Testing the validity of 360-video for analysing visual exploratory activity in soccer

Full text

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
Testing the validity of 360-video for analysing visual exploratory activity in
1
soccer
2
3
James Feista, Naomi Datsonb*, Oliver R. Runswickc, and Chris Pococka
4
5
aInstitute of Applied Sciences, University of Chichester, U.K
6
bDepartment of Sport and Exercise Sciences, Manchester Metropolitan University Institute of
7
Sport, U.K.
8
cDepartment of Psychology, Institute of Psychiatry Psychology & Neuroscience, King’s
9
College London, U.K.
10
11
12
Correspondence concerning this article should be addressed to Dr Naomi Datson, Institute of
13
Sport, Manchester Metropolitan University, Manchester, UK. Email: N.Datson@mmu.ac.uk
14
15
Open Science Framework Project Link:
16
https://osf.io/ezd24/?view_only=5ed7ce365dea454b9a5b4f7be73b1194
17
18
ORCID
19
James Feist https://orcid.org/0009-0007-7708-925X
20
Naomi Datson https://orcid.org/0000-0002-5507-9540
21
Oliver Runswick https://orcid.org/0000-0002-0291-9059
22
Chris Pocock https://orcid.org/0000-0001-5929-7273
23
24
25
26
27
28
29
30

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
Testing the validity of 360-video for analysing visual exploratory activity in
31
soccer
32
Extended reality (XR) technologies present new opportunities to measure sports performance in
33
immersive and representative environments. Viewed through head-mounted displays (HMDs), 360-
34
video offers the opportunity to capture visual exploratory activity (VEA) using representative stimuli
35
in controlled scenarios. This study aimed to i) assess the construct and face validity of a 360-video
36
simulation for capturing VEA in women’s soccer and ii) understand players’ perceptions
37
of acceptability and tolerability of the simulation. Footage was recorded using a stationary GoPro 360
38
Max camera at eye height in six pitch locations. VEA was measured by the number of ‘scans’ away
39
from the ball before the ball reached the 360-video camera. Eleven sub-elite women’s soccer players
40
and eleven novices viewed 40 soccer videos in a HMD, with videos ending after a pass from a
41
teammate. Upon receiving the pass, participants verbalised and acted an action response. Participants
42
answered open-ended questions on acceptability, physical fidelity, and tolerability. Results supported
43
construct and face validity, with good acceptability, tolerability, and physical fidelity. Soccer players
44
(Mdn = 0.31 scans/s) had significantly higher scan frequencies than novices (Mdn = 0.06 scans/s, p <
45
0.001) and generated significantly more detailed responses per trial (p < 0.001). 360-video offers a
46
valid and acceptable method for capturing VEA and has potential to offer new measures for talent
47
identification processes. Future work should focus on efficacy of 360-video for skill development.
48
Key Words: immersive, scan frequency, women’s football, visual perception, virtual
49
reality, simulation.
50
51
52
53
54
55
56
57
58
59
60
61

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
Introduction
62
In soccer, players must effectively process surrounding information to select the most
63
appropriate action (Pagé et al., 2019). This process relies on effective visual exploratory
64
activity (VEA), defined as a head or body movement where a player’s face is temporarily
65
directed away from the ball to locate teammates, opposition players or empty space, before
66
engaging with the ball (Jordet et al., 2020). Studies have found positive relationships between
67
VEA and pass completion rates in youth men’s (Aksum, Pokolm et al., 2021; Pokolm et al.,
68
2022), professional men’s (Jordet et al., 2013), and women’s soccer (Feist et al., 2024).
69
Skilled players frequently scan their environment to identify nearby opponents, teammates,
70
and potential passing options (Pokolm et al., 2022). However, research into VEA in
71
experimental settings remains limited. One study presented 12 male soccer players with video
72
scenarios on four computer screens positioned behind them, requiring them to identify a “free
73
teammate” after observing a pass on a front-facing screen (McGuckian et al., 2019). Results
74
showed that time constraints significantly influenced head movements as well as a significant
75
relationship between head movements and the speed of a simulated passing response
76
(McGuckian et al., 2019). Whilst this was a novel design, the study’s use of multiple screens
77
lacked realism, highlighting the need for more representative tools. Emerging XR
78
technologies such as 360-video (Höner et al., 2023) and Virtual Reality (VR; Wirth et al.,
79
2021; Wood et al., 2021) present promising avenues for training and testing VEA.
80
360-video is a video recording technique where all directions are recorded at the same
81
time (Kittel et al., 2023). When displayed via a head-mounted display (HMD) users can scan
82
representative environments and change their viewpoint with their head movements (Lindsay
83
et al., 2023). Unlike traditional video, 360-video enables participants the opportunity to
84
explore game-based situations as if they were players in the game (Musculus et al., 2021).
85
This technology has increased the opportunities to study perceptual-cognitive skills such as
86
decision-making in cricket (Discombe et al., 2022), basketball (Pagé et al., 2019), soccer
87
(Höner et al., 2023; Musculus et al., 2021) and boxing (Taupin et al., 2023). Research has
88
utilised 360-video to assess in-game decision-making in soccer, showing that 24 male soccer
89
players rated the motivational effect, acceptability and immersion positively, highlighting
90
benefits of HMDs (Höner et al., 2023). Although the terms 360-video and VR are often used
91
interchangeably, they are separate platforms with different functionality. VR is a computer
92
simulated environment that requires time and programming expertise to develop, which is
93
typically beyond the capacity of many sporting organisations (Panchuk et al., 2018).
94

