Action and puzzle video games prime different speed ...

Perception, 2009, volume 38, pages 1678 ^ 1687

doi:10.1068/p6324

Action and puzzle video games prime different speed/accuracy tradeoffs Rolf A Nelson, Ian Strachan

Department of Psychology, Wheaton College, Norton, MA 02766, USA; e-mail: [email protected] Received 11 November 2008, in revised form 6 February 2009

Abstract. To understand the way in which video-game play affects subsequent perception and cognitive strategy, two experiments were performed in which participants played either a fastaction game or a puzzle-solving game. Before and after video-game play, participants performed a task in which both speed and accuracy were emphasized. In experiment 1 participants engaged in a location task in which they clicked a mouse on the spot where a target had appeared, and in experiment 2 they were asked to judge which of four shapes was most similar to a target shape. In both experiments, participants were much faster but less accurate after playing the action game, while they were slower but more accurate after playing the puzzle game. Results are discussed in terms of a taxonomy of video games by their cognitive and perceptual demands.

1 Introduction Video games have increased greatly in their sophistication and popularity in recent years. While there has been a great deal of effort focused on the relationship between video-game violence on real-world behavior and development (cf Dill and Dill 1998; Anderson and Bushman 2001), recent studies have shown specific effects that videogame play can have on visual perception and attention. Green and Bavelier (2003), for example, demonstrated that previously inexperienced video-game players can improve their performance on attentional tasks by playing an action video game an hour a day for ten days. A number of studies have shown interesting performance differences between non-video-game players and avid video-game players, including on an attentional-blink and a useful-field-of-view task (Green and Bavelier 2003); enumeration and multiple object tracking (Green and Bavelier 2006a); a variety of spatial tasks (Pepin and Dorval 1988; Subrahmanyam and Greenfield 1994; Green and Bavelier 2006b, 2007; Quaiser-Pohl et al 2006); inhibition-of-return and visual-search tasks (Castel et al 2005); and even lowlevel measures such as contrast sensitivity (Li et al 2008). In order to make a serious effort at understanding the perceptual and cognitive effects of video games, it must be realized that the variety of video games is quite broad. Most studies on the perceptual effects of video games in recent years have utilized a particular genre, that of the fast-action first-person shooter (FPS). However, it is misleading to base conclusions about video games in general on a single genre, just as it would be misleading to base one's conclusions about the effects of television by considering only crime shows. In recent years, video-game taxonomy (see Wolf 2001: Apperly 2006) has coalesced into a number of distinct genres. Although many games may overlap in genre, they are often broken down into categories such as sports, simulation, puzzle, FPS, role-playing, and so forth, with each of these categories including several subgenres. In the study of the effects of video-game play on perceptual and attentional processes, the relevant categorization into genres may be different than for other purposes. For example, the game Halo 3 has a science fiction theme, while the game Medal of Honor 4 has a military theme. If they were films, they might be categorized into war and science-fiction genres. However, the relevant aspect here is that they are both FPS games with a similar style of play öfast reactions, accuracy, and good spatial

