PT142 · S1 · Q21 In a study, pairs of trained dogs

We're trying to explain the behavior of the dogs who stopped obeying, but all the dogs in the study obeyed the order (by that I mean that we are told "after both obeyed, one was given a treat and the other wasn't"). So since the behavior we're trying to explain or interpret comes after both dogs have obeyed, it doesn't really matter which dog was more inclined to obey in the first place. They both obeyed. Since they both obeyed, we wouldn't know how to use this answer choice to judge what happened next. Which dogs were the ones who weren't given a treat? Dogs who were accustomed to receiving regular rewards, the other dogs, a mix of both? Since we don't know which type of dogs were given the treat vs. not given the treat, we can't use this answer as a possible alternate explanation for anything.

Whether we say, "YES, at least one dog got to be in the treat group and the no-treat group" or whether we say "NO, dogs were either assigned to the treat group or the no-treat group for all the trials", it doesn't make a difference. It's not an indictment of a study if you take your initial pool of (hopefully representative) subjects and then randomly assign half of them to one group and the other half to the other group. That is how respectable medical clinical trials work (half the group gets the drug / half gets a placebo). So the NO answer does not weaken. The YES answer might feel like it weakens if we were to think, "if a dog was previously given treats and now is in a trial where they're not being given treats, couldn't this be an alternate explanation for why they disobey? They're not mad that they're being treated unfairly; they're mad that they're no longer getting a treat like they used to." Two things: 1) "treated unfairly" is broad enough that it could actually still apply to that scenario -- the dog who used to be given treats and now isn't feels like it's being treated unfairly (not in respect to the other dog paired with it in the trial, but in respect to how it got used to being treated in the previous trial ... "No fair! You used to give me treats for obeying.") 2) the quantity is super weak. Since the question is posed as "were dogs who were given treats in one trial ever given no-treats in a different trial", this is only talking about at least one dog. It also doesn't specify any chronological sequence, for whether this dog who played both roles was a treat-received dog or a treat-denied dog first. Thus, it's really hard to get an Alternate Explanation answer out of YES that has much punch or that is clearly different from "That's unfair! I feel cheated"

If we were thinking about an alternate explanation like, "It's not unfairness that's leading to disobeying; it's just the lack of a treat.", then we might be attracted to this answer. After all, it seems to deal with the connection between getting a treat and obeying (not the connection between unfair treatment and disobeying). But if we say "YES, there was at least one case in which a dog who was getting a reward became more inclined to obey the command", that isn't strongly suggesting an alternate explanation. And if we say, "NO, none of the dogs given a reward became extra inclined to obey the command", that also isn't any sort of objection.

We aren't going to learn anything about what caused the decline in obedience among the unrewarded dogs by learning the precise number of reps before they began to disobey. It's not like we can say, "If it only took 5 repetitions, then it wasn't about unfairness." There's not going to be any common sense link between how many repetitions it took and what the underlying cause was for disobeying.