The link between playing action video games and enhanced cognitive and visual-perceptual abilities is well documented. However, the definition of action video games and why playing them leads to improved abilities has not been resolved.
Most action video game studies have used first person shooter games (but see 
). These video games contain several common properties: unpredictability, intense speed, high perceptual, cognitive and motor load, the selection between multiple action plans and an emphasis on peripheral processing 
. Action video game related perceptual enhancements have been reported in the following areas: improved peripheral vision 
, reduced crowding effect 
, a shorter period of backward masking 
and improved contrast sensitivity 
. Action video game play has also been linked to enhanced cognitive abilities such as attending to multiple objects simultaneously 
, superior spatial skills 
as well as reduced attentional blink effects 
. Furthermore, these enhancements have also been seen in higher-order executive control functions such as task switching, working memory, inhibition and reduced attentional capture 
Although there is a large amount of evidence of the benefits of action video games, not all cognitive abilities that have been measured were improved by action video game play. For instance, no differences have been found between action video game players and non-gamers in the ability to inhibit attention from returning to previously attended locations 
, a phenomenon known as the inhibition of return 
. Furthermore, the ability to orient to an exogenous cue has generally been found to be equivalent between action and non-action video gamers 
Despite their potential benefits, action video games may not be suitable for everyone. Many action video games contain violent themes, making their suitability for young children questionable 
. Hence, there is a practical need to investigate whether non-action games benefit cognitive and perceptual abilities.
Currently, little is known about the cognitive benefits of non-action video games. Two studies have demonstrated evidence of improved mental rotation skills in young children trained in Tetris 
in addition to enhanced executive control abilities in older adults after real time strategy game training 
. Nevertheless, these non-action video game training results are not directly comparable to studies using action video games due to differences in sample selection. While most action video game studies used young adult samples, these two studies comprised of young children 
and elderly adults 
. Hence, it remains unclear if young adults trained using these games will show similar cognitive improvements.
To date, there has been a paucity of studies that compared transfer effects of different genres of games within the same study. Such a comparison of multiple genres is important because it is possible that different game types will improve different aspects of cognition and perception.
In most longitudinal studies involving action video games, researchers have compared a control group playing a non-action video game (e.g., Tetris) to a treatment group playing an action video game 
. These studies have generally shown more favorable transfer effects after playing action video games compared to a control non-action video game. While the benefits of action video games are not in question, the methods that have been used in these previous studies may have favored action games. Specifically, many action video game training studies have utilized transfer tasks that mimic the demands of action video games, thus possibly maximizing transfer effects. These transfer tasks are typically fast paced, with multiple objects to track and require quick switches of attention from one target to the next. One commonly used task, the attentional blink task requires fast recovery from detecting one target (T1) in preparation for detecting a second target (T2). Cognitive abilities required in this task are highly practiced in an action video game, where failure to attend to a second target (e.g., enemies appearing in succession) may result in losing points or even the game. Similarly, the cognitive abilities required in a multiple object tracking transfer task are also frequently practiced in action video games where there are high demands to attend and respond to numerous items simultaneously to progress in the game.
Group compositions in cross-sectional studies of action video games may also have unfairly favored experienced action video game players. Specifically, in cross-sectional studies 
, participants who reported casual playing of various genres of non-action video games have been grouped together with non-gamers in a single control group when compared with experienced action video gamers. Drawing conclusions on the superiority of action video games based on such selection criteria is problematic because the effects of playing non-action games may have been masked by the inclusion of non-gamers. Grouping game players of different genres within the same control group poses another problem. Since playing different genres of non-action video games may train different abilities, it is possible that the effects of different non-action games may cancel each other out. One way around this problem would be to directly compare performances between experienced players of specific game genres in a single study, although we suspect this would be operationally difficult because gamers tend not to play a single genre of game exclusively.