Action and puzzle video games

1679

navigation being important for successful game play. On the basis of this taxonomy of skills utilized, we have chosen two games that require very different skills. We reasoned that interaction with a puzzle game, in which deliberation and a considered approach is required for success, would incur different effects than interaction with a fast-action shoot-em-up game, in which these cognitive skills are less relevant. One fairly consistent finding on the effects of video-game play is a generalized improvement on reaction-time (RT) tasks (eg Green and Bavelier 2003; Castel et al 2005; Quaiser-Pohl et al 2006; Dye et al submitted). Since an increase in speed on a task is often accompanied by a reduction in accuracy, we wondered if at least part of this reaction-time improvement might be due to strategic factors; that is, a shift in the speed ^ accuracy tradeoff. This sort of shift could result from a demand in the optimal strategy for a particular game, and thus might yield very different results depending on the game. For example, a fast-action FPS could require the rapid deployment of attention and an immediate motor movement, while a more cognitively oriented puzzle game could require a slow and more deliberate response. In the present experiments we investigated whether video-game play can influence strategic deployment of speed and accuracy in a task which utilizes both. We measured performance on a task that could be influenced by different strategic approachesöparticipants could either be very fast at the task, or very accurate. Speed ^ accuracy tradeoffs have been shown to be influenced most prominently by including incentives for emphasizing one or the other (eg Fitts 1966; Wood and Jennings 1976; Dickman and Meyer 1988), but also by situational variables such as nearness to completion of a task (Fo«rster et al 2003). It was our prediction that playing a video game would cause a generalized shift in strategy reflecting the skills used in that game. Playing a game which requires very fast deployment of visual attention and motor movement could prime a strategy of speed over accuracy, while playing a game which emphasizes a slower, more thoughtful pace could prime the opposite pattern. 2 Experiment 1: Speed and accuracy in a location task 2.1 Method 2.1.1 Participants. Twenty students (eleven male, nine female) from Wheaton College in Norton, MA, were recruited for the study. Ages ranged between 19 and 23 years (mean 21:8 years). They received compensation in the form of a $10 gift certificate to the campus bookstore for their participation. 2.1.2 Procedure. After signing a consent form, participants were asked to fill out a brief survey on the amount of experience they had playing video games. Along with a few other background questions, participants were asked to circle a number on a scale of 1 to 10 how often they played FPS action games such as Quake, Doom, or Halo. Underneath the rating scale were the words ``never'', ``seldom'', ``sometimes'', ``frequently'', and `òften''. Participants were then given a location task. The task was given on a Dell Optiplex computer with a 21-inch ViewSonic CRT monitor at a resolution of 1024 by 768 pixels. Participants sat approximately 24 inches from the screen, giving the visible screen a total size of approximately 19.5 deg of visual angle wide by 14.5 deg of visual angle high. Stimuli were presented via the program SuperLab. All participants received the following instructions: `Ìn this test, you will have a number of trials. On each trial, a plus sign will appear in the middle of the screen. When this happens, you will move the mouse to the plus sign and click it. Next, a square with an ``X'' in it will flash somewhere on the screen. Your task is to click the mouse QUICKLY and ACCURATELY at the center of the spot where the box was. After you have clicked on it, the next trial will begin. To familiarize you with the procedure, there will be five practice trials. Do you have any questions?''

1680

R A Nelson, I Strachan

Ti m

e

Blank screen until response

Randomly-located target (200 ms)

SOA (1500 ms)

Fixation cross (until clicked)

Figure 1. The time course of the trials for experiment 1.

Figure 1 shows the sequence for each trial. Initially, a fixation cross appeared at the center of the screen. When participants clicked on the cross with the mouse, a 1500 ms SOA was followed by the appearance of a rectangular box with an `6' inside it. This target appeared randomly at one of twenty locations on the screen, varying from approximately 2.8 deg to 10.2 deg from the center. The target was present for 200 ms, a time selected so that participants would not be able to click on it while still present. As soon as participants clicked the mouse on a location, the fixation cross appeared at the center of the screen to begin the next trial. There were five blocks of 20 trials presented. Each block contained all 20 locations. Following the pretest, participants were introduced to one of two video games (ie ten participants per game). After being familiarized with the game, they played it for an hour. Performance on these games was recorded in order to ensure that the participants were sufficiently engaged. The action game Unreal Tournament is an FPS game. The pace of the game is very fast, with the goal being to kill as many opponents as possible while avoiding being killed. Each participant went through the game tutorial (and was observed by the experimenter), which lasted approximately 5 min and familiarized them with the controls, before engaging in four 15-min sessions. All participants were started on the easiest level, `novice', If they performed sufficiently well during each session (as measured by killing more enemies than they were killed), they advanced to a more difficult level. Thus, participants could advance from `novice' to àverage' to èxperienced' to `skilled' during the course of the hour if they performed optimally. The puzzle game Portal is superficially similar to an FPS game (1) in that it involves navigating around a complex spatial environment from a first-person perspective; however, the focus of the game-play is on solving puzzles. Rather than using a weapon to kill opponents, players must aim a gun at walls and objects to create a temporary `portal' (1) The

Source game engine for Portal has in fact been used for FPS games, most notably Half-Life 2.


