When is Reading Just as Effective as One-on-One Interactive Human Tutoring?

Many human tutorial dialogues have a similar format (Graesser, Person & Magliano, 1995) wherein the tutor poses a question, the student answers it, and they work collaboratively to improve the student’s answer. We hypothesized that the well-known effectiveness of human tutoring was due to the last phase, and in particular, to the interactive nature of the tutorial dialogue there. To test the hypothesis, we compared tutoring which held constant the first two steps, the question and the student’s initial answer, and varied whether the student engaged in tutorial dialogue or simply read an explanation of how to derive a correct answer. To our surprise, the initial experiment showed equal learning gains. Two further experiments painted a confusing picture, but it is safe to say that interactive tutoring is not always superior to reading.