Now I don't understand how the experiment works at all. You can't make the choice for them; and you can't be sure their choices are random. They may have preferences before the experiment, based on whatever. The conclusion that the act of choosing changes their preference seems impossible to prove.
Per an earlier study (which may or may not have weathered the replication crisis, I'm not sure) they proved it for adults with anterograde amnesia (inability to form new memories).
If memory serves:
They had patients sort paintings by preference, told them they could have a copy of whatever they ranked 2nd or 3rd (of course they'd pick the one they ranked 2nd), told them (falsly) they'd be mailed their copy, came back in a week and had them rank the same paintings again. More often than should be explained by chance, patients would rank what had been 2 as 1 or what had been 3 as 4 or 5; their preferences for these paintings changed in the direction of their choice, despite choosing freely and even though they didn't remember making the decision.
That sounds like a lousy thing to do to someone with amnesia. I hope they got copies of paintings in the end, even if they don't remember how they got them.
Heh, I'm really undecided on the morality of inconsequentially lying to someone who can't remember it, but I certainly hope they were appropriately rewarded for their participation in the study whether or not that includes paintings.