1681

between spatial locations. So, for example, players may shoot the gun at a distant wall to create a hole. They may then shoot the gun at the floor in front of them and create a second hole. If they step through the floor in front of them, they will appear out of the distant wall where they had created the first hole. The physics of the game preserves momentum, so that if a player jumps into a hole far below them, he/she will be propelled with an equivalent speed (but not direction) from the far wall. New players to the game must spend time to understand these novel physics, so cognitive strategy should be less concerned with rapid response than with one that solves the problem at hand. After observing participants during the first level to ensure they understood the controls, the experimenter then began an hour of play. Progress was measured by the number of levels through which they successfully navigated. When participants finished the hour of video-game play, they repeated the spatiallocation task (the post-test), identical to the one they had taken earlier. 2.2 Results 2.2.1 Reaction time. There was a significant difference in RT patterns for the two groups. Figure 2 shows RT scores before and after playing each game. A 262 ANOVA with epoch (pre-test=post-test) as a within-subjects variable, and video game as a between-subjects variable indicated an interaction (F1, 18 18:64, p 5 0:001), such that participants showed a much greater speed increase on the task after playing the action game than the puzzle game. 1100

Pre-test

1000

Post-test

RT=ms

900 800 700 600 500 400

puzzle

Game type

action

Figure 2. Reaction-time (RT) results for experiment 1. Participants were faster after playing an action game than after playing a puzzle game.

Individual comparisons indicate that pre-test RTs did not differ between the two game types (t18 5 1), but post-test RTs were considerably faster for the action-game players than for the puzzle-game players (t18 5:44, p 5 0:0001). In fact, although performance was faster after playing the action game than before for the same participants (t9 3:44, p 5 0:01), performance was actually slower after playing the puzzle game (t9 2:63, p 0:01). 2.2.2 Accuracy. There was also a significant difference in the pattern of accuracy between game conditions. Figure 3 shows accuracy in terms of mean distance from the target (ie greater distance corresponds to more error). While both the action- and puzzlegame players performed similarly in the pre-test (t18 5 1), puzzle-game players were significantly more accurate than action players in the post-test (t18 3:10, p 5 0:01). This resulted in a significant interaction between game type and the pre-test/post-test conditions (F1, 18 10:69, p 5 0:01). Paired comparisons indicate that participants who played the action game became less accurate (t9 ÿ2:90, p 5 0:01), and participants who played the puzzle game became more accurate (t9 1:82, p 0:05).

1682


Distance from target=pixels

50 Pre-test

45 40

Post-test

35 30 25 20 15 10 5 0

puzzle

action

Game type

Figure 3. Accuracy results for experiment 1. Action gamers were less accurate than puzzle gamers.

Figure 4 shows errors in more detail, offering a visualization of the ways in which participants strayed from the target. Each scatter dot represents a single response on an X and Y axis relative to the target position. A positive point reflects a response farther from the fixation than the target, while a negative point reflects a response closer to the fixation than the target. In this figure, one can see the more scattered and less precise responses that came after playing the action video game. As indicated by linear regression trend lines in the figures, overall error patterns show that participants tended to err along the linear trajectory running from the fixation to the target.