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
Although 360-video sacrifices interactive elements it can be produced at much lower costs
95
and provides an immersive view of the real world that athletes rate highly for the ability to
96
visually explore a realistic environment (Runswick, 2023). Therefore, 360-video appears to
97
be practical technology for measuring visual exploratory activity.
98
Despite multiple experimental studies investigating VEA in male soccer (e.g.,
99
McGuckian et al. 2019; Aksum, Brotangen et al., 2021), understanding of VEA in women’s
100
soccer remains limited. Research in women’s soccer has focused on the technical and tactical
101
demands of the game (de Jong et al., 2020; Kubayi & Larkin, 2020), with differences found
102
in ball possession tactics between successful and unsuccessful teams (Dipple et al., 2022;
103
O’Donoghue & Beckley, 2023). Successful teams have been found to be more centralised,
104
performing more effective ball movements and transfers (de Jong et al., 2022). An
105
observational study of VEA in elite women’s central midfield players which analysed 30
106
central midfield players during the knock-out stages of UEFA Women’s EURO 2022 (Feist
107
et al., 2024). The study found higher scan frequencies significantly predicted more successful
108
actions with the ball. Scan frequencies were significantly higher in central defensive midfield
109
pitch locations, compared with attacking or wide locations (Feist et al., 2024). In light of
110
these findings, understanding how to measure and train VEA appears crucial. This would
111
help to develop players’ ability to explore their environment effectively and guide subsequent
112
actions with the ball.
113
Following Harris et al.’s (2020) framework for validating simulated environments, an
114
evidence-based approach to developing 360-videos which ensures construct validity
115
(accurately reflecting performance differences; Harris et al., 2021) and face validity (true
116
representation of the task; Bright et al., 2012) is required. Examining construct validity in
117
360-video is crucial to provide an objective measure of a simulated test’s ability to capture
118
elements of sporting performance across skill levels (Harris et al., 2020). Birckhead et al.
119
(2019) provides a methodological framework which assesses users’ perceptions of
120
acceptability and tolerability of a simulation. Acceptability refers to a user’s willingness to
121
try the technology, while tolerability addresses any underreported emotional or physical
122
effects, typically assessed via questions regarding simulation sickness (Birckhead et al.,
123
2019). Understanding these factors is the first step for the use of 360-video to capture VEA in
124
women’s soccer. The present study aims to i) assess the construct and face validity of a 360-
125
video simulation for capturing visual exploratory activity in women’s soccer, and ii)
126
understand players’ perceptions of acceptability and tolerability of a 360-video simulation in
127

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
women’s soccer. For construct validity, we hypothesise that sub-elite women’s soccer players
128
will have significantly higher scan frequencies compared to novices. We further hypothesise
129
that soccer players will provide more varied and detailed verbal descriptions of their next
130
intended action compared to novices.
131
Method
132
Participants
133
An a priori power analysis was conducted using G*Power (version 3.1; Faul et al.,
134
2007) and the effect size (Hedge’s g = 1.13) for distinguishing competitive and social soccer
135
players on a soccer skills test reported by Runswick et al. (2022). With a one-tailed α of 0.05,
136
a power (1-β) of 0.80, a minimum sample size of 20 (10 participants per group) was required
137
to detect this effect. Eleven sub elite female soccer outfield players (M age = 22, SD = 5
138
years) and eleven novices (M age = 20, SD = 2 years) were recruited, with expertise classified
139
based on Swann et al.’s (2015) continuum. Inclusion criteria required participants to be over
140
16 years of age; report normal or corrected to normal vision and be injury-free. Sub-elite
141
outfield soccer players currently competed in Tier 6 or higher in the English women’s
142
football pyramid. Novices had no experience of playing any form of competitive soccer.
143
Ethical approval was obtained from the lead author’s institution and written informed consent
144
was provided by all participants, including those featured in the video stimuli.
145
Filming 360-video soccer stimuli
146
360-video footage was created by filming 9v9 and 7v7 soccer training matches (see
147
Figures 1 and 2). Compared to competitive 11v11 matches, these reduced player numbers
148
allowed all players to be clearly visible in the HMD (see Höner et al., 2023). All visual
149
stimuli were recorded on three-quarters of a full-size pitch using a Go-Pro 360 max (30FPS at
150
5.6k) camera positioned in central areas of the pitch on a stationary tripod at eye height (1.68
151
m from the ground). Pedersen et al. (2019) reported the average height of women in their
152
sample to be 168cm. Therefore, based upon this finding and that of other similar studies
153
camera height (Runswick, 2023; Kittel et al. 2019), the camera was placed 1.68m above the
154
ground at ‘eye height’. This camera angle provided a first-person perspective in the HMD to
155
enhance the sense of being in the game itself.
156
As shown in Figure 1 and 2, the GoPro 360 max camera was positioned in four pitch
157
locations: defensive midfield centre left (DMCL), defensive midfield centre right (DMCR),
158

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
attacking midfield centre right (AMCR) and attacking midfield centre left (AMCL). For each
159
location, the ball began in one of three positions: (1) with the right back, (2) with a throw-in
160
taken by the left back in a defensive midfield location of the pitch or (3) at the feet of the
161
striker in a central attacking pitch location. These starting locations reflected frequent
162
scenarios from the UEFA Women’s EURO 2022 based upon findings from Feist et al.
163
(2024). Players received contextual information about the match (0-0; first half) and were
164
instructed to perform as if they were in a competitive match. Play began with the ‘in-
165
possession’ team (orange bibs) which aimed to pass the ball towards the tripod (with the
166
intention of hitting the tripod). Once a pass struck or came within 1 metre of the tripod,
167
players continued until a whistle signalled the scenario’s end. A total of 108 scenarios across
168
four pitch locations were recorded over four sessions. The lead author reviewed all scenarios,
169
excluding trials in which possession was lost before reaching the camera. Five trials where
170
possession broke down before reaching the camera were randomly selected as ‘washout
171
trials’ for the final testing video. In total, forty scenarios (twenty 9v9 trials and twenty 7v7
172
trials) were selected including the five ‘washout’ trials where possession ended without
173
requiring participant responses. These trials were included to ensure participants remained
174
engaged in the task, but intended actions were recorded for the 35 trials where participants
175
‘received’ the ball.
176
Figure 1
Schematic illustration of the 9 vs 9 soccer training game. The central midfield player
(orange cross located in the white circle) represents the position of the 360-video camera.

