Casual Video Games as Training Tools for Attentional Processes in ...

CYBERPSYCHOLOGY, BEHAVIOR, AND SOCIAL NETWORKING Volume 18, Number 11, 2015 ª Mary Ann Liebert, Inc. DOI: 10.1089/cyber.2015.0316

Casual Video Games as Training Tools for Attentional Processes in Everyday Life Michael J. Stroud, PhD,1 and Susan Krauss Whitbourne, PhD 2

Abstract

Three experiments examined the attentional components of the popular match-3 casual video game, Bejeweled Blitz (BJB). Attentionally demanding, BJB is highly popular among adults, particularly those in middle and later adulthood. In experiment 1, 54 older adults (Mage = 70.57) and 33 younger adults (Mage = 19.82) played 20 rounds of BJB, and completed online tasks measuring reaction time, simple visual search, and conjunction visual search. Prior experience significantly predicted BJB scores for younger adults, but for older adults, both prior experience and simple visual search task scores predicted BJB performance. Experiment 2 tested whether BJB practice alone would result in a carryover benefit to a visual search task in a sample of 58 young adults (Mage = 19.57) who completed 0, 10, or 30 rounds of BJB followed by a BJB-like visual search task with targets present or absent. Reaction times were significantly faster for participants who completed 30 but not 10 rounds of BJB compared with the search task only. This benefit was evident when targets were both present and absent, suggesting that playing BJB improves not only target detection, but also the ability to quit search effectively. Experiment 3 tested whether the attentional benefit in experiment 2 would apply to non-BJB stimuli. The results revealed a similar numerical but not significant trend. Taken together, the findings suggest there are benefits of casual video game playing to attention and relevant everyday skills, and that these games may have potential value as training tools.

Introduction

I

nterventions designed to benefit the performance of older adults in laboratory tasks increasingly support the plasticity view of aging.1 A subset of these seek to capitalize on the seemingly natural potential of video games to enhance attention.2 If these interventions prove successful, they could help to offset typical age-related attentional declines.3–5 The key test of video game interventions will be not only whether players can enhance performance on the games themselves, but whether the performance enhancement transfers to attentional tasks in everyday life.6 Games that present players with stimuli approaching those they might encounter in everyday search tasks could therefore have beneficial effects not only on attention, but on driving-related tasks. Previous research using the Nintendo Wii failed to yield evidence for transfer.7 There is evidence that video game training, in contrast, improves executive control in older adults,8 a skill relevant to driving. Studies on video game training (Tetris and Medal of Honor) and driving skills has yielded mixed results, with positive effects on visual attention but no effects on driving simulator performance.9 1 2

Video games also have motivational properties that could make them useful in training older adults. Older adults have reported in prior training studies that they dislike first-person shooter games because they find them violent and intimidating.10,11 So-called ‘‘casual’’ video games may be better training platforms for this population. These video games involve simplified rules, short time commitments, and nonviolent stimuli often based on cartoon images. Casual video games are typically available without cost through social media or mobile device platforms. The primary purpose of the present set of experiments was to investigate the attentional components of Bejeweled Blitz (BJB), a one-minute per game match-3 casual video game widely available via social media. This genre of games is highly popular among adults,12 making it a potentially useful tool for studying the benefits of such games for maximizing performance in other cognitive tasks. Baniqued et al.13 identified several cognitive components of casual video games, and recommend that future training studies should be conducted to assess the possible transfer of cognitive ability. The current experiments focus on one specific casual game to attempt to validate their findings.

Department of Psychology, Merrimack College, North Andover, Massachusetts. Department of Psychological and Brain Sciences, University of Massachusetts Amherst, Amherst, Massachusetts.