Action

ÿ250

ÿ150

250

250

200

200

150

150

100

100

50

50

0

ÿ50ÿ50 ÿ100 ÿ150 ÿ200 ÿ250

50

150

250

ÿ250

ÿ150

ÿ250

ÿ150

50 0 ÿ50ÿ50 ÿ100 ÿ150 ÿ200 ÿ250

Pre-test

50

150

250

50

150

250

250 200 150 100 50

250 200 150 100

Puzzle

0 ÿ50ÿ50 ÿ100 ÿ150 ÿ200 ÿ250

50

150

250

ÿ250

ÿ150

0 ÿ50ÿ50 ÿ100 ÿ150 ÿ200 ÿ250

Post-test

Figure 4. Plot of location errors in experiment 1. Action gamers were less accurate in the posttest than in the pre-test, and as compared to puzzle gamers.

2.2.3 Video-game expertise and performance. On a scale of 1 to 10, participants averaged a score of 3.55 on the survey of experience playing action games, with individual scores ranging all the way from 1 to 10. There was no difference between the averages of the action group (mean 3:6) and the puzzle group (mean 3:5) (t 5 1), ruling out video-game experience as a factor in the results described above.


1683

All participants were able to understand the controls and successfully play the games. For the action game, all participants finished at least 1 level (average 3.1 levels finished). For the puzzle game, participants finished between 8 and 15 levels (average 11.4 levels). 2.3 Discussion Results convincingly demonstrate a priming effect for two different types of video games. Playing an action video game resulted in faster reaction times and lower accuracy on a location task, while playing a puzzle game resulted in slower reaction times and higher accuracy. It is interesting that participants tended to overshoot the target, especially in the post-test for the action-game condition. It might be expected that the fastest RTs (which occurred in this condition) would result from undershooting the target. That is, since the time to move the mouse outward from the center increases the farther the response is, it might be that participants are increasing speed by clicking closer along the trajectory to the target. However, this is not the case. It would appear that they are amplifying the same type of error that participants in all conditions are making, just at a faster speed. How generalizable are these strategic shifts in tasks after video-game play? Experiment 1 demonstrated speed ^ accuracy priming in a task that was closely related to the types of movement used in the video game. While playing the action game, participants moved the mouse around the screen as rapidly and as accurately as possible to target enemies before shooting at them. Likewise, in the location task, they moved the mouse to a cued location as rapidly as possible before clicking on it. In order to understand the nature of the strategy shift that was observed in experiment 1, we wanted to measure performance on a task which did not employ the same motor-movements or location-accuracy skills as the video game itself did. The task in experiment 2 was quite different, while still requiring both speed and accuracy. Rather than mimicking the movements made in the video game, this task required participants to make judgments about the similarities of shapes. We reasoned that if the strategy shift we saw in experiment 1 was at a more cognitive (rather than early visual or motor processing) level, we could reasonably expect a similar speed ^ accuracy shift in a task that does not require the same type of spatial attention and motor movements. 3 Experiment 2: A matching-figures task The task was loosely based on the Kagan's Matching Familiar Figures Test (Kagan 1965), and is also similar to the one used by Dickman and Meyer (1988) in measuring speed ^ accuracy tradeoffs. In this task, a sample shape is shown along with comparison shapes, and response time and error rate are measured in choosing the correct match. 3.1 Method 3.1.1 Participants. Twenty college students (twelve male, eight female) from the same population as in experiment 1 participated, and no participants were in both experiments. Ages ranged between 19 and 22 years (mean 20:7 years). They received the same compensation as those in experiment 1. 3.1.2 Procedure. General procedures were the same as in experiment 1 except for the task. The instructions given for the matching-figures task were: `Ìn this test, you will be shown shapes that look like this.'' [A set similar to figure 5 was shown on the screen.] ``Your task is to find which of the four shapes matches the top one. To do this you will press the corresponding button on the button box.'' [The button box is shown.] ``These may seem difficult, but it is important that you respond as QUICKLY and as ACCURATELY as possible. If your response is longer than 5 seconds, you will get a prompt and the trial will be repeated. Before the actual trials begin, you will be given a

1684


few practice trials.'' [Practice trials are given.] ``Now the actual trials will begin. If you have any questions, or are unsure about what to do, please ask the experimenter now. Remember to respond as QUICKLY and as ACCURATELY as you can.''

