The Effects of Video Game Experience and Active Stereoscopy on Performance in Combat Identification Tasks

Abstract

Objective

We investigated the effects of active stereoscopic simulation-based training and individual differences in video game experience on multiple indices of combat identification (CID) performance.

Background

Fratricide is a major problem in combat operations involving military vehicles. In this research, we aimed to evaluate the effects of training on CID performance in order to reduce fratricide errors.

Method

Individuals were trained on 12 combat vehicles in a simulation, which were presented via either a non-stereoscopic or active stereoscopic display using NVIDIA’s GeForce shutter glass technology. Self-report was used to assess video game experience, leading to four between-subjects groups: high video game experience with stereoscopy, low video game experience with stereoscopy, high video game experience without stereoscopy, and low video game experience without stereoscopy. We then tested participants on their memory of each vehicle’s alliance and name across multiple measures, including photographs and videos.

Results

There was a main effect for both video game experience and stereoscopy across many of the dependent measures. Further, we found interactions between video game experience and stereoscopic training, such that those individuals with high video game experience in the non-stereoscopic group had the highest performance outcomes in the sample on multiple dependent measures.

Conclusion

This study suggests that individual differences in video game experience may be predictive of enhanced performance in CID tasks.

Application

Selection based on video game experience in CID tasks may be a useful strategy for future military training. Future research should investigate the generalizability of these effects, such as identification through unmanned vehicle sensors.

Keywords

combat identification decision making active stereoscopic displays video game experience learning and memory training technologies

Introduction

Combat Identification of Military Vehicles and Fratricide

Combat identification (CID) of military vehicles is a challenging task faced by soldiers, manned vehicle operators, and unmanned vehicle operators (Keebler, Jentsch, & Hudson, 2011). CID has been defined as the capability to rapidly and accurately identify potential targets (U.S. Department of Defense, 1996) and has been established as a key feature in missions using teleoperation (“Ground Combat Systems,” 2010). One major issue associated with CID of military vehicles is fratricide, also referred to as “blue on blue” (Regan, 1995). Fratricide is basically the act of committing a false alarm in battle by firing at one’s allies. Due to the highly similar nature of military vehicles (O’Kane, Biederman, Cooper, & Nystrom, 1997) and the fact that soldiers often make quick decisions during battle, there are many fratricide mistakes made in the identification of combat vehicles (Briggs & Goldberg, 1995; Regan, 1995). In recent conflicts, the estimates for combat vehicles lost due to fratricide have ranged from 30% to 77% (McCarthy, 2003; Reagan, 1995). Therefore, it is pertinent to develop ways to enhance the warfighter’s ability to quickly and accurately identify combat vehicles, both friendly and enemy (Keebler et al., 2011).

Although modern technology, such as blue force tracking (i.e., real-time satellite tracking of friendly “blue force” vehicles and units), can aid in CID tasks, it has been found that the introduction of lag to these systems severely reduces their efficacy (Bryant & Smith, 2013). Therefore, it is simply not enough to solve this problem through technology; it must be solved through the science of training (Salas & Cannon-Bowers, 2001) so that warfighters are able to make accurate decisions without technological aids. For example, Biederman and Shiffrar (1987) found that training using critical cues of an object (i.e., distinguishing features unique to that object) can lead novices to outperform experts on difficult perceptual discrimination tasks (i.e., chicken sexing). Further, they applied their findings to the identification of military vehicles, specifically, main battle tanks (Biederman & Shiffrar, 1987). Others have also found that training using critical features of vehicles leads to the highest performance outcomes in CID tasks (Bramley, 1973).

In this study, we aimed to better understand the effects of individual differences in video game experience and active stereoscopic displays on training CID of military vehicles. Specifically, we hope to provide evidence that video game experience may aid in the selection of individuals who can rapidly learn to competently identify military vehicles (McKinley, McIntire, & Funke, 2011) while also evaluating the use of stereoscopic displays for system design. Next, we review a proposed technology for training, namely, an active stereoscopic simulation-based training. Following, we will review the potential effects of video game experience in the context of CID, specifically through the mechanisms of object recognition.

Active Stereoscopic Displays

Authors of previous research have examined potential technologies that can aid in training CID tasks. Specifically, a multitude of training materials have been examined for their effectiveness at enhancing individuals’ abilities to identify combat vehicles. These materials have included military-issued graphic training aids (U.S. Army Training Support Centers, 1996), 1:35-scale die-cast models, and simulations (Keebler et al., 2007, 2011). In this paper, we aim to further examine the effectiveness of one technology reviewed by Keebler and colleagues (2011), namely, an active stereoscopic simulation-based training.

Active stereoscopy consists of shutter glass technology that is synchronized with on-screen images to induce stereoscopic depth (Turner & Hellbaum, 1986). Active stereoscopy has been found to enhance performance in teleoperative military simulations in which participants must identify improvised explosive devices (Chen, Oden, Kenny, & Merritt, 2010). Further, Chen, Haas, and Barnes (2007) found that stereoscopy enhanced perception of negative terrain and mobility obstacles, yet they noted that stereoscopic displays have potential perceptual and physical limitations. Further studies have shown that stereoscopic information can reduce errors in visual identification (Edelman & Bulthoff, 1992), and others have shown it can enhance the ability for students to understand complex object motion (i.e., plate tectonics; Kim, 2006). Nakayama, Shimojo, and Silverman (1989) argued that stereoscopic information could aid in the combining of partially hidden images—a relevant notion for identification of camouflaged objects. Related to this finding, Bishop and Pettigrew (1986) have argued that one major role of stereoscopic information is to see through camouflage, noting its specific presence in predatory animals. Contrary to the positive findings of stereoscopic systems, some have argued that stereoscopic systems can lead to a decoupling between depth cues (i.e., accommodation and convergence) due to the actual depth of objects being the surface of the display, which has been theorized to cause interference in processing of visual information (Bowman, Kruijff, LaViola, & Poupyrev, 2004).

