Graesser, A. C., Person, N. K., & Magliano, J. P. (1995). Collaborative dialogue patterns in naturalistic one-on-one tutoring. Applied Cognitive Psychology, vol. 9, pp495-522.

@Article{,
  author = 	 {A. C. Graesser, N. K. Person, J. P. Magliano},
  title = 	 {Collaborative dialogue patterns in
naturalistic one-on-one tutoring},
  journal = 	 {Applied Cognitive Psychology},
  year = 	 {1995},
  OPTvolume = 	 {9},
  OPTpages = 	 {495--522},
}

Author of the summary: Jim R. Davies, 2000, jim@jimdavies.org

Cite this paper for:

• tutoring is superior to classroom learning
• Evidence against active student learning
• tutors, not students, ask most questions
• most student questions do not address knowledge deficits but confirm what they already know [p501]
• amount of knowledge deficit questions did not correlate with performance.
• Sophisticated teaching strategies like the socratic method were non-existant with unskilled tutors [p502]
• Tutors typically use curriculum scripts
• tutors typically coach students through example problems or collaboratively solve them.
• Tutors used almost no authentic anchored cases but only symbolic examples or concrete referents. [p503]
• Conversational turns were 5 in the research methods domain, 10 in the algebra domain. Typical classroom conversations have 3 turns. [p504]
• Successive revision of questions on the tutor's part get easier to answer
• 34% of neutral feedback (on audio) turned out to be positive or negative when the video was looked at [p507]
• Fox (1991) reported skilled tutors use silence to signify negative feedback
• pumping: giving positive or neutral feedback to get a student to refine an answer
• Deep reasoning questions are more prevelant in tutoring [p511]
• Answering "yes" to a "do you understand?" question is not correlated to performance. Answering "no" is positively correlated. [p512]
• Neither frequency of student questions nor knowledge deficit questions correlated with performance, but freq of deep reasoning questions did.
• Modest correlation between completely correct answers to tutor questions and performance [p513]
• It is very difficult for tutors to identify bugs. Here ITS systems may outperform real tutors. [p514]
• "Tutors adopt a gentle indirect stance when students commit errors." unlike Anderson's LISP tutor. This may be good for motivation.

Summary author's notes:

• none