Figure 5 demonstrates the type of figures that were used in the task. There were a total of 20 different such sets, and for each set there were 4 different trials consisting of the similar shape in each of the 4 positions. Thus, in all, there were 80 different trials, and they were presented randomly. The stimulus set was designed by changing just one vertex of a shape. The differences in the shapes were intentionally very slight to increase the difficulty of the task and to decrease the corresponding accuracy. Responses were collected via a Cedrus RB-834 USB button box using the four central large buttons.

Figure 5. A sample stimulus for experiment 2. Participants' task was to decide which of the four shapes matched the one on top.

3.2 Results 3.2.1 Reaction time. 68 trials, or about 2% of the total, were discarded because participants took over 5 s to respond. Results, shown in figure 6, were quite similar to those of experiment 1. Again, a 262 ANOVA with epoch (pre-test/post-test) as a within-subjects variable, and video game as a between-subjects variable indicated an interaction (F1, 18 19:46, p 5 0:001), such that participants showed a much greater speed increase on the task after playing the action game than the puzzle game. Participants were much faster after playing the action game than before (t9 3:31, p 5 0:01), and slower after playing the puzzle game than before (t9 3:08, p 5 0:01). 3000

Pre-test

2500

Post-test

RT=ms

2000 1500 1000 500 0

action

Game type

puzzle

Figure 6. Reaction-time (RT) results for experiment 2. Action gamers showed a speed increase, while puzzle gamers showed a speed decrease.


1685

3.2.2 Accuracy. There was again an interaction such that participants were more accurate after the puzzle game than the action game (F1, 18 23:85, p 5 0:001). Individual comparisons (see figure 7) show that accuracy was worse after playing the action game than before (t9 2:77, p 0:01), and that their accuracy was better after playing the puzzle game than before (t9 5:07, p 5 0:0001). 3.2.3 Video-game expertise and performance. Participants averaged a score of 4.75 on the survey of experience playing action games, with individual scores ranging from 1 to 10. There was no difference between the averages of the action and the puzzle group (t18 5 1). In the action condition, all participants finished at least one level (average 2:7 levels finished). For the puzzle game, participants finished between 6 and 18 levels (average 11:7 levels). 90 Accuracy=% correct

80

Pre-test

70

Post-test

60 50 40 30 20 10 0

action

Game type

puzzle

Figure 7. Accuracy results for experiment 2. Action gamers became less accurate, while puzzle gamers became more accurate.

3.3 Discussion In addition to replicating and verifying the results of experiment 1, experiment 2 provided clear evidence that this priming effect can not be attributed solely to a change in motor skills or to task-specific perceptual learning. The matching-figures task was very different in terms of the skills needed and motor-movements involved than the location task used in experiment 1. One objection that might be made is that in experiment 2 (unlike in experiment 1), the pre-test scores from the two types of games were different. RTs for the puzzle gamers on the pre-test were in fact faster than for the action gamers (t18 3:30, p 5 0:01). This could not be a result of any experimental manipulation, since the pretest was given to all participants in an equal manner at the beginning of the experiment. We attribute this difference to a greater range of flexibility in strategy for this task. Note that RTs are overall much higher than for experiment 1 (roughly two to three times higher), and there is a greater variability for individual average response times. For experiment 1, the standard deviation of responses between subjects for the pretest was 219 ms, while in experiment 2, this standard deviation was 741 ms. Clearly there was more individual variation in the initial strategy employed. However, it is important to keep in mind that this was a within-subjects design that was primarily interested in the way that participants' individual strategies shifted öand results conclusively demonstrated that those who played the action game showed faster RTs and less accuracy on the post-test, while those who played the puzzle game showed the reverse. It could be argued that the different pre-test scores indicated different strategies, which in turn caused a different response to the games. Under this scenario, pre-test RTs (and not game) should be the best predictor of post-test RTs. To address

