Call Us: US - +1 845 478 5244 | UK - +44 20 7193 7850 | AUS - +61 2 8005 4826

evidence for alternative accountability measures

A large-scale experiment evaluated the benefits of intelligent tutoring in an urban high school (Koedinger, Anderson, Hadley, and Mark, 1997). Researchers compared achievement levels of ninth-grade students who received the PUMP curriculum, which is supported by an intelligent tutor, the PUMP Algebra Tutor (PAT) (experimental group), with those of students who received more traditional algebra instruction (control group).* The results, presented below, demonstrate strong learning benefits from using the curriculum that included the intelligent tutoring program.

The researchers did not collect baseline data to ensure similar starting achievement levels across experimental and control groups. However, they report that the groups were similar in terms of demographics. In addition, they looked at students’ mathematics grades in the previous school year to check for differences in students’ prior knowledge that would put the experimental group at an advantage. In fact, the average prior grades for the experimental group were lower than those for the control group.

  • The researchers note that their research strategy is first to establish the success of the whole package and then to examine the effects of the curriculum and intelligent tutoring components independently; this work is still to be finished.

On the other hand, some research suggests that the relationship between formative assessment and cognitive theory can be more complex. In a study of Anderson’s geometry tutor with high school students and their teachers, Schofield and colleagues found that teachers provided more articulate and better-tuned feedback than did the intelligent tutor Nevertheless, students preferred tutor-based to traditional instruction, not for the reasons one might expect, but because the tutor helped teachers tune their assistance to problems signaled by a student’s interaction with the tutor. Thus, student interactions with the tutor