Although there is much research involving the use of stereoscopic displays, there have been no studies on the effects of stereoscopy in relation to CID tasks, although some have argued it could be beneficial in these types of tasks (Keebler et al., 2011). Therefore, in this study, we aim to understand the effects of active stereoscopy in a simulation-based training. Next we will discuss the theoretical underpinnings of video game experience in relation to object perception and their theoretical impact on CID of military vehicles.

Video Game Experience

Since the widespread emergence of video games began over 30 years ago, the impact of video game experience has become confounded with long-term exposure to video games in some individuals. Consequently, researchers have examined the effects of video games through two lenses: as a training intervention and as an individual difference. We should note that differences observed across individuals are due to experience in the same way that video games as training lead to changes through experience. The difference is the duration of exposure. As an individual difference, researchers have asked whether people who habitually play video games are unique in their perceptual skills. As training, researchers have asked how short-term video game play affects subsequent task performance. Therefore, we will examine research along each of these lines and the application of each to object perception.

Video game experience as an individual difference

A continued research challenge is assessing the effects of longer-term video game exposure as well as transfer of video game experience to tasks outside the laboratory. Examining the individual differences between those who play video games extensively and those who do not can inform the real-world risks and benefits of video games while providing guidance for selection and training. At the same time, correlational findings must be examined cautiously, as people with individual differences, such as spatial ability, may prefer video games, leading them to be labeled video game players in subsequent research (for a review of these limitations, see Boot, Blakely, & Simons, 2011).

Video game experience has a long list of correlates, including route planning and execution (Guildea & Bailey, 1994; Rehfeld, Jentsch, Curtis, & Fincannon, 2005), spatial ability (Schuster, Fincannon, Jentsch, Keebler, & Evans, 2008), hand-eye coordination (Gagnon, 1985; Griffith, Voloschin, Gibb, & Bailey, 1983), laparoscopic surgery performance (Rosser et al., 2007; Schlickum, Hedman, Enochsson, Kjellin, & Fellander-Tsai, 2009), and age and gender (Powers, Brooks, Aldrich, Palladino, & Alfieri, 2013). Here, we focus on the correlates that may explain differences in object recognition performance.

A number of differences in perception have been observed in expert video game players that are relevant to object recognition. Video game players tend to have better visual processing and visual attention (Achtman, Green, & Bavelier, 2008). Contrast sensitivity was found to be better in video game players at higher spatial frequencies (Li, Polat, Makous, & Bavelier, 2009). A subsequent experiment showed that 9 weeks of playing an action video game, Call of Duty 2, further increased contrast sensitivity in video game players relative to a non-action video game, The Sims 2. This finding casts doubt on self-selection as an explanation for the effects of video game experience. Video game players appear to have differences in attention as well; they are more sensitive to attention-capturing events, such as an abrupt change in motion (West, Stevens, Pun, & Pratt, 2008), and have better selective attention (Karle, Watter, & Shedden, 2010).

Although the literature has typically focused on the link between action video game experience and a specific perceptual skill, Riesenhuber (2004) concluded that “just as playing the increasingly complex and realistic video games of today draws on a varied set of skills, it is likely that stimulus- and task-driven plasticity at multiple levels of visual processing contribute to the observed advantages” (p. 73). In an exploratory study of non–video game players, Baniqued and colleagues (2013) examined the link between cognitive skills and different types of web-based causal games. Using a principal component analysis, they identified cognitive skills tapped by each game and found relationships between tests of working memory, fluid intelligence, and perceptual speed with games that tapped those components.

In summary, the literature suggests that video game players have a number of performance advantages that are used during object recognition. These advantages may impact object recognition by accelerating the process of attending to and identifying the outlines and geons of objects and, therefore, may affect perception and recognition of combat targets.

Video games as training

Experimental research shows a diverse collection of cognitive skills that improve as a direct result of video game play. As a training intervention, researchers have examined the specific effects of relatively short-term exposure to video games in novice individuals. The goal of this research has been to facilitate development of computer-based training products that target specific skills. Frequently, these participants are novices in both the performance task and the video game used for training.

Wilson and colleagues (2009) distinguished video games as training from simulation-based training. Although there is some overlap, video games generally carry contrived rules and goals, whereas simulation reproduces a real-world task. Part-task trainers in aviation or health care are examples of simulation-based training; they simulate a specific procedure with the aim of improving performance on that procedure. Serious games, which are not yet well distinguished from other forms of computer-based training (Mayer, 2012), fall into this category, as well. Transfer-appropriate processing theory explains the benefit of simulation-based training (Blaxton, 1989). That is, specific training is most effective when cognitive skills used during training match those required at task performance.

The performance benefits from short-term exposure to video games are more complex. There is little evidence that in the short term, all types of video game play globally improve performance (Lee et al., 2012). At the same time, the list of cognitive improvements observed after short-term video game play is lengthy (Powers et al., 2013). The content of the video game matters, but the specific mechanisms are not well understood. This finding was illustrated in research on the game Space Fortress (Mané & Donchin, 1989), in which players shoot missiles to destroy a fortress while defending their own ship and managing the location of mines. Hart and Battiste (1992) trained student pilots with either Space Fortress or Apache Strike, a commercial, off-the-shelf (COTS) game in which players fly a helicopter. Space Fortress training was associated with better pilot performance than in groups that played Apache Strike or no game at all. Although the mechanisms behind the effectiveness of one game versus the other were not established by this experiment, this research demonstrated that video games do not unilaterally improve performance. The authors suggested that Space Fortress improved attention management and the ability to cope with high-workload tasks.

In a meta-analysis, the genre of game moderated effect size (Powers et al., 2013). Motion-based games in which players parrot predefined behaviors had the largest effect size, but there was no difference in action versus non-action games. However, Powers and colleagues (2013) included “educational games, sports games, and simulation games such as Word Whomp, Mario Kart, and The Sims 2” (p. 5) in their non-action game category. This finding can be reconciled with that of Green and Bavelier (2006a), who found that action video games enhance visual attention. They defined action video games as “those that have fast motion, require vigilant monitoring of the visual periphery, and often require the simultaneous tracking of multiple targets” (Green & Bavelier, 2006a, p. 3). As noted by Powers and colleagues, researchers often select video games to use in experiments because they match information processing in the target task. Oei and Patterson (2013) found that the cognitive benefits extended to non-action video games and mobile video game play, but action games led to improvements across more diverse tasks. Although the content of the video game does matter, more research is needed on the effects of specific game genres in the short term (Powers et al., 2013). Here, we focus on action games and their perceptual and cognitive benefits specific to object perception.