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
177
Figure 2
Schematic illustration of the 7 vs 7 soccer training game. The central midfield player
(orange cross located in the white circle) represents the position of the 360-video camera.

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
After selecting the final testing scenarios, videos were imported into Adobe Premier
178
Pro (San Jose, CA, USA) to create two larger testing videos: one 7v7 video and one 9v9
179
video. The videos had a mean duration of eleven minutes and one second. Based on pilot
180
testing, videos were edited to include a five second freeze frame at the beginning, showing
181
the football starting location and attacking direction. Scenario order (pitch location and ball
182
starting locations) was randomised, but remained consistent across participants (Discombe et
183
al., 2022).
184
Apparatus
185
All trials were presented through a HMD (Meta Quest 2) connected to a ASUS
186
G533QS gaming laptop. An adapted strap was used to tightly secure the headset on
187
participants. Trials were played through SkyBox VR on the Meta Quest 2.
188
Procedure
189
All participants attended a single testing session and wore sports clothing, indoor sport
190
trainers, and an orange bib as they would play as a member of the orange team. Participants
191
viewed two separate three minute videos (an operational definitions video and a testing
192
instructions video) in the HMD while standing. Following this, participants completed five
193
self-guided practice trials, similar to that of Höner et al. (2023), to familiarise themselves
194
with the viewing perspective and task requirements (Murphy et al., 2018). Participants were
195
instructed to imagine themselves as a player on the pitch and to observe each scenario until
196
the trial ended.
197
In thirty-five trials, participants received a pass and were instructed to perform a
198
‘shadow’ action with the ball (‘mime’ a physical action of their intended action), similar to
199
Roca et al. (2013) and Discombe et al. (2022) where soccer players mimed soccer actions and
200
batters mimed a ‘shadow’ cricket shot, respectively. After performing their ‘shadow action’,
201
participants verbalised their intended action with the ball and were presented with a list of
202
potential ‘actions’ to provide guidance: ‘Pass’, ‘dribble’, ‘shoot’, ‘receive and protect the
203
ball’, ‘turn with the ball’ and ‘unsure’. For example, a participant might respond verbally, “I
204
would turn with the ball and pass to the left winger”. Participants completed forty trials split
205
into two separate blocks of twenty 9v9 trials and twenty 7v7 trials with a five-minute seated
206
rest between blocks (similar to that of Musculus et al., 2021). The entire procedure lasted 60
207
minutes.
208

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
Actions were recorded in both the real-world (using a Go-Pro Hero 4, 30FPS at 720p)
209
and the 360-video environment (using QuickTime player on an Apple MacBook Pro, Version
210
12.6.3). All trials were analysed using the first person Oculus Footage, with 20% cross-
211
checked against the external Go-Pro footage. After completing the forty trials, participants
212
completed an adapted presence questionnaire (Witmer et al., 2005) and answered open and
213
closed questions to understand the face validity, acceptability, and tolerability of the task.
214
Participants were also asked if they would be interested in using 360-video for future training
215
and testing.
216
Measures
217
Scan frequency. The total number of scans over the final 10 seconds before the ball
218
reached the 360-video camera divided by the elapsed time (Feist et al., 2024).
219
Scan timing. The time in seconds before trial end when players scanned their
220
environment (Feist et al., 2024). Data is presented as mean scan frequencies across the final
221
five seconds prior to participants receiving the ball in the video.
222
Action Type. The type of action with the ball verbalised by participants summarised as
223
frequency scores for both groups. Presented as frequency scores.
224
Action Detail. For every action type, ‘action detail’ was recorded capturing additional
225
information provided in their response. For example, if a player responded, “I would turn
226
with the ball, dribble down the left wing and cross the ball”, the recorded action type would
227
be ‘turn with the ball’ with two additional action details (‘dribble’ and ‘cross’). This measure
228
is presented as frequency scores.
229
Number of actions generated per trial. Dividing the total number of actions
230
verbalised by the number of trials completed.
231
Number of action details generated per trial. Dividing the total number of additional
232
action details verbalised by the number of trials.
233
Presence. An adapted 22 item presence questionnaire (Witmer et al., 2005), excluding
234
touch was used rated on a seven-point scale across six factors: possibility to act, possibility to
235
examine, realism, quality of interface, sounds and self-evaluation of performance. Scores
236
were calculated per the questionnaire’s guidance
237

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
Acceptability, tolerability, face validity and fidelity of the task. Open and closed
238
questions (adapted from Chertoff et al., 2010 and Höner et al., 2023) were asked to all
239
participants. Sample questions included: ‘How well did you feel you were able to move your
240
head?’ (see Table 1).
241
Table 1
242
Follow up questions asked to participants after completing the 360-video soccer task
243
Question/Measure
Category
How well did you feel you were able to move your head?
How involved did you feel in the match situation?
Did the task lead you to experience any feelings of nausea
or sickness?
How much did the 360-video trials look like real-life
football?
Would you use this 360-video simulation again?
How often would you use this 360-video simulation?
Please respond in number of times per week: 0, 1-2, 3-4,
5-6 or 7.
How much did the 360-video feel like real life football?
What would you use the 360-video footage for?
Is there anything that you think would prevent you from
using 360-videoin football?
What would be important to a good football training
session using 360-video?
Physical Fidelity
Face Validity
Tolerability
Face Validity
Acceptability
Acceptability
Face Validity
Acceptability
Tolerability
Acceptability
244
Data Analysis
245
Reliability
246
A senior lecturer in sport psychology with prior VEA coding experience conducted
247
additional coding on all variables to assess inter-rater reliability. A total of 132 trials (15% of
248
all trials), were re-analysed for inter and intra-rater reliability aligning with previous VEA
249