1686


this possibility, a further analysis was done correlating pre-test and post-test RTs. Although there was a significant correlation for both the puzzle game (r 2 0:44, p 5 0:01) and the action game (r 2 0:29, p 5 0:01) for pre-test and post-test RTs, there was no correlation between overall pre-test and post-test RTs (r 2 0:04, p 4 0:05), strongly indicating that pre-existing strategies (eg differences in pre-test scores) cannot sufficiently explain the results. 4 General discussion It is clear from the present results that video games can prime certain strategic shifts in subsequent tasks which require both speed and accuracy. In fact, it is striking how dramatically these strategies can be shifted by a single hour of video-game play. These results underscore the importance of studying the cognitive and perceptual consequences of video games in terms of the types of skills demanded from the particular video game under study. It is clear that generalized statements about how video games affect cognition are misleading; different genres affect perception and strategy in very different ways. An action video game and a puzzle video game have very different demands, and no doubt there are other demands of video games which fall into other categories. This leads to an important issue: creating a useful taxonomy of video games according to the visuo-spatial skills they employ. In the present study, the games were chosen because the demand on speed and accuracy is very different öin Unreal, one must monitor for moving targets and respond quickly to them, while in Portal, one must survey the situation before acting; a fast response is not a demand characteristic.(2) Other video games could have differing demands on speed and/or accuracy, and given that many video games have a story arc that incorporates multiple types of gameplay, might have different demands at different parts of the game. Besides (1) speed and (2) accuracy, we propose that other visuo-spatial skills defining a useful taxonomy might be (3) target discrimination, (4) multiple object tracking, (5) visual search, (6) divided attention, (7) change detection, and (8) spatial navigation in 2-D or 3-D environments. This is not intended to be an exhaustive list, but rather to enumerate some visuo-spatial skills utilized in video games that are well-studied in human visual perception.(3) An issue which may be interesting to speculate on, and one left to further research, involves the degree to which the type of strategy effect demonstrated here manifests itself in everyday life. It might be wondered how video-game play affects the way in which students interact in the classroom öperhaps students who have just played an action video game before class are less patient but faster in their interactions, while those who have played a puzzle game show a slower but more accurate mode of interaction. It also might be wondered how video games affect those in the workplace, especially those who spend a great deal of time on computers. Is there an effect such that playing puzzle games contributes a slower mode of working, while playing action games contributes to a speedier mode? One must, however, be cautious in overextending present results. Since the time course in the present experiments was shortöthe priming effect was demonstrated immediately after the video game was playedöit is not known if these are persistent effects. It is a topic for further investigation whether these effects might persist over the course of hours, days, or weeks. It may be that these effects accumulate with increased video-game play, such that long-term action gamers favor speed over (2) In

terms of visuo-spatial processing, Unreal requires a strategic adjustment between speed and accuracy, while Portal, for the most part, does not. That is, the time window for a response in Unreal (shooting at a randomly moving target) is limited, while a response in Portal (shooting at a static target, or at a moving target with a predictable trajectory) is less constrained. (3) Naturally, most games will employ many of these skills to one degree or another; boundaries between games in these categories would be graded.