Even short practice with action video games improves perceptual abilities needed for object perception. Generally speaking, object recognition can be thought of as a two-stage process, in which a relevant visual area is examined and individual objects are identified (Nodine & Kundel, 1987). Framed in the terms of Biederman and Shiffrar’s (1987) model, the basic outline of the vehicle and its geons are identified first, followed by closer inspection of the features that distinguish objects within the class.

Video game play has been shown to increase the efficiency of the first stage of object recognition (Green & Bavelier, 2003). For example, Green and Bavelier (2007) found improvements in spatial resolution. Spatial resolution of visual processing refers to the ability to distinguish a visual target from nearby distractor objects. Specifically, they found that non-gamers significantly increased in their spatial resolution after 30 hr of playing a first-person shooter, Unreal Tournament 2004, but not after 30 hr of playing Tetris.

These benefits may be due to improvements in visual working memory (Green & Bavelier, 2006b), contrast sensitivity (Li et al., 2009), executive control (Anguera et al., 2013; Strobach, Frensch, & Schubert, 2012), and visual attention (Chiappe, Gonger, Liao, Caldwell, & Vu, 2013; Green, Sugarman, Medford, Klobusicky, & Bavelier, 2012; Oei & Patterson, 2013; Wu et al., 2012). Although Boot, Kramer, Simons, Fabiani, and Gratton (2008) did not find significant effects for video game experience in a cognitive battery, with the exception of spatial ability, their findings may be explained by a short training time of 20 hrs of video game play (Green et al., 2012). This null finding must be also considered in the context of experimental studies demonstrating cognitive improvements after video game play, even with shorter training times (e.g., Feng, Spence, & Pratt, 2007).

The primitive stimuli used by Green and Bavelier (2007) ruled out strategy development as the mechanism for performance improvement. The authors previously found that 10 days of video game training increased the number of objects participants could attend to accurately (Green & Bavelier, 2003). Green, Puget, and Bavelier (2010) demonstrated that video games improved probabilistic inferences. That is, participants exposed to video game play were better able to make decisions using implicit, probabilistic rules. Powers and colleagues (2013) found that COTS games improve information processing, but their meta-analysis suggested that video games have a greater effect on visual and auditory perceptual skills than they do on executive functioning. Video games also lead to faster reaction times across a number of tasks (Dye, Green, & Bavelier, 2009; Li et al., 2010). Ultimately, short-term video game play is likely to have positive, short-term effects on object perception through multiple cognitive mechanisms, but more research is needed to explain which games will offer the most benefit.

In summary, establishing best practices for CID training is an integral step toward reducing the number of fratricides. Stereoscopic systems have little research in this area, and none in the direct identification of combat vehicles, and yet they are being adopted regularly across a multitude of simulation products and systems. Further, video game experience is an important factor in predicting performance, specifically in target identification (McKinley et al., 2011). Therefore, in this article, we aim to better understand the effects of active stereoscopic displays and video game experience on simulated CID tasks.

Method

Participants

Initially, 71 male participants were recruited from a large southeastern university in exchange for class credit. All participants were over the age of 18 (ranging from 18 to 33, M = 20.3), and none had prior combat training or experience. Participants all had normal or corrected-to-normal vision.

Materials and Procedure

This study was conducted using the military armored vehicle expertise, recognition, identification, classification, and knowledge (MAVERICK) simulation-based training system (Figure 1). This system utilized NVIDIA’s GeForce technology, which has been used successfully to create stereoscopic displays in studies involving teleoperation of unmanned military assets (Chen et al., 2010). The GeForce system allows for active stereoscopy via shutter glass technology (Figure 2). Active stereoscopy indicates that the system “actively” introduces stereoscopic information through the use of shutter glass technology. This technology synchronizes LED lenses that darken alternately between each eye with an on-screen image that is doing the same. This technology induces binocular information, leading to a visual experience of seeing in 3-D.

Figure 1.

Partial view of the MAVERICK (military armored vehicle expertise, recognition, identification, classification, and knowledge) training simulation.

Figure 2.

Side-by-side comparison of the UK Challenger in (a) normal (non-stereoscopic viewing condition) and (b) stereoscopic condition.

The MAVERICK simulation consists of a small first-person shooter gaming environment where participants control an avatar (i.e., an in-game representation of the player). Guided by a trained observer, participants were taught the critical features (Biederman & Shiffrar, 1987; Bramley, 1973), names, and alliance of 12 military vehicles (i.e., BTR, Challenger, BMP, BRDM, M60, M1A1, T80, T72, LAV, M113, ZSU, and Bradley). Participants controlled the avatar but were verbally guided through the environment by a trained experimenter. Upon reaching a vehicle, they were asked to point to specific cues of that vehicle and verbalize the visual information (e.g., the BTR has four wheels per side; the M1A1 has seven wheels per side). They were told whether or not they were correct and were given correct information if wrong. Following, they were told the name of the vehicle and were asked to recite it to acknowledge hearing it. This step was followed by the alliance of the vehicle (i.e., U.S. ally or enemy), again with corrective feedback when needed. Immediately following this training, the simulation was restarted and participants were asked to again walk to each vehicle and announce the name and alliance out loud to the trained observer (i.e., training outcomes for naming vehicles).

Performance was measured with the use of an online survey that provided participants with a sequence of 144 photographic images of the studied vehicles (12 images of each of the 12 vehicles studied). Images were randomized for each subject, and each image was presented for 5 s, followed by a prompt asking the participant to categorize the alliance of the vehicle (ally or enemy) and to identify the vehicle by typing its name. All data were automatically logged into an Excel database. Note that reaction times were measured for the alliance discrimination but not for identification due to large expected variations in time typing the name of the vehicles. After presentation of the 144 items, participants then observed 12 videos of the vehicles performing maneuvers (one of each vehicle) and were asked to provide the same information as previously (alliance and identity of the vehicle).