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
research (Aksum, Pokolm et al., 2021; Feist et al., 2024). Intra-rater reliability was tested
250
following a six-week gap to minimise potential learning effects. Intra-class correlations (ICC)
251
were calculated for the continuous variable ‘number of scans’, the basis for scan frequency and
252
were assessed following Cicchetti (1994) criteria to determine the strength of agreement
253
between different coders and repeated coder observations (see Table 2).
254
Table 2
255
Intra-class correlations for number of scans (continuous variable)
256
257
258
259
Statistical Analysis
260
Normality was assessed using the Shapiro-Wilk test, histograms, boxplots, and
261
zskewness/zkurtosis with ±1.96 criteria applied (O’Donoghue, 2013). Between-group
262
comparisons of questionnaire items used independent samples t-tests for normal data and
263
Mann-Whitney U tests for non-normal data. Levene’s test confirmed equal variances (p >
264
0.05). Mann-Whitney U tests compared scan frequency, actions per trial and action details
265
per trail between groups, with medians and interquartile ranges reported. A two-way mixed
266
ANOVA examined scan timing differences in the final five seconds before ball contact. A 2
267
Group (soccer players, novices) x 6 verbal action response category (pass, shot, dribble,
268
receive and protect the ball, turn with the ball and unsure) ANOVA with Greenhouse-Geisser
269
correction was performed for action type and action detail, with the assumption of sphericity
270
being violated for both tests. Verbal action response categories were treated as repeated
271
measures, similar to that of Roca et al. (2011). Bonferroni-adjusted t-tests were used to
272
determine the source of the effect. Effect sizes for ANOVAs (partial eta squared) were small
273
(≈.01), medium (≈.06), large (≈.14) (Cohen, 1988) and for t-tests (Cohen’s d): small (0.20–
274
0.49), medium (0.50–0.79), large (≥0.80) (Cohen, 1992). Rank Biserial-Correlation (range: -1
275
to +1) provided further measures of effect size. The alpha level was α = 0.05, and analyses
276
were conducted in JASP (version 0.16.4).
277
278
279
Inter-rater
Intra-rater
Variable
ICC (95% CI)
Strength of
Agreement
ICC (95% CI)
Strength of
Agreement
Number
of scans
0.902
(0.865-0.930)
Excellent
0.953
(0.934-0.966)
Excellent

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
Results
280
All participants reported good levels of presence (for presence questionnaire data, see
281
supplementary material).
282
Construct Validity
283
Scan Frequency
284
Soccer players performed significantly higher scan frequencies (Mdn = 0.31 scans/s,
285
IQR = 0.155) compared with novices (Mdn = 0.06 scans/s, IQR = 0.040; U = 10.50, p <
286
0.001, rb = -0.83; Figure 3).
287
Figure 3
Scatter bar displaying median scan frequency (scans/s) between soccer players and novices.
Bars represent median scan frequency scores by skill level. Black dots represent individual
data by participant
Scan Timing
288
For soccer players, the highest mean scan frequency was observed between 1.01 - 3
289
seconds and for novices was between ball contact - 1 second and between 4.01 - 5 seconds
290
prior to receiving a pass from a teammate (see Figure 4). A significant main effect of skill
291

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
level on scan timing, F(1, 20) = 16.68, p < 0.001, η² = 0.364 was found with soccer players
292
scanning significantly more often than novices. There was no significant main effect of
293
time, F(4, 80) = 0.55, p = 0.703, η² = 0.005, and no significant interaction between scan timing
294
and skill level, F(4, 80) = 0.74, p = 0.565, η² = 0.007.
295
Verbal action responses
296
Number of actions and number of action details generated per trial
297
Soccer players generated significantly more actions per trial (Mdn = 1.30, IQR = 0.25)
298
compared to novices (Mdn = 1.00, IQR = 0.05, U = 31.50, p = 0.028). Soccer players also
299
generated more action details per trial (M = 1.06, SD = 0.07) compared to novices (M = 0.45,
300
SD = 0.35, t10.899 = 5.653, p < 0.001, d = 2.410). The number of actions and number of action
301
details generated per trial data is presented in Figure 5.
302
Action Type
303
Figure 4
Means and Standard Errors (presented as error bars) across the final five seconds prior to
receiving the ball for soccer players and novices
0
0.04
0.08
0.12
0.16
0.2
0.24
0.28
0.32
0.36
0.4
Ball Contact -
1 second 1.01 - 2
seconds 2.01 - 3
seconds 3.01 - 4
seconds 4.01 - 5
seconds
Mean Scan Frequency (Scans/second)
Seconds Prior to Ball Contact
Soccer Players Novices

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
Results indicated a significant main effect of verbal action response category, F(2.37,
304
47.37) = 69.09, p < 0.001, η² = 0.755. Bonferroni-corrected follow up test comparisons
305
demonstrated that participants verbalised the action of pass significantly more than all other
306
action categories (p < 0.001). There was no significant main effect of skill level, F(1, 20) =
307
3.30, p = 0.084, η² = 0.003, and no significant interaction between verbal action response
308
category and skill level, F(2.37, 47.37) = 0.49, p = 0.648, η² = 0.005.
309
Figure 5
Scatter bars displaying mean number of verbal action responses (a) and the mean number of
verbal action response details per trial (b) between soccer players and novices
a b
Action Detail
310
There was a significant main effect of verbal action response detail category, F(2.30, 46.09)
311
= 24.26, p < 0.001, η² = 0.450. Follow up test comparisons demonstrated that participants
312
verbalised the action detail of pass significantly more than any other action categories. There
313
was a significant main effect of skill level, F(1, 20) = 28.25, p < 0.001, η² = 0.050 with soccer
314
players verbalising significantly more action details compared to novices. A significant
315
interaction between verbal action response detail category and skill level, F(2.30, 46.09) = 0.49, p
316
= 0.008, η² = 0.093 was found.. Table 3 contains soccer players’ and novices verbal action
317
detail.
318
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
Soccer Players Novices
Mean action details generated per trial
Skill Level
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
Soccer Players Novices
Mean actions generated per trial
Skill Level

