Professional Documents
Culture Documents
PREPRINT GonalvesandMendona2023
PREPRINT GonalvesandMendona2023
net/publication/368748093
CITATIONS READS
0 34
2 authors, including:
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Effect of Expectation on Cognitive Training in Action Games in Virtual Reality Videogames on Visual Attention View project
The Effects of Practice in Action Games on Attention and Memory View project
All content following this page was uploaded by Edimilson dos Santos Gonçalves on 23 February 2023.
https://doi.org/10.17605/OSF.IO/R6JAH
memory
Abstract: Objective: The present study aimed to verify the effects of types of engagement with action
games in the optimization of attentional resources and working memory. Method: A computerized
version of the Continuous Performance Test (CPT), with three different levels of complexity, was
applied in two phases to a sample of 85 participants, divided into four groups according to their weekly
engagement with action games on PlayStation platforms and Xbox. Dependent variables included:
discriminability value (d′), criteria (c), reaction time, and recall rate in the working memory task.
Results: The results showed a significant effect between complexities for each phase. In addition, the
results showed that participants who engaged more than 11 hours a week in action games performed
better compared to players who dedicated themselves to 10 hours or less. Furthermore, players who
play between 1 to 3 times a week achieved better results compared to players who dedicate themselves
4 times a week or more. Conclusion: These results suggest that the number of times/hours devoted
weekly to action games is an important factor in achieving better performance in attention and working
memory tests by video game players. This study is important to evaluate the effects of games on
cognition and their contributions to rehabilitation practices based on cognitive training with games.
1
Master's student in Cognition and Neurosciences at the Graduate Program in Behavioral Sciences at the University of
Brasília (PPG/CdC/UnB). Email: edmilsonds48@gmail.com (ORCID: https://orcid.org/0000-0003-0864-1649)
2
Adjunct professor linked to the Department of Basic Psychological Processes at the University of Brasília (PPB/UnB).
Email: goiaracastilho@gmail.com (ORCID: https://orcid.org/0000-0001-9132-0859)
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
2
Keywords: Action games, attention, working memory, cognitive training, video games.
Authors' Notes
We have no known conflict of interest to disclosure. This study has not yet received a peer review.
Public significance statements: This study enabled a better understanding of the number of hours
required for attention and working memory improvement in video game players. This study brings
important data to the area, as well as contributing to studies that aim to create research and cognitive
training interventions based on video games. Indicating the best configuration for training practices
with games.
of the University of Brasília (PPB/IP/UnB). Thanks to Marcos Felipe Rodrigues de Lima for his
suggestions regarding data analysis. Thanks to the Federal District Research Support Foundation
(FAPDF) for the financial contribution in the form of a scientific initiation scholarship, which made
Financing: This work was supported by the Department of Basic Psychological Processes of the
Institute of Psychology of the University of Brasília (PPB/IP/UnB), by the Dean of Graduate Studies
at the University of Brasília (DPG/UnB) and by the Federal District Research Support Foundation
(FAPDF).
Introduction
Video games demand complex processing of simultaneous information that depends on the
cognitive functions of Video Game Players (VGJPs; Bediou et al., 2018; Latham et al., 2013; Reynaldo
et al., 2021; Stanmore et al., 2017). Research has suggested that video games, especially action games,
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
3
can improve several cognitive skills, with emphasis on perceptual skills (Bediou et al., 2018; Bejjanki
et al., 2014; Sajan et al., 2017), attention (Föcker et al., 2019; Figueiredo & Sbissa, 2013) and working
memory (Ballesteros et al., 2017; Colzato et al., 2013; Toril et al., 2016), among other functions
compiled in meta-analyses in the last decade (e.g., Latham et al., 2013; Stanmore et al., 2017).
Among the various cognitive skills researched, attention and working memory have gained
prominence (Rivero et al., 2012; Sarmet & Pilati, 2016). Attention is an essential competence for
VGPs, since video games present several stimuli in rapid sequence, both sound and visual, with the
filtering of those essentially necessary for a good performance (Figueiredo & Sbissa, 2013). Likewise,
the use of memory resources, especially working memory, is essential for this practice (Ballesteros et
al., 2017; Colzato et al., 2013; Toril et al., 2016), since this type of memory includes a temporary
storage buffer that allows manipulating information relevant to the task (Baddeley, 2012).