We sought to extend previous findings of cognitive improvements seen in action video games by including a wider range of video game genres. That is, in addition to an action game, we purposefully chose games that differ in cognitive demands like visual matching and search (e.g., match-3 and hidden-object game) as well as games that specifically train spatial working memory (e.g., memory matrix). We also included an agent-based life simulation game (The Sims) that did not initially appear to contain any of the aforementioned cognitive demands as a comparison.
Consistent with the claim that transfer of learning depends on the similarities between the learning and transfer task (theory of identical elements) 
, we propose that near transfer will be more likely to occur in behavioral tasks that share similar demands with the trained video games. Hence, frequent training of skills within the game itself should lead to improvements when the same skills are used in the behavioral tasks. Note that this proposal contrasts with far-transfer proposals that suggest training on one task will improve performance on multiple tasks. Far transfer claims include enhanced attentional control and speeded learning of new tasks after action video game training 
. The near and far transfer proposals make different predictions. The former predicts that training on one cognitive task will only lead to improvements in a similar cognitive task, but the latter proposal predicts improvements in several or all tasks.
To test for these transfer effects, we included a wide range of cognitive paradigms whose demands closely mimic the demands of the video game training. This allowed us to determine the limits of how cognitive abilities can be affected by video game playing. Thus, we included measures that have previously shown improvement after action video game training (i.e., cognitive control, apprehension of multiple objects and attentional blink) as well as two tasks that contain features similar to both action and non-action video games used in the current study, but have not been tested in relation to action video game training (i.e., spatial working memory and visual search).
When a visual search and spatial working memory task were administered concurrently, performance in both declined compared to performance on each task separately 
. Hence, the inclusion of this dual task allows us to determine whether training in video games that place demands on either search skills or spatial working memory separately reduces interference between the two tasks. Finally, we included a task to assess verbal working memory span to determine if spatial working memory training can lead to far transfer effects from a visual to verbal modality.
The apparent similarity between the demands in the training games and the behavioral tasks used as pre- and post-measures allowed us to make specific predictions regarding transfer effects. First, the action game makes strong demands on flexible, rapid and multiple target detection but at the same time requires filtering out of distracting stimuli. Training should therefore transfer to attentional blink, multiple object tracking and cognitive control respectively.
In contrast to the action game where enemies are salient and require selective attention, the match-3 game demands tracking the location of multiple static items across a large area, effortful and deliberate searching for items that can be moved to make matches, and strategic planning to produce cascades of matches. To find matches requires conjunctive search of both the item color and the pattern in which it is arranged. Although at the earliest levels, the number of possible matches is plentiful so a match can be found by searching a small area of the game display, as the levels progress, the number of possible matches decreases, requiring the player to examine a larger area to find a match. There is little demand to filter out simple visual information such as color alone since matches may be made using any color at any location of the screen. Any filtering of information would be at a higher level of pattern analysis. There is a demand to keep track of multiple, separate colored groups of objects so that possible areas where matches can occur are remembered. Thus, playing match-3 should lead to improvements in visual search and the tracking of tracking of multiple static objects.
Third, the hidden-object game places demands upon visual search and thus is expected to lead to improvement in search, allowing more resources to be used for a dual task of visual search and spatial working memory. However, it does not demand fast tracking of multiple objects and thus should not improve attentional blink, cognitive control, and multiple-object tracking.
Fourth, the demands in memory matrix game are specific since only spatial working memory is trained. Although there are some memory demands in action video games to keep track of multiple items, they contain many other demands. It is thus possible that transfer will be more effective in the memory-matrix game since it has no competing demands and therefore offers the most training in spatial memory compared to other games. In addition, assuming far transfer, we predict the memory matrix game is most likely to lead to improvements in verbal working memory. This assumes that the memory matrix task trains general memory processes that are not modality specific.
Finally, for the agent-based life simulation game, we do not expect improvements in any behavioral tasks in this study since it does not appear to have similar demands to the behavioral tasks. Although it does have some memory demands to keep track of the goals of each agent, there is no incentive or need to do so as goals and objectives are available on demand in the game menu.