey players actual
inhibitory control in more impulsive participants as performance on the ice, as in our study faster reac-
was evident in their study from signicantly longer tion time to a visual stimulus, faster visual stimulus
stop-signal reaction times. More recently Lage et al. discrimination, better visual memory and a faster
(2011) reported that in handball female athletes ability to shift focus between far and near objects
impulsivity as measured by Conners Continuous signicantly predicted such an important perfor-
Performance Task (CPT-II) and the Iowa mance statistic as goal percentage.
Gambling Task (IGT) was positively correlated
with offensive fouls. The researchers suggested that
this type of non-planning impulsivity results in risky References
decisions that may seem to produce immediate Abernethy, B. (1986). Enhancing sports performance through
rewards but are potentially fraught with longer-term clinical and experimental optometry. Clinical and Experimental
negative consequences (e.g. penalty). Our results Optometry, 69, 189196.
Abernethy, B., Wann, L., & Parks, S. (1998). Training percep-
seem to be in agreement with this conclusion. tual-motor skills for sport. In B. Elliott (Ed.), Training in sport:
Applying sport science (pp. 168). New York, NY: Wiley.
Abernethy, B., & Wood, J. M. (2001). Do generalized visual
Study limitations training programmes for sport really work? An experimental
investigation. Journal of Sports Sciences, 19, 203222.
One of the natural limitations of studying elite ath-
Belsley, D. A., Kuh, E., & Welsch, R. E. (1980). Regression
letes is a highly circumscribed participant pool. diagnostics. New York, NY: Wiley.
Although we were able to test 88.4% of the target Berg, W. P., & Killian, S. M. (1995). Size of the visual eld in
population (hockey players who appeared on the collegiate fast-pitch softball players and nonathletes. Perceptual
roster for the NCAA 20112013 seasons (n = 43)), and Motor Skills, 81, 13071312.
Bulson, R. C., Ciuffreda, K. J., & Hung, G. K. (2008). The effect
the number of participants was still relatively small to
of retinal defocus on golf putting. Ophthalmic and Physiological
allow regression modelling with more than four pre- Optics, 28, 334344.
dictors and was adequate to detect only large effect Chen, J. K., Johnson, K. M., Collie, A., McCroy, P., & Pito, A.
sizes. Thus, some of other potentially important rela- (2007). A validation of the post-concussion symptom scale in
tionships may have been overlooked due to the lack the assessment of complex concussion using cognitive testing
and functional MRI. Journal of Neurology, Neurosurgery, and
of statistical power.
Psychiatry, 78(11), 12311238.
There is always a danger with small sample sizes Ciuffreda, K. J. (2011). Simple eye-hand reaction time in the
that you will be capitalising on chance when using retinal periphery can be reduced with training. Eye & Contact
predictors that have signicant bivariate correlations Lens: Science & Clinical Practice, 37(3), 145146.
with the criterion (the signicance of the correlation Ciuffreda, K. J., & Wang, B. (2004). Vision training and sports. In
G. K. Hung & J. M. Pallis (Eds.), Biomedical engineering princi-
could be due to a chance uctuation in the data), but
ples in sports. New York, NY: Kluwer Academic/Plenum.
in the present study it was probably not the case as Classe, J. G., Semes, L. P., Daum, K. M., Nowakowski, R.,
adjusted indexes of explained variance in the popu- Alexander, L. J., Wisniewski, J., & Bartolucci, A. (1997).
lation (adjusted R2 reported in Table II) remained Association between visual reaction time and batting, elding,
relatively close to the proportions of explained var- and earned run averages among players of the southern baseball
iance in our sample. league. Journal of American Optometric Association, 68(1), 4349.
Coffey, B., & Reichow, A. W. (1989). Athletes vs non-athletes:
Another obvious limitation of the current study Static visual acuity, contrast sensitivity, dynamic visual acuity.
design is the correlational nature of the observed Investigative Ophthalmology and Visual Science, 30(Suppl.), 517.
relationships. Nevertheless, this is one of the rst (abstract).
12 D. Poltavski & D. Biberdorf
Cohen, J. (1988). Statistical power analysis for the behavioral sciences Magill, R. A. (2006). Motor learning and control: Concepts and
(2nd ed.). Hillsdale, NJ: Erlbaum. applications (6th ed.). New York, NY: McGraw-Hill.
Di Russo, F., Pitzalis, S., & Spinelli, D. (2003). Fixation stability Mann, D. Y., Williams, A. M., Ward, P., & Janelle, C. M. (2007).
and saccadic latency in lite shooters. Vision Research, 43, Perceptual-cognitive expertise in sport: A meta-analysis.
18371845. Journal of Sport and Exercise Psychology, 29(4), 457478.
Edman, G., Schalling, D., & Levander, S. E. (1983). Impulsivity Martell, S. G., & Vickers, J. N. (2004). Gaze characteristics of
and speed and errors in a reaction time task: A contribution to elite and near-elite athletes in ice hockey defensive tactics.
the construct validity of the concept of impulsivity. Acta Human movement science, 22, 689712.
Psychologica, 53(1), 18. Mcleod, B. (1991). Effects of eyerobics visual skills training on
Erickson, G. (2007). Sports vision: Vision care for the enhancement of selected performance measures of female varsity soccer players.
sports performance. St. Louis, MO: Butterworth Heineman Perceptual and Motor Skills, 72, 863866.