As a result of this complex cognitive demand, recent research (e.g., Bavelier et al., 2012; Bhat
et al., 2019; Figueiredo & Sbissa, 2013; Green & Bavelier, 2006; Palaus et al., 2020; Rivero et al.,
2012; Toril et al., 2016) point to the development of better skills among VGPs, such as greater
cognitive flexibility, better performance in mental rotation tasks, reduced processing time, among
others. These results suggest that the high interactivity of games contributes to VGPs developing a
better ability to deal with a possible cognitive overload in a multitasking environment (Ballesteros et
Considering the relevant roles of attention and working memory for the performance of VGPs,
several studies have been carried out to investigate the influence of playing games on these cognitive
skills, showing promising results regarding the improvements provided by games (e.g., Bavelier et al.,
2012; Figueiredo & Sbissa, 2013; Green & Bavelier, 2006; Toril et al., 2016), although other studies
also show divergent results, with the absence of effects that improve cognitive skills (Farhy et al.,
https://doi.org/10.17605/OSF.IO/R6JAH
4
In addition, many studies have compared the performance of VGPs with groups formed by
people who are Non VideoGame Players (NVGPs), in which the results have shown better
performance among VGPs in several cognitive skills (Alves et al., 2009; Colzato et al., 2013; Green
& Bavelier, 2006). The differences in performance between VGPs and NVGPs can be partially
explained by the distinction between controlled processes and automatic processes (Lee et al., 2012;
Sarmet & Pilati, 2016), in which VGPs start to automate their attentional processes when exposed to
By way of illustration, in the study by Alves et al. (2009), the performance of VGPs and NVGPs
in sustained attention skills was evaluated using a computerized test, the Continuous Performance Test
(CPT II). The results showed that the JGVs group performed significantly better than the NVGPs group
in 3 of the 5 variables related to CPT II, showing a lower number of omission errors, shorter reaction
time and better consistency in the responses presented. Additionally, the group of NVGPs underwent
a training intervention with games and performed better in the retest. However, opposite results were
verified by Farhy et al. (2018) using the CPT to compare VGPs and NVGPs, showing a greater need
Discussions in meta-analyses have shown limitations in studies that compare VGPs and
NVGPS, in which the different profiles of VGPs are not taken into account, such as the time dedicated
to games, the number of times (sessions) dedicated per week, making comparisons between surveys
according to the profile of each group (Anguera & Gazzaley, 2015; Bediou et al., 2018). These
differences point to different degrees of engagement on the part of the players, making it difficult to
define how much dedication is necessary for training to have significant impacts on the cognition of
VGPs, as pointed out in the studies by Bediou et al. (2018), Latham et al. (2013) and Stanmore et al.
(2017).
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
5
Differences between experienced and novice players, as well as their level of expertise, are important
data to avoid comparison bias between different groups and to better understand the effects of
playing games on cognition (Boot et al., 2011). This information is essential for a better
understanding of the effect of practice in games and its adequate distribution of time to obtain more
limitations of the studies and the contradictions of the findings make it necessary to carry out further
Therefore, this research evaluated the effects of experience with action games on the optimization of
attentional resources and working memory in a dual task context, based on Signal Detection Theory
through a version of the Continuos Performance Test (CPT) , among VGPs with different game
dedication settings. The basic hypothesis of this work is that players who spend more time in action
games will have a greater capacity to optimize their attention and working memory resources when
compared to players with less engagement in a dual task, thus contributing to a better performance.
Method
Participants
The sample consisted of 85 participants aged between 18 and 30 years (M = 21.9, SD = 3.25;
79% male and 21% female). Participants were divided into groups based on their varying degrees of
engagement in action games, defined by the weekly hours (hr) dedicated to gaming (Low = 1 hr to 5
number of times a week (Low = 1 to 3 times a week); Low-Moderate = 4 to 7 times a week; High-
Moderate = 8 to 11 times a week; High = 11 times a week or more) and frequency (Low = A few times
a month; Low-Moderate = A few times a week; High-Moderate = Many times a week; High = Every
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
6
day). All participants reported playing video games for an average of 14 years (SD = 5), until the time
of data collection. The following inclusion criteria were applied for the experiment and sample: having
a minimum routine of 1 to 20 hours a week (or more) maintained for at least 6 months, using a Sony
Playstation (PS) or Microsoft Xbox (XB) console, and being familiar with the joystick commands of
Stimuli
An experiment was carried out with two versions, each with 12 different stimuli, depending
on the console with which the participant was more familiar, Playstation or Xbox. For the Playstation
console, joystick commands were used, which included the symbols: Triangle, Square, Circle, “X,”
“L1,” “L2,” “R1” and “R2,” while on the Xbox console, the symbols were: “Y,” “B,” “X,” “A,”
“LB,” “LT,” “RB” and “RT.” On both consoles, four direction arrows (up, down, right, and left)
were also used. All the symbols were on a dark background and had the colors corresponding to the
VGPs for each console, the stimuli were selected to safeguard the similarity regarding their functions
and/or positions in the Playstation and Xbox consoles, thus making the performance comparison for
the same task. The stimuli were created using the Inkscape 0.92.4 program and presented on the
computer screen using the PsychoPy 3.0 program (Peirce et al., 2019). An example of the stimuli
Design
The research presented a 4 × 3 mixed factorial design, with the variables Groups - VGPs (Low,
Low-Moderate, High-Moderate, and High) and Complexity of the experiment (C1, C2, C3)
manipulated between and within subjects, respectively. The numbers that denote the levels of
complexity, C1, C2, and C3, indicate the amount of non-target stimuli per block.
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
7
The experiment was divided into two phases. In Phase 1, the participant's performance was
evaluated by three dependent variables, aiming to analyze sustained and selective attention. In Phase
2, in addition to the three variables measured in the previous Phase, the correct evocation variable was
included to assess aspects related to working memory. In both Phases, to better assess different aspects
of players' performance, the Conners' Continuous Performance Test II protocol (CPT II, Conners et
al., 2000) was adapted for the present study. Based on the Signal Detection Theory (SDT), this test
presents a series of stimuli, with some considered as target stimuli (signal) and others as non-target
stimuli (noise). The participant's task is to respond to all target stimuli while simultaneously inhibiting
responses to non-target stimuli. From the rates of correct answers, correct rejections, false alarms, and
omissions, the discriminability value (d prime) and the criterion value (c) of the participants were
calculated. (Details about the calculations can be seen in Van der Kellen et al., 2008.) In addition, the
Reaction Time (RT) and the correct evocation rate for the F2 were evaluated. The experiments, their
respective stimuli, and the raw and processed data are available in the Open Science Framework
(https://osf.io/r6jah/).
