Motivation: Compare the effectiveness of human and computer tutors

Definitions:

• Computer Tutoring Systems
  • Intelligent tutoring – step-based or substep-based, students work problems in the system and receive feedback for each step, high interactivity
  • Answer-based tutoring (Computer-Aided Instruction (CAI), Computer-Based Training (CBT), Computer-Aided Learning (CAL)) – answer-based, students work problems outside of the system and enter the answer to receive feedback, low interactivity
• Human Tutoring – students work on problems with subject-matter experts synchronously, high interactivity

Advantages of Human Tutoring: VahLehn discusses several theoretical advantages of human tutors over computer tutors, but most of them are not supported by research because they overestimate human tutors or underestimate computer tutors. He argues that the advantages of human tutors are that they provide more formative feedback, which helps students fix flaws, and scaffolding, which guides students further down a line of reasoning.

Outcome: Compared to no tutoring, effect size of answer-based computer tutors is d = .3, intelligent computer tutors is d = .76; and human tutors is d = .79. VanLehn explains these results by arguing that though computer tutors can’t elicit interactive behaviors as well as human tutors, they can elicit constructive behaviors. According to Chi’s (2009) ICAP framework, which predicts interactive >= constructive > active > passive learning, interactive and constructive behaviors are equally effective.

Why this is important: This review suggests that computer tutors and human tutors can be equally effective, though through different processes. Both of these findings are valuable to educational technology designers because they help us argue for the value of computer tutors and recognize the strengths of human and computer tutors. Furthermore, VanLehn posits the Interaction Granularity Hypothesis, which states that the key difference between feedback and scaffolding of human and computer tutors is the granularity of interaction. The larger the granularity, the more reasoning that is required of the learner between interactions. Both human and computer tutors must strike the correct granularity to most effectively support learning.

VanLehn, K. A. (2011). The relative effectiveness of human tutoring, intelligent tutoring systems, and other tutoring systems. Educational Psychologist, 46(4), 197-221. doi:10.1080/00461520.2011.611369

Chi, M. T. H. (2009). Active-Constructive-Interactive: A Conceptual Framework for Differentiating Learning Activities. Topics in Cognitive Science, 1(1), 73-105. doi:10.1111/j.1756-8765.2008.01005.x

For more information about the article summary series or more article summary posts, visit the article summary series introduction.