Affective support in AutoTutor benefits low-knowledge learners but not high-knowledge learners, and gaze-based disengagement reactivity yields significant gains only on deep learning¶

Claim · All claims
Evidence · 1 study (2 entries) · review r2–r3 · q2 · i2 medium
Currency · newest study 2014 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i2 The Supportive AutoTutor was more effective for low knowledge students (0.71 σ gain on posttest scores) than standard AutoTutor, but only on a second session; high knowledge students did not benefit and sometimes showed lower learning. → Nye 2014 q2 i1 GazeTutor produced small, not statistically significant overall learning gains (0.26σ, p<.178) and moderate, statistically significant gains on deep learning (0.45 σ, p<.035). → Nye 2014 (2)

Evidence¶

Nye 2014¶

Nye, B. D., Graesser, A. C., & Hu, X. (2014). AutoTutor and Family: A Review of 17 Years of Natural Language Tutoring. International Journal of Artificial Intelligence in Education. https://doi.org/10.1007/s40593-014-0029-5

q2 · i2 · review · r2

Review reports D'Mello and Graesser's comparison of Supportive versus standard (neutral) AutoTutor. Printed effect size: 0.71 σ for low-knowledge students, evident only on a second session; a buddy-style Shakeup Tutor showed no positive benefits.

"The Supportive AutoTutor was more effective for low knowledge students (0.71 σ gain on posttest scores) when compared to standard AutoTutor, but such gains were only evident on a second session when student difficulty was more pronounced"

Nye 2014 (2)¶

Nye, B. D., Graesser, A. C., & Hu, X. (2014). AutoTutor and Family: A Review of 17 Years of Natural Language Tutoring. International Journal of Artificial Intelligence in Education. https://doi.org/10.1007/s40593-014-0029-5

q2 · i2 · review · r3

Review reports the GazeTutor evaluation using camera-captured eye movements to react to disengagement. Printed effect sizes: 0.26σ overall (not significant, p<.178) and 0.45 σ on deep learning (significant, p<.035).

"GazeTutor succeeded at increasing attention to the tutor and resulted in small, but not statistically significant overall learning gains (0.26σ, p<.178) and moderate, statistically significant gains on deep learning (0.45 σ, p<.0 35)"

Discussion¶

Related Claims¶

• The talking head contributes little to AutoTutor's learning gains; the content of the conversation matters — related
• Early AutoTutor versions gained about 0.5 σ over textbook controls, with gains depending on test type (cloze 0.64 σ, deep 0.28 σ, shallow 0.15 σ) — related
• Meta-analytic evidence indicates intelligent tutoring systems yield positive learning outcomes compared with non-ITS conditions — a broader claim this one bears on
• Young children's attention to a robot increased sharply when the robot reacted to their gaze — related