Procedure
This research project was approved by the Ethics Committee in Research in Human and Social
Sciences of the University of Brasília (CEP / ICH / UnB), according to opinion 988.985. Participants
were invited to respond to a Google Form, in which basic information about the research was exposed,
along with the Free and Informed Consent Form (TCLE). After consenting to participate, the
participant answered questions about their routine with games (hours, times and frequency of time
spent in games, estimated total time from the moment they started playing, as well as the console used),
https://doi.org/10.17605/OSF.IO/R6JAH
8
Data collection was carried out in a laboratory, in rooms with acoustic insulation, thus avoiding
the interference of additional distracting factors. Subsequently, the participants sat facing a computer
(Notebook Acer Aspire 3, with a 15.6-inch screen, Intel Core i3, and 8GB ram), with an average
distance of 57 cm. Each stimulus had a size of 15º of visual angle. The instructions, defined according
to the Complexity of the task in two different phases, were presented on the screen immediately before
The experiment consisted of a computerized version of the Continuous Performace Test (CTP
II), which aimed to assess the participants' sustained/selective attention (Phase 1) and working memory
(Phase 2). In both Phases, each stimulus remained on the screen for 1 s, followed by a gray rest screen,
between stimuli, lasting 500 ms. The 12 stimuli were presented six times in random order, totaling 216
presentations in three Complexities (2×12×6=216). The three Complexities of the task were defined
from the exposure of different proportions between target and non-target stimuli, with progressive
increase of non-target stimuli in each Complexity. Thus, in Complexity 1 (C1), only one of the stimuli
was defined as non-target (example: “X”), which would be repeated 6 times during this Complexity.
In Complexity 2 (C2), this number increases to 2 (example: “🛆” and “⇾”), with 6 repetitions for each
stimulus, totaling 12 presentations of non-target stimuli. Finally, for Complexity 3 (C3), there were 3
different stimuli (example: “⃪” “⍜” and “R1”) with 6 repetitions for each, totaling 18 presentations
of non-targets.
The participant's task involved inhibiting the response to non-target stimuli. Therefore, Phase
1 of the experiment consisted of a task in which the participants had to press the buttons corresponding
to the stimuli presented on the computer screen, except for previously instructed commands, the so-
called non-target stimuli (for example: Do not respond when “X ” or “A” appears). Phase 2 explored
the same task as Phase 1, but participants were additionally asked to perform a task to count the number
of times non-target stimuli appeared on the screen, thus requiring dual processing of information in
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
9
working memory. At the end of each Complexity, participants should enter the number of computed
The progressive increase in the number of repetitions of the non-target stimuli and the counting
task in Phase 2 represents the effect of cognitive overload during the experiment, aiming to evaluate
possible different performances between the types of engagement. The experiment, carried out in a
single session, lasted an average of 30 minutes and, during the execution of the experiment, the
participant was alone in the collection environment. An example of C3 in both Phases can be seen in
Figure 1.
Figure 1
https://doi.org/10.17605/OSF.IO/R6JAH
10
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
11
Note: The two upper boxes represent the equipment and stimuli used in the experiments, while the two lower boxes
Data Analisys
The rates of correct answers, omissions, false alarms, and correct rejections of the
participants were computed. These values were used to calculate the value of the d prime (d') and
criterion (c) of each participant in each of the Complexities and Phases of the experiment.
Calculations were performed using the Microsoft Excel program, which resulted in a database.
Details on the calculations of d' and c can be seen in Van der Kellen et al., 2008. In addition,
correction criteria for values equal to 0 and 1 for d' were corrected by Hautus (1995 ). The database
was exported to the Jamovi software (The Jamovi Project, 2021), where descriptive and inferential
measures were calculated, such as normality tests, analysis of variances, correlations, and linear
regressions. Access to the database, as well as the outputs of all analyzes performed, are available at
(https://osf.io/r6jah/).
Results
The upper panels of Figure 2 show the average breakdowns as a function of complexities in
Phases 1 (left) and 2 (right). Mixed analyzes of variance (ANOVAs) were performed comparing the
were found regarding d′ values in each Complexity of Phase 1, F (2, 162) = 22.5, p < 0.001, η²ₚ =
0.228. The different combinations were compared using post hoc tests with Tukey's correction. The
d' in C1 (M = 2.310, SE = 0.113) was greater than in C2 (M = 1.970, SE = 0.109) and than in C3 (M
= 1.620, SE = 0.079). Furthermore, C2 was also significantly different from C3. All ps ≤ .003.