Design

The experiment involved a 2 × 2 between-subjects design. The first independent variable was video game experience, and it was measured with the use of three 7-point Likert scales to assess how often each participant played video games in general, how often they played first-person shooters, and how familiar they were with different controller types. Only participants scoring in the top 25% (totaled score across the three items > 19) or the bottom 25% (totaled score across the three items < 15.25) of video game experience, which we will refer to hereafter as VGE, were used in further analyses. This procedure led to a reduced sample of 32 individuals. The second independent variable was manipulated stereoscopy, which could be turned on or off in the MAVERICK system by disengaging the shutter glasses. Therefore, the four between-subjects groups were as follows: non-stereoscopic, low VGE (n = 10); stereoscopic, low VGE (n = 7); non-stereoscopic, high VGE (n = 8); and stereoscopic, high VGE (n = 7).

Results

A series of two-way ANOVAs were used to analyze the data for each of the dependent variables. Next we will discuss the results across each of the five dependent variables: identification performance during training, alliance categorization of photographs, identification performance of photographs, alliance categorization of videos, and identification of videos. Please refer to Table 1 for a summary of means and standard deviations and Table 2 for confidence intervals of all performance outcomes.

Table 1:

Estimated Means and Standard Errors for Performance/Categorization Variables

Dependent Variable	Non-Stereo or Stereo	VGE (Low = 0, High = 1)	M	SE
CID performance during training (out of 12)	Non-stereo	0	4.60	0.83
	Non-stereo	1	8.88	0.92
	Stereo	0	4.43	0.99
	Stereo	1	5.71	0.99
Alliance categorization of photographs (out of 144)	Non-stereo	0	75.00	5.70
	Non-stereo	1	115.50	6.37
	Stereo	0	69.57	6.81
	Stereo	1	89.00	6.81
CID performance on photographs (out of 144)	Non-stereo	0	22.70	6.88
	Non-stereo	1	76.00	7.69
	Stereo	0	17.43	8.22
	Stereo	1	31.57	8.22
Alliance categorization of videos (out of 12)	Non-stereo	0	6.80	0.85
	Non-stereo	1	9.88	0.95
	Stereo	0	6.14	1.01
	Stereo	1	7.14	1.01
CID performance on videos (out of 12)	Non-stereo	0	2.60	0.74
		1	7.25	0.83
	Stereo	0	1.43	0.88
		1	2.14	0.88

Note. CID = combat identification; VGE = video game experience.

Table 2:

Estimated 95% Confidence Intervals for Performance/Categorization Variables

Dependent Variable	Non-Stereo or Stereo	VGE (Low = 0, High = 1)	95% Confidence Interval
CID performance during training	Non-stereo	0	[2.91, 6.29]
		1	[7.00, 10.76]
	Stereo	0	[2.41, 6.45]
		1	[3.70, 7.73]
Alliance categorization performance on photographs	Non-stereo	0	[63.32, 86.68]
		1	[102.45, 128.55]
	Stereo	0	[55.62, 83.53]
		1	[75.04, 102.96]
CID performance on photographs	Non-stereo	0	[8.61, 36.79]
		1	[60.25, 91.75]
	Stereo	0	[0.59, 34.27]
		1	[14.73, 48.41]
Alliance categorization performance on videos	Non-stereo	0	[5.07, 8.53]
		1	[7.94, 11.81]
	Stereo	0	[4.07, 8.21]
		1	[5.07, 9.21]
CID performance on videos	Non-stereo	0	[1.09, 4.11]
		1	[5.56, 8.94]
	Stereo	0	[–0.38, 3.24]
		1	[0.34, 3.95]

Note. CID = combat identification; VGE = video game experience.

Identification Performance During Training

There was a significant main effect of VGE on training outcomes (i.e., immediately after studying) for naming the identity of the 12 vehicles, F(1, 28) = 8.9, p = .006, partial η² = .24, with the high-VGE group naming significantly more vehicles (M = 7.3, SE = 0.675) than the low-VGE group (M = 4.51, SE = 0.643). The effect of stereoscopy was approaching significance, F(1, 32) = 3.2, p = .085, with the non-stereoscopic group naming more vehicles (M = 6.74, SE = 0.618) than the stereoscopic group (M = 5.07, SE = 0.697; Figure 3).

Figure 3.

Training identification performance across all conditions.

Performance Outcomes for Alliance Categorization of Photographs

There was a significant main effect of VGE on outcomes for alliance categorization performance of the 144 photographs, F(1, 28) = 21.64, p < .001, partial η² = .44, with the high-VGE group correctly categorizing significantly more of the pictures (72% correct; M = 102.3, SE = 4.7) than the low-VGE group (M = 72.3, SE = 4.4), which performed approximately the same as chance (50% correct). There was also a significant main effect of stereoscopy, F(1, 32) = 6.74, p = .019, partial η² = .18, with the non-stereoscopic group scoring higher (M = 95.3, SE = 4.28) than the stereoscopic group (M = 79.29, SE = 4.82; Figure 4), which also performed close to chance (55% correct).

Figure 4.

Alliance categorization across conditions.

Performance Outcomes for Identification Performance of Photographs

There was a significant main effect of VGE on outcomes for identification performance on the 144 photographs, F(1, 28) = 18.83, p < .001, partial η² = .40, with the high-VGE group correctly naming the identity of more pictures of vehicles (M = 53.8, SE = 5.6) than the low-VGE group (M = 20.1, SD = 5.36). There was also a significant main effect of stereoscopy, F(1, 28) = 10.23, p < .003, partial η² = .18, with the non-stereoscopic group scoring higher (M = 49.35, SE = 5.2) than the stereoscopic group (M = 24.5, SE = 5.8). Further, we found a significant interaction between VGE and stereoscopy, F(1, 28) = 6.35, p < .018, partial η² = .19, with the high-VGE, non-stereoscopic group (M = 76, SE = 7.7) scoring significantly higher than the low-VGE, non-stereoscopic group (M = 22.7, SE = 6.9; Figure 5).

Figure 5.

Combat identification performance on photographs across conditions.