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
Table 3
Descriptive analysis of soccer players’ action response verbalisations
Frequency
Action Detail
Action Type
Soccer
Players
Novices
Soccer
Players
Novices
Pass
257
227
228
86
Shot
52
57
3
2
Dribble
118
91
104
45
Receive and protect the ball
13
11
9
2
Turn with the ball
46
37
63
48
Unsure
0
2
0
0
Total
486
425
407
183
Face Validity & Fidelity
319
All soccer players commented on how they were able to move their head freely when
320
wearing the Meta Quest 2 with two players stating that it took them a short amount of time to
321
adjust to wearing a headset. Soccer players shared how the soccer video task felt and looked
322
like real-life soccer with clear visuals of players on the pitch and match realistic sounds.
323
Thematic analysis capturing participants responses can be found in Figure 6.
324
Figure 6
325
Dimensions and Themes that emerged from questions on soccer players perceptions of face
326
validity and physical fidelity of the 360-video soccer simulation task
327

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
328
Acceptability & Tolerability
329
No participants reported motion sickness from the 360-video soccer video stimuli. All
330
soccer players reported that they would be interested in using 360-video in training and
331
testing. When asked how often players would use 360-video, responses ranged from one per
332
month to one-to-two times per week. Nine soccer players explicitly shared the importance of
333
using match-realistic scenarios which could be evaluated with a coach as part of team-based
334
video analysis. Thematic analysis capturing participants responses can be found in Figure 7.
335
Figure 7
336
Dimensions and Themes that emerged from questions on soccer players acceptability and
337
tolerability of the 360-video soccer simulation task
338

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
339
Discussion
340
The study aimed to assess the construct and face validity of a 360-video simulation for
341
capturing VEA in women’s soccer and to understand perceptions of acceptability and
342
tolerability of the task. Results indicated the newly developed 360-video soccer task
343
demonstrates construct and face validity. Soccer players exhibited significantly higher scan
344
frequencies and generated significantly more verbal actions with the ball per trial compared
345
to novices, supporting construct validity. No significant differences were reported across any
346
of the presence questionnaire items, with all participants reporting moderate to high presence
347
in the environment. Overall, the 360-video task indicated construct and face validity was
348
achieved, with good acceptability, tolerability and physical fidelity.
349
As hypothesised, sub-elite soccer players displayed significantly higher median scan
350
frequencies compared to novices. This suggests players actively scanned their environment
351
for critical information to inform actions upon receiving the ball (Aksum, Pokolm et al.,
352
2021). Studies in men’s soccer link higher scan frequencies to improved performance with
353

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
the ball and expertise (McGuckian et al., 2018). In the current study, soccer players highest
354
scan frequencies were between 1.01 – 3 seconds compared to novices’ highest scan
355
frequencies between ball-contact – 1 second and 4.01 – 5 seconds. Once the trials started,
356
novices tended to ‘ball watch’ and would typically only scan their environment as the ball
357
approached, suggesting that novices’ scanning was more reactive, compared to soccer
358
players. These findings demonstrate minor differences in scan timing between the two skill
359
level groups, with soccer players scanning significantly more than novices. Lastly, soccer
360
players generated more action responses per trial and more action details compared to
361
novices. One possible explanation for this is that by scanning their environment more
362
frequently, soccer players were able to generate richer responses on subsequent actions with
363
the ball compared to novices. These findings align with previous research where skilled
364
athletes produced more task-relevant options and detailed verbal responses compared to
365
novices (Murphy et al., 2019; Roca et al., 2011). Therefore, this 360-video task appears
366
representative of real-life soccer by its ability to distinguish between soccer players and
367
novices across measures of VEA and verbal action responses and so may be a valuable tool in
368
assessing VEA in women soccer.
369
Both soccer players and novices reported good levels of presence where participants
370
scored highest for levels of realism and lower for possibility to act. This evidence suggests
371
soccer players perceive the 360-video environment as somewhat immersive indicating its
372
potential as a suitable tool for assessing players’ VEA in match-realistic situations. To
373
understand soccer players perceptions of face validity and physical fidelity open-ended
374
questions were asked to all soccer players. Seven of the eleven soccer players stated they
375
could move their heads and scan their environment freely with the Meta Quest 2 headset,
376
feeling immersed in the match situation suggesting good physical fidelity. This will likely
377
continue to be improved with newer, lighter headsets. Previous research on 360-video’s
378
effectiveness in enhancing decision-making skills among Australian football umpires found
379
athletes reported greater task engagement compared with viewing traditional broadcast
380
footage (Kittel, Larkin, Elsworthy et al., 2020), supporting the immersive feel of 360-video.
381
However, players described limitations such as the ball not being at their feet in the testing
382
room and the inability to move within the 360-video environment. Research highlights
383
primary limitations of 360-video including restricted perception-action loop (i.e. action
384
fidelity) and reliance on stationary footage (Kittel, Larkin, Cunningham et al., 2020). Thus,
385
future research should explore mixed reality benefits which may facilitate perception-action
386