654

CASUAL VIDEO GAME PLAY IN ADULTS

655

In BJB, players score points by aligning at least three gemstones of the same shape and color in rows, columns, or L-shapes within an 8 · 8 matrix. Players make moves by clicking on adjacent gems in order to make a match. BJB is essentially a visual search task with the target always present, with color and shape as the two most prominent features driving attentional search. However, the target of this ‘‘search’’ is not easily defined because the player cannot simply create a mental representation of a specific target to guide search as in laboratory search tasks. Since a successful move in BJB must consist of at least three matching gems, it was speculated that a fundamental strategy consisting of two separate search phases may be adopted by novice players. The goal of the first search, or ‘‘find-two,’’ is to locate any two adjacent gems of the same color/shape. The next phase, ‘‘find-match,’’ is to scan the neighboring gems for a match. If a third gem that matches the two previously located gems is found, then a valid move can be played. Understanding the dynamics of these two aspects of search may be critical to uncovering what may be important for transfer. Experiment 1

Experiment 1 compared older and younger adults in a single session in which they played BJB for 20 one-minute rounds. Method Participants. The older adult sample (61.5% female) ranged from 65 to 85 years of age (Mage = 70.57 years, SD = 5.26 years). All were healthy, independently living community dwellers with normal (or corrected) vision and hearing. They were compared with 33 undergraduates drawn from the participant pool at the University of Massachusetts, Amherst, who were 18–22 years old (M = 19.82 years, SD = 1.01 years).

Using online software,14 an 8 · 8 grid (subtending a visual angle of 11.56) was constructed. The simple task was a conjunction search based on color and shape containing a red open circle target and blue plus signs and open circles, and red plus sign distractors. The complex task manipulated color, shape, and orientation with an orange triangle target and green triangles, orange inverted triangles, and green inverted triangles distractors. In terms of guiding features,15 the simple task contained one ‘‘absolute attribute’’ (color) while the complex task contained two (color and orientation). Participants completed 100 trials (two blocks each with 50 simple and 50 complex), with the target present in 50% of trials. Each trial provided visible feedback on accuracy (but not reaction time). Five trials of the online Tasks and materials.

‘‘Red Light–Green Light Reaction Time Test’’ measured simple reaction time16 by having participants click the mouse when the stoplight changed color. All tasks were administered on a 15 inch PC laptop computer with participants seated 22 inches from the screen, using both the keyboard and an external mouse. Procedure. Older participants were recruited through local newspapers, and testing took place in a quiet, distraction-free laboratory at the university campus where an undergraduate female research assistant completed all testing. All participants wore headphones during testing. They were compensated $20 for their participation. After providing informed consent, participants were given task instructions and practice. For BJB, they were shown the embedded tutorial included in the game. They completed the visual search task in randomly assigned counterbalanced order. BJB was measured in 20 trials. Results

Among the younger adults, 71% (n = 22) reported playing computer games. Among the older adults, 59.6% (n = 31) reported playing at least one computer game, v2(1) = 1.08, p > 0.05. There were significant effects of age, F(1, 79) = 39.25, p < 0.0001, prior online game experience, F(1, 79) = 18.31, p < 0.0001, and round, F(19, 61) = 2.55, p < 0.003) on BJB scores but no interactions (see Table 1 and Fig. 1). The multivariate analysis of variance on simple reaction time, simple visual search, and conjunction visual search, F(3, 75) = 18.56, p < 0.0001, produced no significant univariate age effect on simple reaction time, but there were significant univariate age effects for simple visual search, F(1, 77) = 50.34, p < 0.0001, and complex visual search, F(1, 77) = 23.20, p < 0.0001. A hierarchical linear regression on BJB scores for younger adults revealed that prior experience was the only significant predictor of performance F(1, 28) = 12.78, p < 0.001, b = 0.56, p < 0.001. For older adults, simple visual search was the only predictor of BJB scores, F(3, 47) = 3.36, p < 0.027; b = –0.29, p < 0.034. Discussion

These findings suggest that participants with prior experience in playing online games, regardless of age, achieve higher scores. It is possible that participants with prior video game experience were self-selected, and that only those with interest and sufficient abilities had attempted to play such games on their own.17 However, even after controlling for prior experience, scores on simple visual search significantly predicted the mean scores of older adults averaged across the

Table 1. Means in Experiment 1 by Age Group and Prior Bejeweled Blitz Play Age group Young adults Older adults

Online game experience

BJB mean (SE)

Yes (n = 22) No (n = 9) Yes (n = 31) No (n = 21)

24,135.80 (2,046.65) 14,094.44 (2,046.65) 11,129.91 (1,102.77) 8,395.00 (1,339.85)

BJB, Bejeweled Blitz; RT, reaction time.