Performance Outcomes for Alliance Categorization of Videos

There was a significant main effect of VGE on alliance categorization performance while observing the 12 videos: F(1, 28) = 4.55, p < .042, partial η² = .14, with the high-VGE group correctly deciding on the alliance of more vehicles (M = 8.6, SE = 0.69) than the low-VGE group (M = 6.5, SE = 0.66). The effect of stereoscopy approached significance, F(1, 28) = 3.15, p = .087, partial η² = .10, with the non-stereoscopic group scoring higher (M = 8.17, SE = 0.63) than the stereoscopic group (M = 6.65, SE = 0.72; Figure 6).

Figure 6.

Performance on alliance categorization of videos across all conditions.

Performance Outcomes for Identification of Videos

There was a significant main effect of VGE on outcomes for identification performance on the 12 videos, F(1, 28) = 10.35, p < .003, partial η² = .27, with the high-VGE group correctly naming the identity of more videos of vehicles (M = 4.87, SE = 0.6) than the low-VGE group (M = 2.17, SE = 0.58). There was also a significant effect of stereoscopy, F(1, 28) = 14.18, p < .001, partial η² = .34, with the non-stereoscopic group scoring higher (M = 4.67, SE = 0.53) than the stereoscopic group (M = 1.78, SE = 0.62). There was also a significant interaction between VGE and stereoscopy, F(1, 28) = 5.57, p < .025, partial η² = .17, with the high-VGE, non-stereoscopic group (M = 7.25, SE = .83) scoring significantly higher than the low-VGE, non-stereoscopic group (M = 2.6, SE = 0.74; Figure 7).

Figure 7.

Performance outcomes for identification of videos across all conditions.

Reaction Time Analysis

There was a significant interaction effect of VGE and stereoscopic condition on average reaction time, F(1, 28) = 5.96, p < .021, partial η² = .18, with the high-VGE, stereoscopic group (M = 5043.45 ms, SE = 759.52) obtaining significantly faster reaction times than the high-VGE, non-stereoscopic group (M = 6207.63 ms, SE = 710.47; Figure 8). There were no main effects, and differences for the low-VGE group were non-significant. Refer to Tables 3 and 4 for details on the means, standard errors, and confidence intervals for reaction times.

Figure 8.

Reaction time outcomes across all conditions.

Table 3:

Estimated Means and Standard Errors for Reaction Time in Milliseconds

Dependent Variable	Non-Stereo or Stereo	VGE (Low = 0, High = 1)	M	SE
Reaction time	Non-stereo	0	5815.49	635.46
		1	6207.63	710.47
	Stereo	0	8157.10	759.52
		1	5043.45	759.52

Note. VGE = video game experience.

Table 4:

Estimated 95% Confidence Intervals for Reaction Time in Milliseconds

Dependent Variable	Non-Stereo or Stereo	VGE (Low = 0, High = 1)	95% Confidence Interval
Reaction time	Non-stereo	0	[4513.80, 7117.17]
		1	[4752.30, 7662.96]
	Stereo	0	[4752.30, 7662.96]
		1	[3487.64, 6599.26]

Note. VGE = video game experience.

Signal Detection Analyses

Signal detection analyses were conducted across the four groups. The results indicate that there was no effect for d prime across VGE, F(1, 28) = .08, p = .78, or stereoscopy, F(1, 28) = .19, p = .66). The c’ criterion did not change based on VGE, F(1, 28) = .02, p = .88, or stereoscopic condition, F(1, 28) = .97, p = .33. Tables 5 and 6 present the means, standard errors, and confidence intervals for these analyses.

Table 5:

Estimated Means and Standard Errors for Signal Detection Analyses

Dependent Variable	Non-Stereo or Stereo	VGE (Low = 0, High = 1)	M	SE
d’	Non-stereo	0	1.092	0.720
	Non-stereo	1	0.824	0.893
	Stereo	0	0.785	0.320
	Stereo	1	0.912	0.540
c	Non-stereo	0	0.141	0.352
		1	0.269	0.553
	Stereo	0	0.149	0.337
		1	–0.022	0.282

Note. VGE = video game experience.

Table 6:

Estimated 95% Confidence Intervals for Signal Detection Analyses

Dependent Variable	Non-Stereo or Stereo	VGE (Low = 0, High = 1)	95% Confidence Interval
d’	Non-stereo	0	[0.657, 1.526]
	Non-stereo	1	[0.338, 1.310]
	Stereo	0	[0.265, 1.305]
	Stereo	1	[0.400, 1.440]
c	Non-stereo	0	[–0.116, 0.398]
		1	[–0.019, 0.556]
	Stereo	0	[–0.159, 0.456]
		1	[–0.329, 0.285]

Note. VGE = video game experience.

Discussion

VGE in CID

It appears that CID performance is at least partially predicted by an individual’s level of VGE. With the research design used in this study, we specifically aimed to examine differences between individuals who self-reported high levels of VGE contrasted with those who reported low levels. Due to the selective nature of the military for placement in occupations, it is important to distinguish between these groups. Further, these findings clearly indicate that VGE does in fact contribute to the skills needed for memorizing and later identifying military vehicles in a simulation-based task. Across multiple indices of performance, including simulated vehicles, photographs, and videos, participants performed better when they were self-identified video game experts compared to video game novices.

Active Stereoscopy in CID

Although stereoscopy has been found to enhance learning and visual performance in other domains and tasks, it seems that it does not help in the learning and memorization of military vehicles in a simulation-based task environment. Further, the stereoscopic conditions led to decrements in performance regardless of one’s previous VGE, lending credence to the explanation that it was not inexperience in operating the avatar in the environment that prevented strong performance outcomes. Interestingly, the interaction of stereoscopy and VGE led to deficits on some of the dependent variables for the high-VGE group. Therefore, it seems that the stereoscopic information may very well interfere with the benefits brought about by high levels of VGE.

