If you ever like to think of yourself as some kind of agent whose decisions are controlled by a rational ego (instead of some random words you heard once upon a time), you might find that a bit worrying.

Or I might not. You just finished suggesting that the way people treat me in the future will be based in part on my reputation for being cooperative. If a sociological study frames its questions in ways that make them sound correlated with real-life situations in which it's socially demanded that I be cooperative, why would I want to defect and risk my reputation? For sociological study points? Even if I'm told that my results will remain anonymous, "do bad things when you're pretty sure you won't get caught" isn't exactly a winning life strategy either.

It's also possible that for some people, "number of times I've cooperated when I know it's socially desirable to do so" is part of their utility function, independent of whatever other utility they receive from the act. "Being good for the sake of being good" certainly isn't an unheard-of philosophy in the general population, and I'd expect this group to be greatly overrepresented in a typical study of people who volunteer to participate in studies.