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
links (Kittel et al., 2021). Overall, feedback indicates soccer players perceive the simulation
387
as immersive, suggesting a moderate to high level of presence and face validity.
388
Following guidelines for developing simulated environments (Birckhead et al., 2019),
389
the study assessed participants perceptions of acceptability and tolerability of the task. All
390
soccer players reported no motion sickness and all soccer players expressed interest in using
391
360-video for training and testing purposes. Soccer players frequently mentioned 360-video
392
as a tool to support physical and team-based training suggesting it could be used 1-2 times
393
per week. Previous research found 91% of male soccer players viewed 360-video as a
394
potential training tool (Musculus et al., 2021), with further research reporting soccer players
395
demonstrated positive ratings for motivational effect, acceptability and immersion in a 360-
396
video for decision making (Höner et al., 2023). This evidence suggests 360-video may aid in
397
understanding perceptual-cognitive skills in soccer with both men’s and women’s players
398
indicating high willingness to use the simulation for training and testing. Soccer players
399
suggested cost, lack of in-game movement and time availability as potential barriers to 360-
400
video use. Despite players perceiving 360-video to be high in cost, research suggests that
401
developing 360-video stimuli and importing this into a HMD is a lower cost option compared
402
to creating custom VR software (Kittel, Larkin, Cunningham et al., 2020; Barbour et al.,
403
2024). To summarise, no participants reported motion sickness indicating good tolerability
404
and although soccer players shared potential barriers to the use of 360-video, players also
405
emphasised its value to develop perceptual-cognitive skills. With players expressing a
406
willingness to use 360-video again, the task appears to demonstrate good acceptability and
407
tolerability.
408
Study Limitations & Future Research Directions
409
A limitation of current study is that the soccer players recruited were sub-elite rather
410
than elite. As a result, caution is warranted when generalising the findings to more elite
411
populations. Future research should aim to investigate VEA using 360-video with a more
412
elite cohort of players to better enhance the applicability and transferability of the technology
413
for measuring VEA. Furthermore, consistent with previous literature, asking participants to
414
verbalise their actions and act out soccer specific movements may not have captured their full
415
capabilities (Panchuk et al., 2018; Dicks et al., 2010). While the task distinguished between
416
soccer players and novices in scan frequency and the number of actions generated per trial,
417
with evidence of face validity and immersion, future research is still necessary to further
418

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
validate this simulation. This study provides initial evidence that 360-video may be a useful
419
tool for testing VEA in women’s soccer, however additional research is still needed to
420
examine other forms of fidelity, such as psychological and biomechanical fidelity to
421
understand whether there is any opportunity for training and transfer of learning to soccer
422
performance (Harris et al., 2020). This presents an opportunity to use 360-video to simulate
423
match-realistic game situations and conduct further experimental research in women’s soccer.
424
Practical Implications
425
Based on the study’s findings, we propose some practical implications. Practitioners
426
should consider using first-person game footage as an individualised tool, incorporating
427
additional contextual and perceptual factors to challenge soccer players. Our results suggest
428
soccer players view 360-video as a beneficial addition to physical team-based training. With
429
360-video enabling multiple repetitions of in-game scenarios without injury or fatigue risks
430
(Musculus et al., 2021), this technology could also support rehabilitation for players returning
431
to play from injury (Musculus et al., 2021) or illness.
432
Conclusion
433
This study assessed the construct and face validity of a 360-video simulation for
434
capturing VEA in women’s soccer and to understand players’ perceptions of acceptability
435
and tolerability of the task. Following Harris et al. (2020) and Birckhead et al. (2019)
436
guidelines, we used an evidence-based approach to test the validity of a 360-video soccer
437
simulation. Results demonstrated construct validity with significant differences in scan
438
frequency and the number of actions generated per trial between soccer players and novices.
439
Soccer players had significantly higher scan frequencies and generated significantly more
440
verbal action responses per trial compared to novices. Participants rated the task highly for
441
acceptability, tolerability and physical fidelity, with soccer players sharing expressing
442
immersion in the task. These findings offer preliminary evidence that this 360-video task may
443
be sufficiently representative of soccer for visually examining the environment suggesting it
444
could serve an alternative to traditional video-based methods in understanding how female
445
soccer players visually explore their environment. Future research should now further
446
validate the use of 360-video as a tool for training and testing in women’s soccer.
447
448
449

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
References
450
Aksum, K. M., Brotangen, L., Bjørndal, C. T., Magnaguagno, L., & Jordet, G. (2021).
451
Scanning activity of elite football players in 11 vs. 11 match play: An eye-tracking
452
analysis on the duration and visual information of scanning. Plos one, 16(8), e0244118.
453
https://doi.org/10.1371/journal.pone.0244118
454
Aksum, K. M., Pokolm, M., Bjørndal, C. T., Rein, R., Memmert, D., & Jordet, G. (2021).
455
Scanning activity in elite youth soccer players. Journal of Sports Sciences, 39(21),
456
2401–2410. https://doi.org/10.1080/02640414.2021.1935115
457
Barbour, B., Sefton, L., Bruce, R. M., Valmaggia, L., & Runswick, O. R. (2024). Acute
458
psychological and physiological benefits of exercising with virtual reality. Plos one
459
19(12): e0314331. https://doi.org/10.1371/journal.pone.0314331
460
Birckhead, B., Khalil, C., Liu, X., Conovitz, S., Rizzo, A., Danovitch, I., ... & Spiegel, B.
461
(2019). Recommendations for methodology of virtual reality clinical trials in health
462
care by an international working group: iterative study. JMIR mental health, 6(1), 1-14.
463
https://mental.jmir.org/2019/1/e11973/
464
Bright, E., Vine, S., Wilson, M. R., Masters, R. S. W., and McGrath, J. S. (2012). Face
465
validity, construct validity and training benefits of a virtual reality turp simulator. Int. J.
466
Surg, 10(3), 163–166. https://doi.org/10.1016/j.ijsu.2012.02.012
467
Chertoff, D. B., Goldiez, B., & LaViola, J. J. (2010). Virtual Experience Test: A virtual
468
environment evaluation questionnaire. In 2010 IEEE virtual reality conference (VR),
469
103-110. doi:10.1109/VR.2010.5444804
470
Cicchetti, D. (1994). Guidelines, criteria, and rules of thumb for evaluating normed and
471
standardized assessment instruments in psychology. Psychological Assessment, 6(4),
472
284–290. https://doi.org/10.1037/1040-3590.6.4.284
473
Cohen, J. (1992). Statistical power analysis. Current directions in psychological
474
science, 1(3), 98-101. https://doi.org/10.1111/1467-8721.ep10768783
475
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale,
476
NJ: Erlbaum.
477
de Jong, L. M., Gastin, P. B., Angelova, M., Bruce, L., & Dwyer, D. B. (2020). Technical
478