Differences between Complexities were also verified in Phase 2, F (2, 162) = 199.0, p < 0.001, η²ₚ =
0.711, where the d′ in C3 (M = 1.060, SE = 0.098) was smaller than in C1 (M = 2.700, SE = 0.081)
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
12
and that in C2 (M = 2.560, SE = 0.091), all ps ≤ .001. These results suggest a decrease in the
The middle panels in Figure 2 show the average criteria as a function of complexities in
Phases 1 (left) and 2 (right). A mixed factorial ANOVA indicated significant differences between
Phase 1 Complexities, F (2, 162) = 3.55, p = 0.031, η²ₚ = 0.042. Post hoc tests showed that c in C2
(M = 0.584, SE = 0.010) was higher than in C3 (M = 0.556, SE = 0.009), p = .022, indicating a more
liberal criterion in C3. However, the c in C1 (M = 0.563, SE = 0.010) did not differ from the other
groups, with all ps ≥ .182. In Phase 2, differences between Complexities were also verified, F (2,
162) = 8.960, p < 0.001, η²ₚ = 0.100. Complexity 1 (M = 0.513, SE = 0.006) did not differ from C2
with C3 (M = 0.550, SE = 0.011). Significant differences were also verified between Complexity 2
and 3. All p-values ≤ .015. In summary, these results suggest that in all Complexities, the
participants presented a more conservative criterion, whereas in Phase 1, this criterion became
slightly more liberal between C2 and C3, and in Phase 2 this criterion became more conservative
The bottom panels of Figure 2 show average reaction times as a function of complexities in
Phases 1 (left) and 2 (right). A mixed factorial ANOVA indicated significant differences in reaction
time between Phase 1 Complexities, F (2, 162) = 141.186, p < 0.001, η²ₚ = 0.635. The reaction time
SE = 0.006), ps ≤ .001. C2 and C3 reaction times did not differ, p = .646. In Phase 2, significant
differences were also verified between Complexities F (2, 162) = 207.261, p < 0.001, η²ₚ = 0.719.
The post hoc studies verified that C1 (M = 0.545, SE = 0.007) differed significantly from C2 (M =
0.562, SE = 0.006) and C3 (M = 0.624, SE = 0.007), as well as C2 showed differences from C3. All
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
13
with p-value ≤ .001. These results show the effect of Complexity, especially in Phase 2, where
Figure 2.
Comparisons Between Complexities in Each Phase Based on d prime, c (Criterion) and Reaction
Note: The upper tables represent the participants' performance in terms of d prime (d') in both phases. The intermediate
tables represent the performance regarding the values of the criterion used (c) in both phases. The lower frames represent
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
14
the reaction time (RT) in both phases. C1 = Complexity 1, C2 = Complexity 2 and C3 = Complexity 3. Dedicated Hours
per Week
The main objective of this study was to understand how different types of engagement in
action games can affect the attentional and working memory capacities of video game players. To
identify differences between groups with different configurations of weekly hours, four groups were
defined according to the number of weekly hours dedicated to games: Group 1 = 1 hr to 5 hr per
week (n = 32), Group 2 = 6 hr to 10 hr per week (n = 24), Group 3 = 11 hr to 20 hr per week (n = 19)
Mixed analyzes of variance (ANOVAs) were performed to compare the performance of the
groups and their interaction effect in each of the Complexities and both Phases. From the value of the
estimated marginal means of d' from Phase 1, no significant difference was verified between the
groups, F (3, 81) = 2.58, p = 0.059, η²ₚ = 0.087. However, when looking at the interaction effect
between groups and Complexities, significant differences between groups were identified for each
Phase 1 Complexity, F (6, 162) = 21.6, p < 0.001, η²ₚ = 0.444. The upper panels of Figure 3 show the
means of discriminability as a function of complexities in Phases 1 (left) and 2 (right). The results of
post hoc tests with Turkey's correction showed that Group 1 (M = 0.804, SE = 0.119) was
significantly different from Group 3 (M = 2,670, SE = 0.154) and Group 4 (M = 2,174, SE = 0.212)
in C3, with p-values ≤ .001, but did not differ from Group 2 (M = 0.846, SE = 0.137), with p-value =
1.000. Complementarily, it was verified that Group 2 differed significantly from Groups 3 and 4 in
C3, with p-values ≤ .001. No differences were identified between groups in the other Complexities,
In Phase 2, a significant difference was verified between the groups, F (3, 81) = 5.77, p =
0.001, η²ₚ = 0.176, regarding d prime values. Post hoc tests showed that Group 1 (M = 1810, SE =
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
15
0.124) was significantly different from Group 3 (M = 2290, SE = 0.161), with p-value = .001, and
Group 2 (M = 1.740, SE = 0.144) differed from Group 3, with p-value = .005. Other pairwise
comparisons, as well as comparisons with Group 4 (M = 2.020, SE = 0.222) did not show significant
differences, assuming p-value ≤ .055. As for the interaction effect between groups and Complexity, a
significant difference was also verified, F (6, 162) = 5.77, p = 0.001, η²ₚ = 0.176. The same effect of
Phase 1 was verified for Phase 2, in which Group 1 (M = 0.348, SE = 0.147) was significantly
different from Group 3 (M = 1.973, SE = 0.191) and Group 4 (M = 1.533, SE = 0.263) in Complexity
3, with p-values ≤ 0.009, but it did not differ from Group 2 (M = 0.401, SE = 0.170), with p = 1.000.
Furthermore, it was found that Group 2 was significantly different from Group 3 and 4, with p-values
≤ .024. These results suggest that the difficulty level of C3 was more sensitive to verify differences
between participants who dedicate less than 10 hours a week and those who dedicate 11 hours or
more. These effects can be seen in Figure 3. Raw and processed data from all analyzes performed are
available at (https://osf.io/r6jah/).
Figure 3.