RT, msec (SD) 293.14 328.33 387.42 395.64

(47.16) (107.69) (229.15) (175.37)

Simple search, msec (SD) 1,127.52 1,098.55 2,288.86 2,340.44

(303.39) (371.26) (800.93) (825.68)

Complex search (SD) 2,850.42 2,513.35 5,979.12 5,972.72

(992.15) (821.18) (1,920.32) (4,343.06)

656

STROUD AND WHITBOURNE 45000 40000

FIG. 1. Bejeweled Blitz (BJB) scores by age group, prior BJB play, and trial in experiment 1.

BJB Round Score B

35000 30000 25000 20000 15000 10000 5000 0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

BJB Round Young Adults

20 rounds. At least for older adults, then, BJB may be tapping some components of simple visual search. Experiment 2

The purpose of experiment 2 was to attempt to test whether, among younger adults, playing BJB could result in a positive transfer of skill to a related visual search task. Method Participants. The sample consisted of 58 healthy, normal vision, undergraduate, Merrimack College students (72.5% female), 18–25 years old (Mage = 19.75 years, SD = 1.35 years) recruited from a psychology course for which they received experimental credit. Two participants were eliminated due to a software malfunction. Procedure. Participants completed the training rounds with BJB while seated 24 inches from a Lenovo all-in-one PC with a 17 inch screen running Windows 7. The visual search stimuli were constructed from cropped screenshots of BJB. Each of the six search arrays consisted of nine objects arranged in a 3 · 3 grid subtending a visual angle of 9.05. Targets were defined as any two gems adjacent to one another in the horizontal or vertical direction. Each participant completed 80 total trials with a target appearing in 50% of the trials. The search task was delivered through Superlab Stimulus Presentation Software v4.5. Participants were randomly assigned to one of the three training conditions (none, 10, or 30). For conditions with BJB, participants received the embedded tutorial in the actual BJB game first. They then completed the training rounds (10 vs. 30) followed by the visual search task. Participants in the zero training round condition (control condition) were given the search task only with no mention of BJB. Results

Seventy-five percent of the sample reported that they played BJB, but only 21% reported frequent playing. Paired-samples t tests comparing first and last scores for evidence of improvement across the training intervals showed that BJB scores

Old Adults

did not significantly improve for participants completing only 10 training rounds of BJB, but scores did significantly improve for participants playing 30 training rounds, t(17) = 2.42, p < 0.027 (see Fig. 2). Error rates on the visual search task were relatively low (between 6% and 9%), indicating that the task was not too difficult. Further, there was no significant difference in error rates across all training round conditions, or between absent and present trials ( p > 0.31). Response times were submitted to a 3 · 2 (none vs. 10 vs. 30 · absent vs. present) ANOVA with training round as a between-subjects factor. Planned comparisons were conducted between the 0 and 30, and the 0 and 10 training rounds for both absent and present trials for response times only (see Fig. 3). There was a main effect of target presence F(1, 53) = 137, p < 0.001 and of training round ( p < 0.001), but the interaction between training round and target presence fell just short of significance, F(2, 53) = 2.70, p = 0.077. Comparisons revealed no significant difference between the 10 and no training round conditions for both target absent and present trials ( p > 0.37). However, response times were significantly faster in the 30 training round condition than in the 10 training round condition for both target absent and target present trials ( p < 0.015). Discussion