These effects may be explained by the work of Cockburn and McKenzie (2002), who found that stereoscopic interfaces can appear “cluttered” and “less efficient” to the user than non-stereo systems. Further, Garg, Norman, Eva, Spero, and Sharan (2002) found that those trained in a comparable non-stereoscopic system outperformed individuals who instead learned wrist anatomy in a stereoscopic simulation. Others (Bowman et al., 2004) have argued that accommodation cues are usually not present in 3-D user interfaces because objects are always in focus at the same depth (i.e., the screen) although they simulate depths beyond the screen. They have argued that this design can lead to an accommodation-convergence mismatch, which sends conflicting signals to the visual centers of the brain. This argument may explain our results. In other words, the conflict caused by incongruent depth cues may overcome the enhancements caused by video game expertise. Authors of future work will need to examine this topic, which we will discuss in the next section.

Yet another argument for our interaction findings could be based in the theory of transfer-appropriate processing (Blaxton, 1989; Graf & Ryan, 1990). Specifically, individuals who are video game experts are likely to have learned on non-stereoscopic displays. Therefore, the introduction of a stereoscopic display may interfere with their capabilities as gamers because the type of processing they normally use during play may not transfer well to stereoscopic displays. We discuss future directions for this avenue of research in the next section.

Further, our analysis of reaction times demonstrated an interesting result. Although there were no effects between groups on our signal detection analyses (d prime and c), the reaction time differences for the high-VGE, stereoscopic group were coupled to significantly worse performance outcomes compared to the high-VGE, non-stereo group. This speed-accuracy trade-off will need further research to better understand why individuals in a stereoscopic training system were more likely to respond quickly yet inaccurately.

Limitations and Future Work

Due to the nature of this experiment (i.e., in a laboratory setting), it could be argued that the external validity of this study is limited. Real combat decisions are made under extraordinary pressure and involve a multitude of factors that were not reproduced in our study, such as loud noises, obscured views, command-and-control needs, and team communication and decision making. Despite this limitation, this study still maintains implications for training because knowledge of vehicles is essential to visual identification. Degraded environments may inhibit the warfighter’s ability to acquire targets, but without detailed knowledge of the vehicles, CID is simply not possible. Another limitation is the way in which we measured VGE (i.e., three questions using Likert scales). The measure used in this study was not specific in regard to which type of video games participants play, aside from first person shooters, which makes it impossible to know which types of games and experience led to an individual’s increases in performance. This limitation also leads to issues concerning exposure to stimuli from military-style games, something that must be accounted for in future studies of these relationships. Further, more work needs to be done to create taxonomies of video games in relation to the real-world tasks that they may influence. The aim of this study was to provide groundwork demonstrating a relationship, if any, between VGE and real-world performance in CID tasks. Although further empirical work is needed, we believe that this study demonstrates a potential relationship between these two variables.

Future research will also need to focus on understanding the differences in performance between males and females in CID tasks. This study involved only male participants, which limits the generalizability of these findings. Further, we cannot account for gender differences or possible interactions between gender and VGE. Research shows that on one hand, men and women tend to play different types of video games (Quaiser-Pohl, Geiser, & Lehmann, 2006), but on the other hand, playing action video games can shrink the differences between male and female spatial abilities (Feng, Spence, & Pratt, 2007). Authors of future work should aim to identify performance differences isolated by types of video games across genders to gain a better understanding of the underlying cognitive mechanisms enhanced by video games and how these enhancements translate to real-world operational environments.

It is important for researchers to continue to develop a better understanding of the role VGE and stereoscopy plays in CID tasks, with specific attention to the interactive effects of convergence-accommodation mismatch and VGE. Related to this goal, work should be conducted on transfer-appropriate process in relation to video game experts who regularly game using 3-D displays compared to those who do not. Also, more work needs to focus on which aspects of VGE led to the results found in this study. Further delineating which types of video games lead to specific increases in real-world tasks will be foundational to better understanding the effects found in this study. Although work has already begun in this area (Baniqued et al., 2013), the relationship between specific video game qualities and performance in real-world tasks is relatively untouched. Stereoscopy is relatively new as an affordable technology at the consumer level, and therefore, many video game players may simply not be accustomed to using these systems. Therefore, authors of future research should investigate changes related to stereoscopy if more video game players begin to use the technology.

Key Points

Self-reported video game experience seems to be a key predictor in simulated combat identification of military armored vehicles.

Active stereoscopy, contrary to our predictions, actually inhibits performance in CID tasks and can have adverse effects regardless of an individual’s video game experience.

Individuals who had high levels of video game experience and who were not in the active stereoscopy condition outperformed the rest of our sample on a multitude of CID measures.

Footnotes

Joseph R. Keebler is an assistant professor at Wichita State University. His research is focused on training and teamwork in medical, military, and consumer systems. As director of the Training Research and Applied Cognitive Engineering Laboratory, he works on multiple efforts examining these topics. He received his PhD in applied/experimental human factors psychology from the University of Central Florida in 2011 and continued on in a postdoc at the Institute for Simulation and Training studying teamwork in complex systems. His current projects are examining topics such as the efficacy of augmented reality training systems in identifying combat vehicles, teamwork on virtually dispersed teams, and the development of the ideal handoff protocol for pediatric hospitalists.

Florian Jentsch is a professor at the University of Central Florida, with joint appointments in the Department of Psychology and the Institute of Simulation and Training. He is also director of the Team Performance Laboratory. He received his PhD in human factors psychology in 1997 from the University of Central Florida, and he holds master’s degrees in aeronautical science and aeronautical engineering. He is an associate editor for Human Factors and Cognitive Technology and a consulting editor for the International Journal of Applied Aviation Studies. His research interests include team performance, training, aviation, human-robot interaction, and simulation technology. He has coauthored over 200 publications, books, and presentations.

David Schuster is an assistant professor at San Jose State University. He received his PhD in psychology, specializing in applied experimental and human factors, from the University of Central Florida in 2013. His research interests include the cognitive aspects of cybersecurity, situation awareness in human-automation teams, and the impact of video games on task performance.

References

Achtman

R. L.

Green

C. S.

Bavelier

(2008). Video games as a tool to train visual skills. Restorative Neurology and Neuroscience, 26, 435–446.

Anguera

J. A.

Boccanfuso

Rintoul

J. L.

Al-Hashimi

Faraji

Janowich

. . . Gazzaley

(2013). Video game training enhances cognitive control in older adults. Nature, 501, 97–101.