Comparison of the d prime Values of Each Group According to the Number of Hours Dedicated
https://doi.org/10.17605/OSF.IO/R6JAH
16
Note: The upper tables represent the performance of the participants regarding the d prime (d’) in both phases regarding
the number of hours dedicated per week. The middle tables represent the participants' performance on d prime (d') in
both phases in terms of the number of times players play action games per week. The lower tables represent the
performance of the participants regarding the d prime (d') in both phases regarding the weekly and monthly frequency
https://doi.org/10.17605/OSF.IO/R6JAH
17
Concerning c (criterion) values, no significant differences were identified between groups in
Phase 1 (M = 0.568, SE = 0.004), F (3, 81) = 2.00, p = 0.121, η²ₚ = 0.069, or in Phase 2 (M = 0.529,
SE = 0.003), F (3, 81) = 0.267, p = 0.849, η²ₚ = 0.010. In addition, the interaction effect between
groups and Complexities was not identified in Phase 1 (M = 0.568, SE = 0.005), F (6, 162) = 1.93, p
= 0.193, η²ₚ = 0.051, or in Phase 2 (M = 0.529, SE = 0.006), F (6, 162) = 0.304, p < 0.934, η²ₚ =
0.011. These results suggest a conservative criterion that was maintained during the two phases of
the experiment.
Lack of effect was also identified in comparisons between groups regarding reaction time in
Phase 1 (M = 0.560, SE = 0.004), F (3, 81) = 1.96, p = 0.126, η²ₚ = 0.068, or in Phase 2 (M = 0.577,
SE = 0.004), F (3, 81) = 1.40, p = 0.247, η²ₚ = 0.049. This was also observed regarding the
interaction effect between groups and Complexities in Phase 1 (M = 0.560, SE =0.004), F (6, 162) =
0.941, p = 0.468, η²ₚ = 0.034, as well as in Phase 2 (M = 0.577, SE = 0.004), F (6, 162) = 0.173, p =
0.984, η²ₚ = 0.006. This result shows that, despite the effect of Complexity, changes in reaction time
were not identified between groups and their interaction with Task Complexity. Finally, no
significant differences were found between the correct evocation rate between the groups, X² (3) =
3.34, p = 0.342, ε² = 0.039, suggesting that the hours devoted weekly did not interfere with the
additional task of counting in Phase 2. Raw and processed data from all analyzes performed are
available at (https://osf.io/r6jah/).
A second criterion used to verify engagement with games was the number of times played per
week, thus resulting in three groups: Group 1 = 1–3 times per week (n = 28); Group 2 = 4–7 times a
week (n = 34); Group 3 = 8 times or more per week (n = 23). From the mixed analysis of variance
(ANOVAs), no significant differences were identified between the groups, regarding the value of d',
from the marginal means estimated in Phase 1, F (2, 82) = 0.818, p = 0.445, η²ₚ = 0.020. As for
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
18
Phase 2, significant differences were found between groups, F (2, 82) = 0.3.55, p = 0.033, η²ₚ =
0.080. The intermediate panels in Figure 3 show the means of discriminability as a function of
complexities in Phases 1 (left) and 2 (right). From the post hoc comparisons, it was verified that
0.113), with p-value = .047. However, no differences were verified between Group 1 and Group 3
(M = 2190, SD = 0.137), or between Group 2 and Group 3, with analyzes having a p-value ≥ .080.
Comparisons regarding the effect of interaction between groups and Complexities showed
significant differences in Phase 1, F (4, 164) = 4.20, p = 0.003, η²ₚ = 0.009, however, post hoc
analyzes did not show significance between equal complexities versus different groups, assuming p-
values ≥ 0.121. In Phase 2, significant differences were also detected by ANOVA, F (4, 164) = 3.78,
p = 0.006, η²ₚ = 0.084. The post hoc results showed that Group 1 (M = 1790, SE = 0.124) differed
significantly from Group 3 (M = 2190, SE = 0.137) in Complexity 3, assuming p-value = .037. Other
comparisons, including comparisons with Group 2 (M = 2.190, SE = 0.113), had no effect, having p-
values ≥ .072. This result shows that players who dedicate themselves between 1–3 times a week,
obtained better results in d' values, compared to players who dedicate themselves 4 times or more a
About c (criterion) values, no significant differences were identified between groups in Phase
1 (M = 0.567, SE = 0.001), F (2, 82) = 0.517, p = 0.598, η²ₚ = 0.012, or in Phase 2 (M = 0.529, SE =
0.001), F (2, 82) = 0.513, p = 0.601, η²ₚ = 0.012. In addition, the interaction effect between groups
and Complexities was not identified in Phase 1 (M = 0.567, SE = 0.002), F (4, 164) = 0.842, p =
0.501, η²ₚ = 0.020, or in Phase 2 (M = 0.529, SE = 0.003), F (4, 164) = 0.977, p = 0.421, η²ₚ = 0.023.
These results suggest a slight decrease in a conservative criterion that was maintained during the two
https://doi.org/10.17605/OSF.IO/R6JAH
19
Lack of effect was also identified in comparisons between groups regarding reaction time in
Phase 1 (M = 0.560, SE = 0.001), F (2, 82) = 0.159, p = 0.853, η²ₚ = 0.004, or in Phase 2 (M = 0.576,
SE = 0.001), F (2, 82) = 0.202, p = 0.818, η²ₚ = 0.005. This was also observed regarding the
interaction effect between groups and Complexities in Phase 1 (M = 0.560, SE = 0.001), F (4, 164) =
0.403, p = 0.806, η²ₚ = 0.010, as well as in Phase 2 (M = 0.576, SE = 0.001), F (4, 164) = 0.786, p =
0.536, η²ₚ = 0.019. This result shows that, despite the effect of Complexity, changes in reaction time
were not identified between groups and their interaction with Task Complexity. Finally, no
significant differences were found between the correct evocation rate between the groups, X² (2) =
1.66, p = 0.436, ε² = 0.019, suggesting that the hours devoted weekly did not interfere with the
additional task of counting in Phase 2. Raw and processed data from all analyzes performed are
available at (https://osf.io/r6jah/).