These results reveal that playing 10 consecutive games of BJB has no significant impact on the subsequent visual search task compared with completing the search task only. However, participants who completed 30 consecutive games of BJB performed significantly better than participants who played 10 or no games at all. Further, this benefit to visual search reaction times came without any cost to accuracy. This increase in performance for the participants playing 30 rounds was revealed in both target present and absent trials, suggesting that not only is target detection more efficient, but that the ability to decide when to terminate search also improves. Based on the above analysis of the search components of BJB, successful play involves being able to locate two gems quickly (find-two). In the visual search task used here, the target was present on only half of the trials, which was not


657

45000 40000

BJB Round Scorer

35000 30000 25000

FIG. 2. Results from experiment 2 of the training sessions for the 10 and 30 conditions.

20000 15000 10000 5000 0 1

2

3

4

5

6

7

8

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 BJB Training Round 10 Games

30 Games

revealed to the participant. It was speculated that visual attention must be engaged similarly for BJB and the visual search task, and thus the findings suggest that an attentionally demanding, casual video game can result in a positive transfer of skill on a similarly designed task when given sufficient training. One potential criticism of experiment 2 involves the similarity between the items in BJB and the subsequent search task. One could argue that playing BJB for an extensive amount of time simply leads to familiarity with the objects and thus success in a task involving the same types of items. Therefore, this would not reflect transfer of attentional processes from one task to another. In order to address this criticism, experiment 3 used different shapes in the search task. Experiment 3

The purpose of experiment 3 was to examine the potential transfer effect of using a search task containing elements that are different from BJB. Since the transfer effect from experiment 2 may involve a multitude of factors, only one aspect of the stimuli is changed: the identity of the objects.

FIG. 3. Reaction times (error bars display the standard error of the mean) for the visual search task across all three training round conditions in experiment 2.

Method Participants. Fifty-nine Merrimack College healthy, normal-sighted undergraduates (77% female) aged 18–31 years old (Mage = 21 years, SD = 1.80 years) completed the study for course credit. Tasks and materials. The same equipment was used as in experiment 2, with fruit objects as shapes to replace gemstones (Fig. 4). Procedure.

The procedure was identical to experiment 2.

Results

Thirty-five percent of participants reported playing BJB before, but only 11% did so often or frequently. There were no improvements in the 10-round condition ( p = 0.31), but scores significantly improved across 30 rounds, t(19) = 2.78, p = 0.012 (see Fig. 5).

FIG. 4. Example of an array for experiment 3. Although the figure is in black and white, each object was presented in a color similar to the gems in BJB. Therefore, the same distribution of colors was present and the only difference was the actual shape of the objects.

658

STROUD AND WHITBOURNE 60000

FIG. 5. Results of the training sessions for the 10 and 30 conditions from experiment 3.

BJB Round Score

50000

40000

30000

20000

10000

0 1

2

3

4

5

6

7

8

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 BJB Training Round Blitz 10

Responses times were faster when the target was present compared with absent, F(1, 57) = 26.17, p < 0.001, and performance improved across target absent trials across the three training rounds, but there was no significant interaction between training round and target presence, F(2, 57) = 1.228, p = 0.30 and no significant differences between training rounds (see Fig. 6). General Discussion

The simplicity of casual games coupled with their cartoonlike content mean that performance on the tasks involved in the current research might provide a more accurate measure of processing speed than is true of the more complex hardcore video games that additionally present violent scenarios potentially offensive to adults seeking more innocuous content. The results of the current experiments represent novel findings that may provide fundamental steps toward revealing the cognitive benefits to adults of playing simple but appealing online games. Experiment 1 showed that both older and younger participants with online game experience in fact had higher BJB

FIG. 6. Reaction times (error bars display the standard error of the mean) for the visual search task for fruit stimuli across all three training rounds in experiment 3.