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
determinants of success in professional women’s soccer: A wider range of variables
479
reveals new insights. PloS one, 15(10), 1-12.
480
https://doi.org/10.1371/journal.pone.0240992
481
de Jong, L. M. S., Gastin, P. B., Bruce, L., & Dwyer, D. B. (2022). Teamwork and
482
performance in professional women’s soccer: A network-based analysis. International
483
Journal of Sports Science & Coaching, PloS one 15(10): e0240992
484
https://doi.org/10.1177/17479541221092355
485
Dicks, M., Button, C. and Davids, K. (2010). Examination of gaze behaviors under in situ
486
and video simulation task constraints reveals differences in information pickup for
487
perception and action. Atten. Percept. Psychophys, 72, 706–720.
488
https://doi.org/10.3758/APP.72.3.706
489
Dipple, J. W., Bruce, L., & Dwyer, D. B. (2022). Identifying the optimal characteristics of
490
ball possession and movement in elite women’s soccer. International Journal of
491
Performance Analysis in Sport, 22(4), 594-603.
492
https://doi.org/10.1080/24748668.2022.2101837
493
Discombe, R. M., Bird, J. M., Kelly, A., Blake, R. L., Harris, D. J., & Vine, S. J. (2022).
494
Effects of traditional and immersive video on anticipation in cricket: A temporal
495
occlusion study. Psychology of Sport and Exercise, 58, 102088.
496
https://doi.org/10.1016/j.psychsport.2021.102088
497
Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G* Power 3: A flexible statistical
498
power analysis program for the social, behavioral, and biomedical sciences. Behavior
499
research methods, 39(2), 175-191. https://doi.org/10.3758/BF03193146
500
Feist, J., Datson, N., Runswick, O. R., Harkness-Armstrong, A., & Pocock, C. (2024). Visual
501
exploratory activity in elite women’s soccer: an analysis of the UEFA Women’s
502
European Championship 2022. International Journal of Sport and Exercise Psychology,
503
1-23. https://doi.org/10.1080/1612197X.2023.2300386
504
Harris, D. J., Bird, J. M., Smart, P. A., Wilson, M. R., & Vine, S. J. (2020). A framework for
505
the testing and validation of simulated environments in experimentation and
506
training. Frontiers in Psychology, 11:605, 1-10.
507
https://doi.org/10.3389/fpsyg.2020.00605
508

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
Harris, D. J., Buckingham, G., Wilson, M. R., Brookes, J., Mushtaq, F., Mon-Williams, M.,
509
& Vine, S. J. (2021). Exploring sensorimotor performance and user experience within a
510
virtual reality golf putting simulator. Virtual Reality, 25, 647-654.
511
https://doi.org/10.1007/s10055-020-00480-4
512
Höner, O., Dugandzic, D., Hauser, T., Stügelmaier, M., Willig, N., & Schultz, F. (2023). Do
513
you have a good all-around view? Evaluation of a decision-making skills diagnostic tool
514
using 360-videos and head-mounted displays in elite youth soccer. Frontiers in Sports
515
and Active Living, 5, 1-11. https://doi.org/10.3389/fspor.2023.1171262
516
Jordet, G., Aksum, K. M., Pedersen, D. N., Walvekar, A., Trivedi, A., McCall, A., Ivarsson,
517
A., & Priestley, D. (2020). Scanning, Contextual Factors, and Association With
518
Performance in English Premier League Soccerers: An Investigation Across a Season.
519
Frontiers in Psychology, 11, 1–16.
520
https://doi.org/10.3389/fpsyg.2020.553813
521
Jordet, G., Bloomfield, J., & Heijmerikx, J. (2013). The Hidden Foundation of Field Vision
522
in English Premier League (EPL) Soccer Players. MIT Sloan Sports Analytics
523
Conference, Boston.
524
Kittel, A., Larkin, P., Cunningham, I., & Spittle, M. (2020). 360 virtual reality: a SWOT
525
analysis in comparison to virtual reality. Frontiers in Psychology, 11, 1-5.
526
https://doi.org/10.3389/fpsyg.2020.563474
527
Kittel, A., Larkin, P., Elsworthy, N., Lindsay, R., & Spittle, M. (2020). Effectiveness of 360
528
virtual reality and match broadcast video to improve decision-making skill. Science and
529
Medicine in Football, 4(4), 255-262. https://doi.org/10.1080/24733938.2020.1754449
530
Kittel, A., Larkin, P., Elsworthy, N., & Spittle, M. (2021). Transfer of 360 virtual reality and
531
match broadcast video-based tests to on-field decision-making. Science and Medicine
532
in Football, 5(1), 79-86. https://doi.org/10.1080/24733938.2020.1802506
533
Kittel, A., Larkin, P., Elsworthy, N., & Spittle, M. (2019). Using 360 virtual reality as a
534
decision-making assessment tool in sport. Journal of science and medicine in
535
sport, 22(9), 1049-1053. https://doi.org/10.1016/j.jsams.2019.03.012
536

