Students tutored by an xPST-created web statistics tutor (xSTAT) scored higher on exam transfer problems than untutored students¶

Claim · All claims
Evidence · 1 study · causal r2 · q3 · i2 medium
Currency · newest study 2015 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i2 Randomly assigned students using the xSTAT ITS averaged 74% versus 62% on transfer problems at final exam time, a significant difference with d=0.62. → Stephen B. Gilbert 2015

Evidence¶

Stephen B. Gilbert 2015¶

Stephen B. Gilbert, Stephen B. Blessing, Enruo Guo. (2015). Authoring Effective Embedded Tutors: An Overview of the Extensible Problem Specific Tutor (xPST) System. Int J Artif Intell Educ 25:428–454. https://doi.org/10.1007/s40593-015-0045-0

q3 · i2 · causal · r2

Randomized comparison in the xSTAT Web xPST statistics tutor study: 24 students used the ITS-enabled web problems and 25 did the same multi-step real-data problems without ITS feedback. The quote reports the final-exam transfer difference, "a significant difference ( t(47)=2.03, p<.05, d=0.62)".

"The non-ITS condition averaged 62 % ( SD=6.99) and the students with tutored instruction averaged 74 % ( SD=4.27) on these problems, a significant difference ( t(47)=2.03, p<.05, d=0.62)."

Discussion¶

Related Claims¶

• Students tutored by xSTAT reported higher satisfaction than untutored students — related
• Five non-programmers authoring with final Web xPST produced 74 effective statistics tutors in a month, with coding a small share of total authoring time — related
• Ten non-programmer graduate students quickly learned early Web xPST and produced mostly high-quality tutors, with authoring time dropping across successive tasks — related
• In a head-to-head comparison, xPST and CTAT both let non-programmers create high-caliber tutors, with no significant time differences by the third problem — related