Blitz 30

scores. Among young adults, neither reaction time nor visual search performance significantly predicted BJB scores. However, for older adults, simple visual search speed significantly predicted mean BJB scores. Thus, the results provide a link between visual search performance, prior gaming experience, and performance playing BJB. This leaves the question open as to whether playing BJB can result in an attentional benefit on a separate task; this was the focus of experiments 2 and 3. In experiment 2, young adults only participated in conditions in which they received varying amount of BJB practice, with the findings indicating positive transfer from 30 rounds of BJB onto a visual search task composed of BJB-like stimuli. The find-two search would require an initial broad distribution of attention aimed at locating relatively denser areas of color (and possibly shape). Relatively speaking, the find-two search phase may be considered more difficult or complex compared with the find-match phase. With six possible color/shape combinations to search for, this is essentially a search for multiple targets, which introduces possible additional costs.18 Initially, this aspect of search would be heavily driven by bottom-up characteristics of the display. For the find-two search, observers simply need to scan one of four possible locations (fewer if the two gems reside along the borders) in order to locate the third gem or to move on. The target in the subsequent visual search task in experiment 2 was to locate two adjacent gems, which is essentially the find-two search phase from BJB. Thus, the results suggest that by playing BJB for 30 minutes, the attentional benefit may be enhancing this aspect of search. Experiment 3 was an attempt to tease apart the factors contributing to the attentional benefit seen in experiment 2. Although the results were not statistically significant, they do reveal some promising trends (see Fig. 7). Playing 10 rounds of BJB does not appear to have any transfer effect on subsequent search compared with playing nothing, regardless of whether the transfer task involves searching for gems or fruit. However, after 30 games, there is a numerical advantage (significant for the array with BJB gems) when the target is present and more so when the target is absent compared with either 10 games or none. This difference in target present trials for gems may suggest some


659

Future research should focus on long-term training as well as other cognitive abilities. With a greater understanding of the cognitive components of this highly popular and widely available online game, BJB could very well serve as an ideal training mechanism for improving attentional skills of younger adults and cognitive plasticity in older adults. Acknowledgments

FIG. 7. Combined reaction time results for experiments 2 and 3. The first three sets of bars corresponds to experiment 2 with the search task for BJB gems and the last three sets of bars represent the search task for fruit.

target-specific learning acquired during the training rounds. However, for the fruit stimuli, this may reflect a more general attentional transfer. The comparison between experiments 2 and 3 may provide some insight into how BJB enhances attention. When the stimuli are familiar from repeated exposure, there appears to be a heightened attentional awareness. Based on the results of experiment 1, it may be that older adults would receive the greatest benefit from playing BJB, as it translates to more complex aspects of visual search. Future research should aim at capturing the benefits of playing BJB as it specifically applies to an older population. If a link can be demonstrated between training on BJB and visual search, then attentionally demanding real-world tasks (e.g., driving) can be considered as well. Researchers are actively exploring the use of video games as a training method for older adults. These games can provide similar features as successful cognitive training programs, including task variability, feedback, motivation, and adaptability.19 Recent findings involving casual video games of a different nature from BJB (working memory and reasoning) have shown improvements in divided attention tasks.20 Ballesteros et al. demonstrated that 20 hours of video game training (across 10–12 weeks) had immediate but not long-term (3 months) cognitive benefits in older adults.21 The results of the current research suggest that a casual game such as BJB may be related to and provide attentional benefits to a similar attentionally demanding task. However, this conclusion does not come without limitations. The current study focuses only on the spatial distribution of attention, which is a large component of visual search–based games such as BJB. Further, only a maximum of 30 minutes of training (for essentially novice players) was included. To the best of the authors’ knowledge, this is the only study to show that BJB may be tapping some components of simple visual search. This finding, if replicated, could provide support for the value of providing older adults with training or practice in BJB. It is hoped that these experiments represent the essential groundwork involved in understanding the cognitive benefits of playing casual video games.