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
Kittel, A., Spittle, M., Larkin, P., & Spittle, S. (2023). 360° VR: Application for exercise
537
and sport science education. Frontiers in Sports and Active Living, 5, 1-6.
538
https://doi.org/10.3389/fspor.2023.977075
539
Kubayi, A., & Larkin, P. (2020). Technical performance of soccer teams according to match
540
outcome at the 2019 FIFA Women’s World Cup. International Journal of Performance
541
Analysis in Sport, 20(5), 908–916. https://doi.org/10.1080/24748668.2020.1809320
542
Lindsay, R., Spittle, S., & Spittle, M. (2023). Skill adaption in sport and movement:
543
Practice design considerations for 360 VR. Frontiers in Psychology, 14, 1-5.
544
https://doi.org/10.3389/fpsyg.2023.1124530
545
O'Donoghue, P. (2013). Statistics for sport and exercise studies: An introduction.
546
Routledge.
547
O’Donoghue, P., & Beckley, S. (2023). Possession tactics in the UEFA women’s EURO
548
2022 soccer tournament. International Journal of Performance Analysis in Sport, 23(1),
549
48-64. https://doi.org/10.1080/24748668.2023.2206273
550
McGuckian, T. B., Cole, M. H., Chalkley, D., Jordet, G., & Pepping, G. J. (2019). Visual
551
exploration when surrounded by affordances: frequency of head movements is
552
predictive of response speed. Ecological Psychology, 31(1), 30-48.
553
https://doi.org/10.1080/10407413.2018.1495548
554
McGuckian, T. B., Cole, M. H., Jordet, G., Chalkley, D., & Pepping, G. J. (2018). Don’t turn
555
blind! The relationship between exploration before ball possession and on-ball
556
performance in association soccer. Frontiers in Psychology, 9, 1–13.
557
https://doi.org/10.3389/fpsyg.2018.02520
558
Murphy, C. P., Jackson, R. C., & Williams, A. M. (2019). Informational constraints, option
559
generation, and anticipation. Psychology of Sport and Exercise, 41, 54-62.
560
https://doi.org/10.1016/j.psychsport.2018.11.012
561
Murphy, C. P., Jackson, R. C., & Williams, A. M. (2018). The role of contextual information
562
during skilled anticipation. Quarterly Journal of Experimental Psychology, 71(10),
563
2070–2087. https://doi.org/10.1177/1747021817739201
564
Musculus, L., Bäder, J., Sander, L., & Vogt, T. (2021). The influence of environmental
565
constraints in 360-videos on decision making in soccer. Journal of Sport and Exercise
566

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
Psychology, 43(5), 365-374. https://doi.org/10.1123/jsep.2020-0166
567
Pagé, C., Bernier, P. M., & Trempe, M. (2019). Using video simulations and virtual reality to
568
improve decision-making skills in basketball. Journal of sports sciences, 37(21),
569
2403-2410. https://doi.org/10.1080/02640414.2019.1638193
570
Panchuk, D., Klusemann, M. J., & Hadlow, S. M. (2018). Exploring the effectiveness of
571
immersive video for training decision-making capability in elite, youth basketball
572
players. Frontiers in psychology, 9, 1-9. https://doi.org/10.3389/fpsyg.2018.02315
573
Pedersen, A. V., Aksdal, I. M., & Stalsberg, R. (2019). Scaling demands of soccer according
574
to anthropometric and physiological sex differences: a fairer comparison of men’s and
575
women’s soccer. Frontiers in psychology, 10, 1-11.
576
https://doi.org/10.3389/fpsyg.2019.00762
577
Pokolm, M., Rein, R., Müller, D., Nopp, S., Kirchhain, M., Aksum, K. M., ... & Memmert,
578
D. (2022). Modeling Players’ Scanning Activity in Football. Journal of Sport and
579
Exercise Psychology, 44(4), 263-271. https://doi.org/10.1123/jsep.2020-0299
580
Roca, A., Ford, P. R., McRobert, A. P., & Williams, A. M. (2013). Perceptual-cognitive
581
skills and their interaction as a function of task constraints in soccer. Journal of Sport
582
and Exercise Psychology, 35(2), 144-155. https://doi.org/10.1123/jsep.35.2.144
583
Roca, A., Ford, P. R., McRobert, A. P., & Williams, A. M. (2011). Identifying the processes
584
underpinning anticipation and decision-making in a dynamic time-constrained
585
task. Cognitive processing, 12, 301-310. https://doi.org/10.1007/s10339-011-0392-1
586
Runswick, O. R. (2023). Player Perceptions of Face Validity and Fidelity in 360-Video and
587
Virtual Reality Cricket. Journal of Sport and Exercise Psychology, 45(6), 347-354.
588
https://doi.org/10.1123/jsep.2023-0122
589
Runswick, O. R., Rawlinson, A., Datson, N., & Allen, P. M. (2022). A valid and reliable
590
test of technical skill for vision impaired football. Science and Medicine in
591
Football, 6(1), 89–97. https://doi.org/10.1080/24733938.2021.1885725
592
Swann, C., Moran, A., & Piggott, D. (2015). Defining elite athletes: Issues in the study of
593
expert performance in sport psychology. Psychology of sport and exercise, 16, 3-14.
594
https://doi.org/10.1016/j.psychsport.2014.07.004
595

This manuscript has not been Pre-print
peer reviewed (07/02/2025)
Taupin, M, L., Romeas, T., Juste, L., & Labbé, D. R. (2023). Exploring the effects of 3D-
596
360-videoVR and 2D viewing modes on gaze behavior, head excursion and workload
597
during a boxing specific anticipation task. Frontiers in Psychology, 14, 1-12.
598
https://doi.org/10.3389/fpsyg.2023.1235984
599
Wirth, M., Kohl, S., Gradl, S., Farlock, R., Roth, D., & Eskofier, B. M. (2021). Assessing
600
visual exploratory activity of athletes in virtual reality using head motion
601
characteristics. Sensors, 21(11), 3728. https://doi.org/10.3390/s21113728
602
Witmer, B. G., Jerome, C. J., & Singer, M. J. (2005). The factor structure of the presence
603
questionnaire. Presence: Teleoperators & Virtual Environments, 14(3), 298-312.
604
https://doi.org/10.1162/105474605323384654
605
Wood, G., Wright, D. J., Harris, D., Pal, A., Franklin, Z. C., & Vine, S. J. (2021). Testing the
606
construct validity of a soccer-specific virtual reality simulator using novice, academy,
607
and professional soccer players. Virtual Reality, 25, 43-51.
608
https://doi.org/10.1007/s10055-020-00441-x
609