1687

accuracy,(4) and the reverse for long-term puzzle gamers. In order to make a convincing causal link between long-term video-game play and resulting cognitive strategies, a study with training on these types of games over days or weeks would be necessary. Nevertheless, present results are conclusive in demonstrating that video games can cause a shift in cognitive strategy outside the game. Given the cultural ubiquity of video games, it is surprising that there is so little research on the cognitive consequences of video-game play. We hope that this will change as it becomes clear how demands of video games may transfer to strategic approaches. Acknowledgments. This work was supported in part by a summer fellowship from the Mars Foundation, and by the Wheaton Research Partnership. References Anderson C A, Bushman B J, 2001 `Èffects of violent video games on aggressive behavior, aggressive cognition, aggressive affect, physiological arousal, and prosocial behavior: A metaanalytic review of the scientific literature'' Psychological Science 12 353 ^ 359 Apperly T, 2006 ``Genre and game studies. Toward a critical approach to video game genres'' Simulation and Gaming 37 6 ^ 23 Castel A D, Pratt J, Drummond E, 2005 ``The effects of action video game experience on the time course of inhibition of return and the efficiency of visual search'' Acta Psychologica 119 217 ^ 230 Dickman S, Meyer D, 1988 `Ìmpulsivity and speed ^ accuracy tradeoffs in information processing'' Journal of Personality and Social Psychology 54 274 ^ 290 Dill K E, Dill J C, 1998 ``Video game violence: A review of the empirical literature'' Aggression and Violent Behavior 3 407 ^ 428 Dye M, Green C, Bavelier D, submitted ``Faster reaction times across domains: transfers of learning with action video games'' Current Directions in Psychological Science Fitts P, 1966 ``Cognitive aspects of information processing: III. Sets for speed versus accuracy'' Journal of Experimental Psychology 71 849 ^ 857 Fo«rster J, Higgins E, Bianco A, 2003 ``Speed/accuracy decisions in task performance: Built in trade-off or separate strategic concerns?'' Organizational Behavior and Human Decision Processes 90 148 ^ 164 Green C S, Bavelier D, 2003 `Àction video game modifies visual selective attention'' Nature 423 534 ^ 537 Green C S, Bavelier D, 2006a `Ènumeration versus multiple object tracking: The case of action video game players'' Cognition 101 217 ^ 245 Green C S, Bavelier D, 2006b `Èffect of action video games on the spatial distribution of visuospatial attention'' Journal of Experimental Psychology: Human Perception and Performance 32 1465 ^ 1478 Green C S, Bavelier D, 2007 `Àction-video-game experience alters the spatial resolution of vision'' Psychological Science 18 88 ^ 94 Kagan J, 1965 Matching Familiar Figures Test (Cambridge, MA: Harvard University) Li R, Polat U, Makous W, Bavelier D, 2008 `Àction video game playing alters early visual processing'' poster presented at the annual meeting of the Vision Sciences Society Pepin M, Dorval M, 1988 `Èffects of practicing video games on two measures of spatial visualization capability'' Revue Quebecoise de Psychologie 9 34 ^ 43 Quaiser-Pohl C, Geiser C, Lehmann W, 2006 ``The relationship between computer-game preference, gender, and mental-rotation ability'' Personality and Individual Differences 40 609 ^ 619 Subrahmanyam K, Greenfield P M, 1994 `Èffect of video game practice on spatial skills in girls and boys'' Journal of Applied Developmental Psychology 15 13 ^ 32 Wolf M, 2001 ``Genre and the video game'', in The Medium of the Video Game Ed. M Wolf (Austin, TX: University of Texas Press) chapter 6 Wood C, Jennings R, 1976 ``Speed ^ accuracy tradeoff functions in choice reaction time: Experimental designs and computational procedures'' Perception & Psychophysics 19 92 ^ 101

(4) Since

the skills required in action games included both speed and accuracy, and because video-game play has been shown to improve the kinds of skills involved in playing the game, it seems reasonable to suggest that more-practiced video-game players would not simply have different speed/accuracy tradeoffs than less-practiced players, but that their own tradeoffs may shift with experience.

ß 2009 a Pion publication

ISSN 0301-0066 (print)

ISSN 1468-4233 (electronic)

www.perceptionweb.com

Conditions of use. This article may be downloaded from the Perception website for personal research by members of subscribing organisations. Authors are entitled to distribute their own article (in printed form or by e-mail) to up to 50 people. This PDF may not be placed on any website (or other online distribution system) without permission of the publisher.