If you’re here, you’ve probably seen articles that talk about scientific studies of games and gaming. In public health we want to study and protect the health of populations. We do this through studies that emphasize validity, which basically means a lack of bias (when what we see doesn’t reflect the real world).
Because we pay such careful attention to validity, we can be pretty sure about how well our studies explain real-world relationships between causes and outcomes, who our conclusions apply to, and what the limitations are of our conclusions. We have to do things this way because the mathematical models—the tools of probability that we use to draw conclusions—assume that we’ve done it.
In an ideal world we would always use random samples to make sure that we’re looking at people who represent the population we want to model – i.e., use math to draw conclusions about. If we can’t use random samples, we at least try to use representative samples, but this makes it harder to draw valid conclusions. Think about the word modeling as it applies to drawing. If you wanted to draw a meaningful portrait of, say, an elderly woman, you wouldn’t use a 20-something year old man as a model. If you drew him, you wouldn’t capture the lines that show the elderly woman’s experience, or the facial expressions that show how she feels (we’d say you don’t have the ability to generalize from one to the other, which is external validity). If the younger man has lines, they might be the result of some kind of badass scar, acne or an eyebrow piercing (this is internal validity: his lines are more likely scars from a bear fight because he’s a man—sexist, yet more probable in the real world). The elderly woman might have badass scars too, but they’re not as likely to be from a bear fight.
Any scars more likely to be due to bear fight
How does all of the above apply to video games?
There is a ton of research on games. Some of it uses representative samples, like the Monitor Internet and Youth study in the Netherlands, which is an annual survey given to thousands of schoolchildren. Most video game research uses volunteers, e.g. hundreds of MMORPG players who fill out surveys on MMO websites (too many to link!). Sometimes a few gamers get MRI scans, and once or twice hundreds of thousands of players had their in-game behavior studied. The upshot is, different studies have different amounts and types of bias; all studies have bias; and it’s up to us to determine what those biases are and how much they affect what we observe in studies vs. what’s true in real life. In this blog I’ll talk about talk about truth and bias in research about video games and health, including violence prevention and problematic gaming/game addiction. Hopefully I’ll get your input too.
The motto of my undergrad university is a biblical quote from the Gospel of John: Veritas liberabit vos: The truth shall set you free. When we’re talking about public health, a better motto is Pilate’s response: Quid est veritas? —What is truth?
Model: Daniel Henney