The authors wish to thank the following University of Massachusetts Amherst undergraduate research assistants: Stacy Ellenberg, Kyoko Amimoto (volunteer), Erica Alberti, Avery Ducey, and Wilton Dadmun III. They also wish to thank Merrimack College undergraduate research assistants, Erica Garguilo and Joshua Czerepica. Thanks to Raymond Shaw and Christina Hardway. They would also like to thank two anonymous reviewers for their helpful comments. This research was supported by a grant from PopCap Games (University of Massachusetts Amherst #111-0612). Author Disclosure Statement

No competing financial interests exist. References

1. Goh JO, Park DC. Neuroplasticity and cognitive aging: the scaffolding theory of aging and cognition. Restorative Neurology & Neuroscience 2009; 27:391–403. 2. Bavelier D, Green CS, Pouget A, et al. Brain plasticity through the life span: learning to learn and action video games. Annual Review of Neuroscience 2012; 35:391– 416. 3. Madden DJ. Aging and visual attention. Current Directions in Psychological Science 2007; 16:70–74. 4. Rabbitt P. Do individual differences in speed reflect ‘‘global’’ or ‘‘local’’ differences in mental abilities? Intelligence 1996; 22:69–88. 5. Salthouse TA. Selective review of cognitive aging. Journal of the International Neuropsychological Society 2010; 16:754–760. 6. Green CS, Bavelier D. Exercising your brain: a review of human brain plasticity and training-induced learning. Psychology & Aging 2008; 23:692–701. 7. Ackerman PL, Kanfer R, Calderwood C. Use it or lose it? Wii brain exercise practice and reading for domain knowledge. Psychology & Aging 2010; 25:753–766. 8. Basak C, Boot WR, Voss MW, et al. Can training in a realtime strategy video game attenuate cognitive decline in older adults? Psychology & Aging 2008; 23:765–777. 9. Belchior PDC. (2007) Cognitive training with video games to improve driving skills and driving safety among older adults. 68, ProQuest Information & Learning, US. 10. McKay SM, Maki BE. Attitudes of older adults toward shooter video games: an initial study to select an acceptable game for training visual processing. Gerontechnology 2010; 9:5–17. 11. Spence I, Feng J. Video games and spatial cognition. Review of General Psychology 2010; 14:92–104. 12. Whitbourne SK, Ellenberg SR, Akimoto K. Reasons for playing casual video games and perceived benefits among adults 18 to 80 years old. Cyberpsychology, Behavior, & Social Networking 2013; 16:892–897.

660

13. Banquid P, Lee H, Voss M, et al. Selling points: what cognitive abilities are tapped by casual video games? Acta Psychologica 2013; 142:74–86. 14. Cognitive Science Software. (2005) http://cognitrn.psych .indiana.edu/CogsciSoftware/VisualSearch/ (accessed Oct. 2, 2015). 15. Wolfe JM, Horowitz TS. What attributes guide the deployment of visual attention and how do they do it? Nature Reviews Neuroscience 2004; 5:1–7. 16. Allen J. (2002) Red light–green light reaction time test. http://faculty.washington.edu/chudler/java/redgreen.html (accessed Nov. 16, 2013). 17. Achtman RL, Green CS, Bavelier D. Video games as a tool to train visual skills. Restorative Neurology & Neuroscience 2008; 26:435–446. 18. Stroud MJ, Cave KR, Menneer T, et al. Why is it difficult to search for two colors at once? How eye movements can reveal the nature of representations during multi-target search. Journal of Experimental Psychology: Human Perception & Performance 2012; 38:113–122.

STROUD AND WHITBOURNE

19. Lustig C, Shah P, Seidler R, et al. Aging, training, and the brain: a review and future directions. Neuropsychology Review 2009; 19:504–522. 20. Baniqued P, Kranz M, Voss M, et al. Cognitive traning with casual video games: points to consider. Frontiers in Psychology 2014; 4:1–19. 21. Ballesteros S, Mayas J, Prieto A, et al. A randomized controlled trial of brain training with non-action video games in older adults: results of a 3-month follow-up. Frontiers in Aging Neuroscience 2015; 7:1–12.

Address correspondence to: Dr. Michael J. Stroud Department of Psychology Merrimack College Sullivan Hall 307F 315 Turnpike Street North Andover, MA 01845 E-mail: [email protected]