Finally, participants were asked about their frequency with action games, thus forming four
groups: Group 1 = A few times a month (n = 16); Group 2 = A Few Times a Week (n = 13); Group 3
= Many Times a Week (n = 32); Group 4 = Every Day (n = 24). Comparisons based on mixed
analysis of variance (ANOVAs) for d' values in Phase 1 did not show significant differences between
groups, F (3, 81) = 1.97, p = 0.125, η²ₚ = 0.068. Regarding the interaction effect, a significant effect
was found, F (6, 162) = 4.87, p < 0.001, η²ₚ = 0.153. The lower panels of Figure 3 show the means of
discriminability as a function of complexities in Phases 1 (left) and 2 (right). Post hoc tests showed
that Group 2 (M = 1573, SE = 0.170) showed significant differences with Group 4 (M = 1874, SE =
0.196) at C3, with p-value = .049. Other pairwise comparisons, including comparisons with Group 1
(M = 0.835, SE = 0.241) and Group 3 (M = 0.757, SE = 0.267), were not significant, assuming p-
values ≥ .056. In Phase 2, no effect was detected concerning groups either, F (3, 81) = 2.11, p =
0.105, η²ₚ = 0.073. Regarding the interaction between group and Complexity, a significant effect was
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
20
identified, F (6, 162) = 2.83, p = 0.012, η²ₚ = 0.095, however, post hoc tests did not identify any
significant effect with equal Complexities and different groups, having these analyzes with p-values
With regard to c (criterion) values, no significant differences were identified between groups
in Phase 1 (M = 0.567, SE = 0.003), F (3, 81) = 0.380, p = 0.767, η²ₚ = 0.014, or in Phase 2 (M =
0.529, SE = 0.003), F (3, 81) = 0.747, p = 0.527, η²ₚ = 0.027. In addition, the interaction effect
between groups and Complexities was not identified in Phase 1(M = 0.567, SE = 0.004), F (6, 162) =
1.94, p = 0.078, η²ₚ = 0.067, or in Phase 2 (M = 0.529, SE = 0.005), F (6, 162) = 0.202, p = 0.976, η²ₚ
= 0.007. Similar to previous analyses, these results suggest the adoption of a conservative criterion
Lack of effect was also identified in comparisons between groups regarding reaction time in
Phase 1 (M = 0.564, SE = 0.003), F (3, 81) = 0.175, p = 0.164, η²ₚ = 0.061, or in Phase 2 (M = 0.579,
SE = 0.003), F (3, 81) = 1.36, p = 0.261, η²ₚ = 0.034. This was also observed regarding the
interaction effect between groups and Complexities in Phase 1 (M = 0.564, SE = 0.003), F (6, 162) =
1.12, p = 0.353, η²ₚ = 0.040, as well as in Phase 2 (M = 0.579, SE = 0.003), F (6, 162) = 1.16, p =
0.329, η²ₚ = 0.003. This result shows that, despite the effect of Complexity, changes in reaction time
were not identified between groups and their interaction with Task Complexity. Finally, no
significant differences were found between the correct evocation rate between the groups, X² (3) =
7.19, p = 0.066, ε² = 0.085, suggesting that the hours devoted weekly did not interfere with the
additional task of counting in Phase 2. Raw and processed data from all analyzes performed are
available at (https://osf.io/r6jah/).
between participants who declared themselves to be Playstation (n = 63) and Xbox (n = 21) players.
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
21
The results showed a significant difference between Playstation (M = 2050, SE = 0.086) and Xbox
(M = 1540, SE = 0.151) players in Phase 1, F (1, 83) = 8.37, p = 0.005, η²ₚ = 0.092 , as well as in
Phase 2, F (1, 83) = 5.20, p = 0.025, η²ₚ = 0.059, for Playstation (M = 2.15, SE =0.0825) and Xbox
(M = 1.77, SE = 0.144) players. In addition, significant differences were also found in the interaction
effects between Playstation and Xbox players as a function of Phase 1 Complexities, F (2, 166) =
4.42, p = 0.013, η²ₚ = 0.051, in which Playstation players (M = 2.58, SE = 0.110) were superior to
Xbox players (M = 1.72, SE = 0.192) at C1. Other comparisons showed no effect, with p-values ≥
.089. No effect was not verified in Phase 2, F (2, 166) = 2.35, p = 0.099, η²ₚ = 0.028. This result
suggests that players who play games on the Playstation platform have better results compared to
Xbox players.
Concerning c values, no differences were found between players in Phase 1, F (1, 83) =
0.901, p = 0.345, η²ₚ = 0.011, but significant differences were found between Playstation players (M
= 0.537, SE = 0.006) and Xbox (M = 0.504, SE = 0.012) in Phase 2, F (1.83) = 5.42, p = 0.022, η²ₚ =
0.061. Interaction effects between player groups and complexities were not verified in Phase 1, F (2,
166) = 1.77, p = 0.174, η²ₚ = 0.021, and in Phase 2, F (2, 166) = 0.297, p = 0.744, η²ₚ = 0.004.
