Training in real-time strategy games in older adults (Basak et al., 2008)


Grandpa Azuma from Yakitate Japan!!

To follow up my previous post, another study published by the same lab at the University of Illinois at Urbana-Champaign used the same experimental paradigm in assessing video games effect on cognitive abilities, but this time with elderly people. This study was widely publicized in many news media. This study is published in the journal Psychology and Aging. What I’m reading is the approved version and not the final version as I am still waiting for them to publish it online.


Declines in various cognitive abilities, particularly executive control functions, are observed in older adults. An important goal of cognitive training is to slow or reverse these age-related declines. However, opinion is divided in the literature regarding whether cognitive training can engender transfer to a variety of cognitive skills in older adults. In the current study, the authors trained older adults in a real-time strategy video game for 23.5 hr in an effort to improve their executive functions. A battery of cognitive tasks, including tasks of executive control and visuospatial skills, were assessed before, during, and after video-game training. The trainees improved significantly in the measures of game performance. They also improved significantly more than the control participants in executive control functions, such as task switching, working memory, visual short-term memory, and reasoning. Individual differences in changes in game performance were correlated with improvements in task switching. The study has implications for the enhancement of executive control processes of older adults.

I found several links to complement this post, from the university’s news report, an audio report, gamepolitics, Washington Post, Telegraph, and

Previous research on slowing down age-related cognitive impairments focused on specific abilities and it was noted that these tools did not transfer to other abilities rendering them less versatile. The researchers’ goal was to examine how individualized feedback, multi-tasking and prioritization of tasks would transfer to executive control and memory processes of older adults, that is whether more complex tools might help transfer the learned skills into other cognitive skills.


Participants: 40 participants, 20 in the experimental group where they played a real-time strategy game and 20 in the control group. Participants have no gaming experience (0 hour per week of play). Both groups are similar in terms of education, age (average age is 70, SD ≈ 5) with no notable differences in other areas. An observation is that the male-to-female ratio in both groups is 1 to 3.

One reason why they chose 70 year old participants is that there is observable cognitive impairment usually after the age of 60.


Since it’s the same psych lab from the last post, they used the same experimental paradigm. Easy for researchers and us to compare data between young adults and older adults if they ever want to make some comparisons in another study.

Operation span: Do math problems while trying to remember 3-6 words. Doesn’t tell how many trials.

Task switching: Because of a single word in the article, I got confused when they wrote “letter”, instead of “number”. Participants had to switch between two tasks, either to press a key whether the number is even or odd or whether the number is higher or lower than five. Which task they must perform depends on the background colour, the high/low task is blue while the odd/even task is pink. They also have to use one hand for one task and the other hand for the other task. Tell me if I lost you. There were 120 test trials for single tasks (just one task), then a practice trial for dual-task (switching between tasks), then a 160 dual-task trials.

N-back task: letters appear in different locations on a screen. They were asked if the letter was at the same location as the letter from 1 or 2 letters back. For example, “a” shown at top right, “b” shown at bottom left, “c” shown at top right (participants must a press key if this letter was the same location or not as from two letters ago, that would be a’s position). “d” shown at left, participant must give an answer by pressing a key, yes or no. There were 24 practice trials and 144 test trials.

Visual short-term memory: sets of 2, 4 or 6 coloured lines are shown with different orientations (horizontal, vertical, diagonal, etc.) and were shown again, either with one change (in colour or orientation) or no change. Sets of 4 coloured lines were considered into the analysis based on previous findings on older adults. Participants were asked if they saw differences. This is done with 16 practice trials with 96 test trials.

Ravens matrices: Wikipedia has an article on it. “In each test item, a candidate is asked to identify the missing segment required to complete a larger pattern. Many items are presented in the form of a 3×3 or 2×2 matrix, giving the test its name.”

Stopping task: new task, participants were to respond to specific letters as fast as possible. However, if they heard a tone or beep they were asked not to respond. The tone is presented 25% of trials and was initially presented  250 milliseconds delay after a target letter. If the participants successful inhibited their response, then the delay time is reduced, if unsuccessful the delay time is increased. Participants went through 224 trials.

Functional field of view test: Search for a white triangle in a circle among square distracters. Participants have 24 practice trials, and 120 test trials.

Attentional blink test: Watch a series of letters rapidly blinking, participants must identify the white letter and whether there was an x presented sometime after the white letter. 15 practice trials and 144 test trials.

Enumeration task: Brief presentation of dots and asked how many were there. 32 practice trials and 160 test trials.

Mental Rotation: Wikipedia has an article on it. Simply look at 3D tetris-like shape, look at another one and say if it’s the same object or not as soon as possible.  The object are simply rotated at specific degrees. There were 30 practice trials and 128 test trials.

Games used: Rise of Nation: Gold Edition. The game was chosen for its complexity and its requirements for multi-tasking in players and task switching (monitoring resources to directing battles or choosing what to research).