Baniqued

P. L.

Lee

Voss

M. W.

Basak

Cosman

J. D.

DeSouza

. . . Kramer

A. F.

(2013). Selling points: What cognitive abilities are tapped by casual video games? Acta Psychologica, 142, 74–86.

Biederman

Shiffrar

M. M.

(1987). Sexing day-old chicks: A case study and expert systems analysis of a difficult perceptual-learning task. Journal of Experimental Psychology: Learning, Memory, and Cognition, 13, 640–645.

Bishop

P. O.

Pettigrew

J. D.

(1986). Neural mechanisms of binocular vision. Vision Research, 26, 1587–1600.

Blaxton

T. A.

(1989). Investigating dissociations among memory measures: Support for a transfer-appropriate processing framework. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15, 657–668.

Boot

W. R.

Blakely

D. P.

Simons

D. J.

(2011). Do action video games improve perception and cognition? Frontiers in Psychology, 2(226), 1–6. doi:10.3389/fpsyg.2011.00226

Boot

W. R.

Kramer

A. F.

Simons

D. J.

Fabiani

Gratton

(2008). The effects of video game playing on attention, memory, and executive control. Acta Psychologica, 129, 387–398. doi:10.1016/j.actpsy.2008.09.005

Bowman

D. A.

Kruijff

LaViola

J. J.

Jr. Poupyrev

(2004). 3D user interfaces: Theory and practice. Boston, MA: Addison-Wesley.

10.

Bramley

(1973). Equipment recognition training: Learning whole or individual features? Occupational Psychology, 47, 131–139.

11.

Briggs

R. W.

Goldberg

J. H.

(1995). Battlefield recognition of armored vehicles. Human Factors, 37, 596–610.

12.

Bryant

D. J.

Smith

D. G.

(2013). Impact of blue force tracking on combat identification judgments. Human Factors, 55, 75–89.

13.

Chen

J. Y. C.

Haas

E. C.

Barnes

M. J.

(2007). Human performance issues and user interface design for teleoperated robots. IEEE Transactions on Systems, Man, and Cybernetics, 37, 1231–1245.

14.

Chen

J. Y. C.

Oden

R. V.

Kenny

Merritt

J. O.

(2010). Effectiveness of stereoscopic displays for indirect-vision driving and robotic teleoperation. Aberdeen Proving Ground, MD: Army Research Laboratory.

15.

Chiappe

Gonger

Liao

Caldwell

J. L.

K.-P. L.

(2013). Improving multi-tasking ability through action videogames. Applied Ergonomics, 44, 278–284.

16.

Cockburn

McKenzie

(2002). Evaluating the effectiveness of spatial memory in 2D and 3D physical and virtual environments. Spatial Cognition, 4, 203–210.

17.

Dye

M. W. G.

Green

C. S.

Bavelier

(2009). Increasing speed of processing with action video games. Current Directions in Psychological Science, 18, 321–326. doi:10.1111/j.1467-8721.2009.01660.x

18.

Edelman

Bulthoff

H. H.

(1992). Orientation dependence in the recognition of familiar and novel views of three-dimensional objects. Vision Research, 32, 2385–2400.

19.

Feng

Spence

Pratt

(2007). Playing an action video game reduces gender differences in spatial cognition. Psychological Science, 18, 850–855.

20.

Gagnon

(1985). Videogames and spatial skills: An exploratory study. Educational Communications and Technology Journal, 33, 263–275.

21.

Garg

A. X.

Norman

G. R.

Eva

K. W.

Spero

Sharan

(2002). Is there any real virtue of virtual reality? The minor role of multiple orientations in learning anatomy from computers. Academic Medicine, 77(10), S97–S99.

22.

Graf

Ryan

(1990). Transfer-appropriate processing for implicit and explicit memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16, 978–992.

23.

Green

C. S.

Bavelier

(2003). Action video game modifies visual selective attention. Nature, 423, 534–537.

24.

Green

C. S.

Bavelier

(2006a). Effect of action video games on the spatial distribution of visuospatial attention. Journal of Experimental Psychology: Human Perceptual Performance, 32, 1465–1478. doi:10.1037/0096-1523.32.6.1465

25.

Green

C. S.

Bavelier

(2006b). Enumeration versus multiple object tracking: The case of action video game players. Cognition, 101, 217–245.

26.

Green

C. S.

Bavelier

(2007). Action-video-game experience alters the spatial resolution of vision. Psychological Science, 18, 88–94. doi:10.1111/j.1467-9280.2007.01853.x

27.

Green

C. S.

Puget

Bavelier

(2010). Improved probabilistic inference as a general mechanism for learning with action video games. Current Biology, 20, 1573–1579.

28.

Green

C. S.

Sugarman

M. A.

Medford

Klobusicky

Bavelier

(2012). The effect of action video game experience on task-switching. Computers in Human Behavior, 28, 984–994.

29.

Griffith

J. L.

Voloschin

Gibb

G. D.

Bailey

A. J. R.

(1983). Differences in eye-hand motor coordination of video-game users and non-users. Perceptual and Motor Skills, 57, 155–158. doi:10.2466/pms.1983.57.1.155

30.

Ground combat systems. (2010, October). ARMY Magazine. Retrieved from http://www.ausa.org/publications/armymagazine/archive/2010/10/Documents/Weapons4_Gound_1010.pdf

31.

Guildea

Bailey

J. H.

(1994). Correlates of route traversal performance in a virtual environment. In Proceedings of the Human Factors and Ergonomics Society 38th Annual Meeting (p. 966). Santa Monica, CA: Human Factors and Ergonomics Society.

32.

Hart

S. G.

Battiste

(1992). Field test of video game trainer. In Proceedings of the Human Factors and Ergonomics Society 36th Annual Meeting (pp. 1291–1295). Santa Monica, CA: Human Factors and Ergonomics Society.

33.

Karle

J. W.

Watter

Shedden

J. M.

(2010). Task switching in video game players: Benefits of selective attention but not resistance to proactive interference. Acta Psychologica, 134, 70–78. doi:10.1016/j.actpsy.2009.12.007

34.