Elsevier. Millslagle, D. G. (2000). Dynamic visual acuity and coincidence
Erickson, G. B., Citek, K., Cove, M., Wilczek, J., Linster, C., anticipation timing by experienced and inexperienced women
Bjarnason, B., & Langemo, N. (2011). Reliability of a compu- players of fast pitch softball. Perceptual and Motor Skills, 90,
ter-based system for measuring visual performance skills. 498504.
Optometry Journal of the American Optometric Association, 82 Milne, D. C., & Lewis, R. V. (1993). Sports vision screening of
(9), 528542. varsity athletes. Sports Vision, 1, 814.
Faul, F., Erdfelder, E., Buchner, A., & Lang, A. G. (2009). Overney, L. S., Blanke, O., Herzog, M. H., & Burr, D. C. (2008).
Statistical power analyses using G* Power 3.1: Tests for corre- Enhanced temporal but not attentional processing in expert
lation and regression analyses. Behavior Research Methods, 41 tennis players. PLoS One, 3, e2380.
Downloaded by [University of North Dakota] at 07:41 25 August 2014
(4), 11491160. Rine, R. M., Roberts, D., Corbin, B., McKean-Cowdin, R.,
Gardner, L. L., & Sherman, A. (1995). Vision requirements in Varma, R., Beaumont, J., Schubert, M. C. (2012). New
sport. In D. F. Loran & C. MacEwen (Eds.), Sports vision. portable tool to screen vestibular and visual function
Boston, MA: Butterworth-Heinemann. National institutes of health toolbox initiative. The Journal of
Hoffman, L. G., Rouse, M., & Ryan, J. B. (1981). Dynamic visual Rehabilitation Research and Development, 49, 209220.
acuity: A review. Journal of American Optometric Association, 52, Rouse, M. W., DeLand, P., Christian, R., & Hawley, J. (1998). A
883887. comparison study of dynamic visual acuity between athletes
Hughes, P. K., Bhundell, N. L., & Waken, J. M. (1993). Visual and non-athletes. Journal of the American Optometric
and psychomotor performance of elite, intermediate and novice Association, 59(12), 946950.
table tennis competitors. Clinical and Experimental Optometry, Schneider, K., Emery, C., Kang, J., & Meeuwisse, W. (2014). Are
76, 5160. clinical measures of cervical exor endurance, divided attention
Isaacs, L. D., & Finch, A. E. (1983). Anticipatory timing of and computerized dynamic visual acuity different in elite youth
beginning and intermediate tennis players. Perceptual and ice hockey players who report a previous history of concussion
Motor Skills, 57, 451454. compared to those who do not? British Journal of Sports
Ishigaki, H., & Miyao, M. (1993). Differences in dynamic visual Medicine, 48(7), 658.
acuity between athletes and nonathletes. Perceptual and Motor Schwab, S., & Memmert, D. (2012). The impact of a sports vision
Skills, 77(3), 835839. training program in youth eld hockey players. Journal of Sports
Junyent, J. L., & Sole, F. L. (1995). Visual training programme Science and Medicine, 11, 624631.
applied to precision shooting. Ophthalmic and Physiological Seiderman, A., & Schneider, S. (1983). The athletic eye: Improved
Optics, 15, 519523. sports performance through visual training. New York, NY: Hearst
Kluka, D. A., Love, P. A., Sanet, R., Hillier, C., Stroops, S., & Books.
Schneider, H. M. (1995). Contrast sensitivity function prol- Stine, C. D., Arterburn, M. R., & Stern, N. S. (1982). Vision and
ing: By sport and sport ability level. International Journal of sports: A review of the literature. Journal of American Optometric
Sports Vision, 2, 516. Association, 53, 627633.
Kofsky, M., & Stareld, B. (1989). SPORTS vision visual training Uchida, Y., Kudoh, D., Higuchi, T., Honda, M., & Kanosue, K.
and experimental program with Australian institute of sport (2013). Dynamic visual acuity in baseball players is due to
basketball players. Australian Journal of Optometry, 6, 1517. superior tracking abilities. Medicine and Science in Sports and
Kohmura, Y., Aoki, K., Honda, K., Yoshigi, H., & Sakuraba, K. Exercise, 45(2), 319325.
(2008). The relationship between dynamic visual acuity and Voss, M. W., Kramer, A. F., Basak, C., Prakash, R. S., & Roberts,
saccadic eye movement. Human Performance Measurement, 5, B. (2010). Are expert athletes expert in the cognitive labora-
2330. tory? A meta-analytic review of cognition and sport expertise.
Laby, D. M., Davidson, J. L., Rosenbaum, L. J., Strasser, C., Applied Cognitive Psychology, 24(6), 812826.
Mellman, M. F., Rosenbaum, A. L., & Kirschen, D. G. Ward, P., & Williams, A. M. (2003). Perceptual and cognitive skill
(1996). The visual function of professional baseball players. development in soccer: The multidimensional nature of expert
American Journal of Ophthalmology, 122, 476485. performance. Journal of Sport and Exercise Psychology, 25,
Lage, G. M., Gallo, L. G., Cassiano, G. J., Lobo, I. L. B., Viera, 93111.
M. V., Salgado, J. V., & Malloy-Diniz, L. F. (2011). Wood, J. M., & Abernethy, B. (1997). An assessment of the
Correlations between impulsivity and technical performance efcacy of sports vision training programs. Optometry & Vision
in handball female athletes. Psychology, 2(7), 721726. Science, 74(8), 646659.
Logan, G. D., Schachar, R. J., & Tannock, R. (1997). Impulsivity Zwierko, T. (2008). Differences in peripheral perception between
and inhibitory control. Psychological Science, 8(1), 6064. athletes and nonathletes. Journal of Human Kinetics, 19, 5362.