Participants were tested on their cognitive skills on their first week as a pre-test measure. The experimental group played RoN for 4 to 5 weeks or 15 gaming sessions of 1.5 hours which totals to 23.5 hours of play time. A second cognitive test was given on the fourth week of the study and a third and post-test was given on the last gaming week or a week after that. Not sure what to make of it.

Participants played RoN in Quick Battle or skirmish mode (i.e. play against a computer opponent in a match). A notable observation that could potentially affect results of the study is the difficulty setting of the game, it is set at easiest and stayed that way throughout the study. Given their abstract stating they found results, it is quite amazing they’ve able to find significant results with the game set at easiest. Nevertheless, it is logical to start at easiest when you’re a newbie player, based on my experiences with new games, starting in normal difficulty would certainly bring immediate defeat. I learned such lessons when playing Sins of a Solar Empire.

It would also be interesting to expand the study by factoring in a difficulty curve by increasing game difficulty as soon as participants achieve mastery (i.e. defeat enemy within a certain time limit). Something I observed myself playing Company of Heroes.


Each measure has some participants excluded from analysis due to errors from experimenters, participants or technical difficulties.

Game performance: participants’ score were statistically improved from pre-test to post-test. Interestingly and if I’m reading this right, their average time in the game went from 250.41 (SD = 211.09) seconds pre-test to 111.86 (SD = 58.26) seconds post-test. That’s short for a regular skirmish, my comfort zone would be 20-40 minutes, but they took less than 10 minutes (even taking into account those who played the longest). This is evidenced by the authors who noted that 17 of 20 participants replayed their game in order to complete their gaming session. Again, it would have been fun to see a difficulty curve in the study. But I would also stress that a higher difficulty would present more occurrences for task switching and a better mental exercise.

Operation span: No statistical differences were found between the experimental and control group.

Task switching: An interaction effect was found in the experimental group, there were no differences found at the pre-test and second cognitive tests, but there was a significant difference at post-test with a large effect size. So, after 23.5 hours of play, older adults who played RoN showed an improvement in reaction time on task switching.  However, such differences were not found for accuracy.

N-back task: Much like task switching. Statistical differences were found at post-test where the experimental group had a faster reaction time than the control group with a large effect size, however their accuracy were not different.

Visual short-term memory: Another time and group interaction, so the experimental group showed a significant difference being more accurate after 23.5 hours of playing with a medium effect size. However, the paragraph and the graph for this measure are sending me into a confused state. The graphs’ error bars are telling me that they’re not different. Hmm… a little help.

Ravens matrices: The same time and group interaction results. The experimental group showed an improvement and therefore are significantly different from the control group after 23.5 hours of play time with a large effect size.

Stopping task: No statistical differences were found between the experimental and control group.

Functional Field of View: No statistical differences were found between the experimental and control group. The authors talked about eccentricity, but this is all perception talk, so I have no clue what it means.

Attentional blink: No statistical differences were found between the experimental and control group. No difference in accuracy.

Enumeration: No statistical differences were found between the experimental and control group.

Mental rotation: No statistical differences were found between the experimental and control group for reaction time. However, a time and group interaction effect was found where the experimental group showed a statistical improvement in accuracy after 23.5 hours of play with a small effect size. This one is actually surprising since RTS games don’t involve much of visuospatial skills.  The authors speculate this might be because players must “perceive the relations between multiple objects and events in the video game.”

Additional analysis: they conducted a correlational analysis to see if game performance (game time) is correlated with the various cognitive abilities. They found a significant correlation between game performance and task switching, namely reaction time. There was also a significant correlation for the N-back task. IMO, no comment for the moment.


Surprisingly, I don’t have much to comment.

The authors argued that the nature of RTS games keeps players on their toes and to have them monitor, juggle, prioritize a lot of things, like units, resources, buildings, defences, etc. Such complexity requires multi-tasking or what the authors appropriately call variable priority training since many goals change in importance according to the situation and context (i.e. more on defence if you sense an incoming tank rush). Of course, all of it involves executive control.

The authors noted a limitation that the control group did not play a video game. Hmmm… I wouldn’t call that a limitation in a real world context. But they convincingly argued that they should tested more participants on other gaming genres.

They asked for more future research that relates to everyday cognitive abilities and real world tasks, such as driving, working, or sports. Maybe I should add, debating or essay writing. IMO, I’m thinking of raising the bar on game performance, let’s try Supreme Commander.

Basak, C., Boot, W. R., Voss, M. W., & Kramer, A. F. (2008) Can training in a real-time strategy video game attenuate cognitive decline in older adults? Psychology and Aging,23 (4), 756-777.


One thought on “Training in real-time strategy games in older adults (Basak et al., 2008)

  1. Pingback: Variations in time schedule and video games affect mental rotation performance (Cherney, 2008) « VG Researcher - Psychology

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s