What does interest have to do with the article and the experiments performed? Seems to me that the performance of the test subject depended upon the type of pr

What does interest have to do with the article and the experiments performed? Seems to me that the performance of the test subject depended upon the type of praise given, not the interest of the subjects (which would have been consistently random across test groups)