Regarding the estimated marginal means of reaction time, significant differences were
verified between Playstation players (M = 0.552, SE = 0.007) and Xbox players (M = 0.585, SE =
0.001) in Phase 1, F (1, 83) = 5.06, p = 0.027, η²ₚ = 0.057) and in Phase 2, F (1, 83) = 4.06, p =
0.047, η²ₚ = 0.047, in which Playstation players (M = 0.570, SE = 0.007) were faster than Xbox
players (M = 0.599, SE = 0.012). Interaction effects between groups and complexities were also
verified in Phase 1, F (2, 166) = 4.15, p = 0.018, η²ₚ = 0.048, where Playstation players (M = 0.517,
SE = 0.007) were faster than Xbox players (M = 0.548, SE = 0.0013) in C1. In Phase 2, no
interaction effects were verified, F (2, 166) = 0.272, p = 0.762, η²ₚ = 0.003. Finally, there were no
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
22
differences between players regarding the correct evocation rate for Phase 2, X² (1) = 2.88, p = 0.996,
ε² = 3.42. Raw and processed data from all analyzes performed are available at (https://osf.io/r6jah/).
Discussion
As for the present study, the results show a contribution on how different types of engagement
with video games can affect the allocation of attentional resources and working memory in players
with different levels of engagement. These results are important, considering the discussions in the
literature about the differences between VGPs and NVGPs, effects of different training configurations,
related to weekly training time, as well as the long-term effects of cognitive training with games in
videogame players, as previously discussed by studies such as that by Bediou et al. (2018), Latham et
al. (2013), Stanmore et al. (2017). Thus, these results contribute to discussions in the literature, such
as the issues raised by Jaeggi et al. (2011), about different training configurations and their effects
In addition, an important result shows that the hypothesis that the greater the number of hours
dedicated to training, the better the players' performance has been partially proven. As for the number
of hours per week, participants who show high engagement, with dedication above 11 hours a week,
aimed at a better performance compared to participants with moderate and low dedication, with
dedication below 10 hours a week. On the other hand, participants who play more than 4 times a week,
performed worse than participants who play up to 3 times a week. These data are important for future
research and interventions in cognitive training with games, showing that researchers must be aware
of the duration of training and interventions and that training lasting more than 11 hours, up to 3 times
a week, presents good results. evidence of positive effects. These contributions are important in light
of the discussions raised by Anguera and Gazzaley (2015) about the different configurations of training
https://doi.org/10.17605/OSF.IO/R6JAH
23
These results are in line with those found by Kühn et al. (2014) and Feng et al. (2007), which
hours a week. Complementarily, the cognitive results found in our research are important in light of
the findings by Kühn et al. (2014), given that the study did not show the cognitive impacts of
neurophysiological changes.
Furthermore, although this study focuses on VGPs in action games, it is not possible to identify
all categories of games accessed by participants during their trajectory as a player. The transience of a
game category to others in short periods is a great challenge to identify which game benefited more a
cognitive skill than another, not to mention other stimuli that can interfere with the gain of cognitive
skills. This effect reflects one of the main methodological limitations of studies on cognitive training
(Anguera & Gazzaley, 2015; Bediou et al., 2018; Boot et al., 2011; Rivero et al., 2012). Despite this,
this study showed a significant difference between Playstation and Xbox players, suggesting evidence
that Playstation Console games present a better development of these cognitive skills related to
attention.
Limitations regarding the sample size and the effects found are also important to be considered.
It is suggested that future research consider larger samples aiming at a greater effect of comparison
between groups. In addition, it was possible to notice that both C1 and C2, as well as the additional
task of counting and evocation, were not sensitive to the differences between the types of players'
engagement, showing that future research should make a greater balance between target and non-target
Finally, we suggest greater investment in longitudinal studies to verify the effects of training
for long periods among video game players, as well as the application of studies with true
experimental designs, with comparison with control groups. Finally, research that proposes to
https://doi.org/10.17605/OSF.IO/R6JAH
24
understand in a deeper and more systematic way the configuration of engagement that is most
Conclusion
In support of what research shows on cognitive training in video games, in addition to the
results of differences between the performance of video game players compared to non-gamers, this
study showed that different forms of game engagement can have differences in attention and working
memory effects. These results are important for discussions in the literature about cognitive training
with games, their effects and adequate conditions for a better allocation of attentional and memory
resources. New studies must be carried out in order to overcome the methodological limitations
mentioned, as well as research must be carried out in order to investigate the effects of games on long-
term cognition.
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
25
References
Alves, L., Carvalho, A. M., Silveira, J., Belizário, J., Fortini, M., Costa, D., ... & Bambirra, E.
http://www.rmmg.org/artigo/detalhes/483
Anguera, J. A., & Gazzaley, A. (2015). Video games, cognitive exercises, and the enhancement of
https://doi.org/10.1016/j.cobeha.2015.06.002
Baddeley, A. (2012). Working memory: Theories, models, and controversies. Annual Review of
Ballesteros, S., Mayas, J., Prieto, A., Ruiz-Marquez, E., Toril, P., & Reales, J. M. (2017). Effects of
video game training on measures of selective attention and working memory in older adults: results
15. https://doi.org/10.3389/fnagi.2017.00354
Bavelier, D., Achtman, R. L., Mani, M., & Föcker, J. (2012). Neural bases of selective attention in
https://doi.org/10.1016/j.visres.2011.08.007
Bediou, B., Adams, D. M., Mayer, R. E., Tipton, E., Green, C. S., & Bavelier, D. (2018). Meta-
Bejjanki, V. R., Zhang, R., Li, R., Pouget, A., Green, C. S., Lu, Z. L., & Bavelier, D. (2014). Action
video game play facilitates the development of better perceptual templates. Proceedings of the
https://doi.org/10.1073/pnas.141705611
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
26
Bhat, S., Bernhardt, L. K., Thinakaran, A., Ganasan, L., & Rao, P. R. J. (2019). Effect of present day
videogames on adolescent cognitive abilities. Indian Journal of Public Health Research &
Boot, W. R., Blakely, D. P., & Simons, D. J. (2011). Do action video games improve perception and
Colzato, L. S., van den Wildenberg, W. P., Zmigrod, S., & Hommel, B. (2013). Action video gaming
and cognitive control: Playing first person shooter games is associated with improvement in
working memory but not action inhibition. Psychological Research, 77(2), 234–239.
