In the context of the water jars experiment, "attitude" is not the explanation though; instead it refers more to the "installation" of the previous solution in the subjects' minds (i.e. the subjects who were "trained" with the 5 pre-test examples effectively had those solutions installed/set in their minds).

You see this in game development a lot- users are trained in the first 1-2 levels "Here's a puzzle - look around for clues". When they later encounter a puzzle in levels 3-n, they often apply the same methodology despite the obviousness that another approach is called for.

You see it in chess development - an app teaches users, "Given a board with these positions, the best option is x". Later during growth, the same board positions may have different solutions yet the student continually applies the original.

Attitude isn't at play here; it's simply expectation.