Keebler

J. R.

Harper-Sciarini

Curtis

Schuster

Jentsch

Bell-Carroll

(2007). Effects of 2-dimensional and 3-dimensional media exposure training on a tank recognition task. In Proceedings of the Human Factors and Ergonomic Society 51st Annual Meeting (pp. 1593–1597). Santa Monica, CA: Human Factors and Ergonomics Society.

35.

Keebler

J. R.

Jentsch

Hudson

(2011). Developing an effective combat identification training. In Proceedings of the Human Factors and Ergonomics Society 55th Annual Meeting (pp. 1554–1558) doi:10.1177/1071181311551324

36.

Kim

(2006). Effects of 3D virtual reality of plate tectonics on fifth grade students’ achievement and attitude toward science. Interacting Learning Environments, 14, 25–34.

37.

Lee

Boot

W. R.

Basak

Voss

M. W.

Prakash

R. S.

Neider

. . . Kramer

A. F.

(2012). Performance gains from directed training do not transfer to untrained tasks. Acta Psychologica, 139, 146–158. doi:10.1016/j.actpsy.2011.11.003

38.

Polat

Makous

Bavelier

(2009). Enhancing the contrast sensitivity function through action video game training. Nature Neuroscience, 12, 549–551. doi:10.1038/nn.2296

39.

Mané

Donchin

(1989). The space fortress game. Acta Psychologica, 71, 17–22.

40.

Mayer

(2012). Towards a comprehensive methodology for the research and evaluation of serious games. Proceedings of Computer Science, 15, 233–247. doi:10.1016/j.procs.2012.10.075

41.

McCarthy

(2003, March 26). Friendly fire kills two UK vehicle crew. Guardian.co.uk. Retrieved from http://www.guardian.co.uk/world/2003/mar/26/iraq.rorymccarthy1

42.

McKinley

R. A.

McIntire

L. K.

Funke

M. A.

(2011). Operator selection for unmanned aerial systems: Comparing video game players and pilots. Aviation, Space, and Environmental Medicine, 82, 635–642.

43.

Nakayama

Shimojo

Silverman

G. H.

(1989). Stereoscopic depth: Its relation to image segmentation, grouping, and the recognition of occluded objects. Perception, 18, 55–68.

44.

Nodine

C. F.

Kundel

H. L.

(1987). Using eye movements to study visual search and to improve tumor detection. Radiographics, 7, 1241–1250.

45.

Oei

A. C.

Patterson

M. D.

(2013). Enhancing cognition with video games: A multiple game training study. PLOS One, 8(3), 1–16.

46.

O’Kane

B.L.

Biederman

Cooper

E. E.

Nystrom

(1997). An account of object identification confusion. Journal of Experimental Psychology: Applied, 3, 21–41.

47.

Powers

K. L.

Brooks

P. J.

Aldrich

N. J.

Palladino

M. A.

Alfieri

(2013). Effects of video-game play on information processing: A meta-analytic investigation. Psychonomic Bulletin and Review, 20, 1055–1079. doi:10.3758/s13423-013-0418-z

48.

Quaiser-Pohl

Geiser

Lehmann

(2006). The relationship between computer-game preference, gender, and mental-rotation ability. Personality and Individual Differences, 40, 609–619.

49.

Regan

(1995). Blue on blue: A history of friendly fire. New York, NY: Avon Books.

50.

Rehfeld

Jentsch

Curtis

Fincannon

(2005, July). Collaborative teamwork with unmanned vehicles in military missions. Paper presented at the 11th Annual Human-Computer Interaction International Conference, Las Vegas, NV.

51.

Riesenhuber

(2004). An action video game modifies visual processing. Trends in Neurosciences, 27, 72–74.

52.

Rosser

J. C.

Lynch

P. J.

Cuddihy

Gentile

D. A.

Klonsky

Merrell

(2007). The impact of video games on training surgeons in the 21st century. Archives of Surgery, 142, 181–186.

53.

Salas

Cannon-Bowers

J. A.

(2001). The science of training: A decade of progress. Annual Review of Psychology, 52, 471–499.

54.

Schlickum

M. K.

Hedman

Enochsson

Kjellin

Fellander-Tsai

(2009). Systematic video game training in surgical novices improves performance in virtual reality endoscopic surgical simulators: A prospective randomized study. World Journal of Surgery, 33, 2360–2367.

55.

Schuster

Fincannon

Jentsch

Keebler

Evans

A. W.

III . (2008, December). The role of spatial ability in the relationship between video game experience and route effectiveness among unmanned vehicle operators. Poster session presented at the meeting of the Army Science Conference, Orlando, FL.

56.

Strobach

Frensch

P. A.

Schubert

(2012). Video game practice optimizes executive control skills in dual-task and task switching situations. Acta Psychologica, 140, 13–24.

57.

Turner

T. L.

Hellbaum

R. F.

(1986). LC shutter glasses provide 3-D display for simulated flight. Journal of Information Display, 2(9), 22–24.

58.

U.S. Army Training Support Centers. (1996). Combat vehicle graphic training aid GTA 13-2-01. Washington, DC: Department of the Army.

59.

U.S. Department of Defense. (1996). Report of the Defense Science Board Task Force on Command, Control, Communications, Computers, Intelligence, Surveillance and Reconnaissance (C4ISR) integration. Retrieved from http://www.fas.org/irp/offdocs/ADA326142.htm

60.

West

G. L.

Stevens

S. A.

Pun

Pratt

(2008). Visuospatial experience modulates attentional capture: Evidence from action video game players. Journal of Vision, 8(16), 1–9.

61.

Wilson

K. A.

Bedwell

W. L.

Lazzara

E. H.

Salas

Burke

C. S.

Estock

J. L.

. . . Conkey

(2009). Relationships between game attributes and learning outcomes: Review and research proposals. Simulation & Gaming, 40, 217–266. doi:10.1177/1046878108321866

62.

Cheng

C. K.

Feng

D’Angelo

Alain

Spence

(2012). Playing a first-person shooter video game induces neuroplastic change. Journal of Cognitive Neuroscience, 24, 1286–1293.