https://doi.org/10.1007/s00426-012-0415-2
Conners, C. K., Staff, M. H. S., Connelly, V., Campbell, S., MacLean, M., & Barnes, J. (2000).
Conners’ continuous performance Test II (CPT II v. 5). Multi-Health Systems Inc, 29, 175–96.
https://www.statisticssolutions.com/free-resources/directory-of-survey-instruments/conners-
continuous-performance-test-ii-cpt-ii/
Farhy, Y., Veríssimo, J., & Clahsen, H. (2018). Universal and particular in morphological
1125–1133. https://doi.org/10.1080/17470218.2017.131091
Feng, J., Spence, I., & Pratt, J. (2007). Playing an action video game reduces gender differences in
9280.2007.01990.x
Figueiredo, O., & Sbissa, P. P. M. (2013). Efeito dos jogos eletrônicos sobre atenção seletiva.
Föcker, J., Mortazavi, M., Khoe, W., Hillyard, S. A., & Bavelier, D. (2019). Neural correlates of
enhanced visual attentional control in action video game players: An event-related potential
https://doi.org/10.1162/jocn_a_01230
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
27
Green, C. S., & Bavelier, D. (2006). Enumeration versus multiple object tracking: The case of action
https://doi.org/10.1016/j.cognition.2005.10.004
Hautus, M. J. (1995). Corrections for extreme proportions and their biasing effects on estimated
values ofd′. Behavior Research Methods, Instruments, & Computers, 27(1), 46-51.
https://link.springer.com/article/10.3758/BF03203619
Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Shah, P. (2011). Short-and long-term benefits of
https://doi.org/10.1073/pnas.1103228108
Kühn, S., Lorenz, R., Banaschewski, T., Barker, G. J., Büchel, C., Conrod, P. J., ... & IMAGEN
Consortium. (2014). Positive association of video game playing with left frontal cortical
Latham, A. J., Patston, L. L., & Tippett, L. J. (2013). The virtual brain: 30 years of video-game play
https://doi.org/10.3389/fpsyg.2013.00629
Lee, H., Voss, M. W., Prakash, R. S., Boot, W. R., Vo, L. T., Basak, C., ... & Kramer, A. F. (2012).
Palaus, M., Viejo-Sobera, R., Redolar-Ripoll, D., & Marrón, E. M. (2020). Cognitive enhancement
via neuromodulation and video games: Synergistic effects? Frontiers in Human Neuroscience,
Peirce, J., Gray, J. R., Simpson, S., MacAskill, M., Höchenberger, R., Sogo, H., ... & Lindeløv, J. K.
(2019). PsychoPy2: Experiments in behavior made easy. Behavior Research Methods, 51(1),
195-203. https://doi.org/10.3758/s13428-018-01193-y
PREPRINT: Effects of action video game engagement on attention and working memory
https://doi.org/10.17605/OSF.IO/R6JAH
28
Reynaldo, C., Christian, R., Hosea, H., & Gunawan, A. A. (2021). Using video games to improve
capabilities in decision making and cognitive skill: A literature review. Procedia Computer
Rivero, T. S., Querino, E. H., & Starling-Alves, I. (2012). Videogame: Seu impacto na atenção,
https://doi.org/10.5579/rnl.2012.109
Sajan, J. E., John, J. A., Grace, P., Sabu, S. S., &Tharion, G. (2017). Wii-based interactive video
367. https://doi.org/10.1080/17518423.2016.1252970
Sala, G., Tatlidil, K. S., & Gobet, F. (2018). Video game training does not enhance cognitive ability:
https://doi.org/10.1037/bul0000139
Sarmet, M. M., & Pilati, R. (2016). Efeito dos jogos digitais no comportamento: análise do General
Stanmore, E., Stubbs, B., Vancampfort, D., de Bruin, E. D., & Firth, J. (2017). The effect of active
https://doi.org/10.1016/j.neubiorev.2017.04.011
The Jamovi Project (2021). jamovi. (Version 2.2) [Computer Software]. Retrieved from
https://www.jamovi.org.
Toril, P., Reales, J. M., Mayas, J., & Ballesteros, S. (2016). Video game training enhances
visuospatial working memory and episodic memory in older adults. Frontiers in Human
https://doi.org/10.17605/OSF.IO/R6JAH
29
Unsworth, N., Redick, T. S., McMillan, B. D., Hambrick, D. Z., Kane, M. J., & Engle, R. W. (2015).
Is playing video games related to cognitive abilities? Psychological Science, 26(6), 759–774.
https://doi.org/10.1177/0956797615570367
Van der Kellen, D., Nunes, L. D., & Garcia-Marques, L. (2008). Sensibilidade e bom senso: