Quick links Home page Staff members Contact details Research overview Research projects Technical reports

REDEEM: Creating Reusable Intelligent Tutoring Systems

Shaaron Ainsworth

Aims

The aims of this project are to:

• construct software tools (REDEEM) that allow existing computer-based instructional materials to be reused with the addition of adaptive teaching strategies
• to benefit from teachers' knowledge by allowing them to graphically describe approaches to course material, teaching strategies and their students' characteristics
• to exploit REDEEM as an ideal "research laboratory" in which to ask questions about the nature of teaching and learning
• to explore the nature of teachers' expertise and REDEEM's role in supporting professional development.
• REDEEM has been used in studies with primary and secondary school teachers (link) and with Naval Trainers (link) and is currently used in the School of Psychology to teach statistics

Read this page to find out about the system, follow this to read about the on-going research projects.

Theoretical Background

Intelligent tutoring systems have failed to gain widespead acceptance outside the laboratory. They are costly to build needing 100 - 1000 hours to produce an hour of instructional material. ITSs are unpopular in some quarters because most teach according to a a fixed, unalterable strategy. Teachers have requested the opportunity to be more involved in creating computer-based environments in their classroom.

These observations prompted the development of REDEEM. REDEEM is one of the easiest ITS authoring tools to use. It allows authors with little technical knowledge to create their own ITS from exisiting computer-based teaching material. They enhance this material by overlaying it with their own instructional knowledge producing richer environments that are designed for the needs of specific learners. REDEEM also incorporates default teaching knowledge supplied from research in the nature of learning and instruction, e.g. help is provided contingently

System Description

There are three major components to REDEEM: course material catalogues, a set of authoring tools and an intelligent tutoring shell.

Course Material

A number of course material packages (e.g. understanding shapes, poetry and history of world war 1, genetics) have been developed in conjunction with teachers. We can also use exisiting courseware and have done so in our ONR funded navy project

A page from the genetics domain

Authoring Tools

These tools allow teachers to describe aspects of the material, instructional strategies and their students. They can also specify under what conditions they would change the way the system teaches.

For each page in the courseware, teachers describe features using a number of dimensional ratings. For example, difficulty and familiarity of the material, the appropriateness of feedback and hints. Teachers can also provide questions, answers and hints. Pages can be related and combined into sections.

A similar approach is taken to the creation of instructional strategies. A graphical interface allows teachers to dimensionally scale and rank the critical aspects of a strategy (e.g. student control, teaching/testing ratio) and to choose a number of potential teaching and testing styles.

Students can be categorised into a set of teacher-defined categories (e.g. relative levels of numeracy) which enable the shell to decide which is the most appropriate initial teaching strategy. If the student category changes during a session, then this could result in a new teaching strategy.

Intelligent Tutoring Shell

The ITS shell will deliver the courseware to students. Its main roles are

• choose the appropriate part of the course to present to the student
• model students' knowledge
• offer questions and monitor student's performance
• provide help
• support different teaching strategies
• provide feedback to the teacher on a student's performance

On-Going Research

REDEEM has been found to be time efficient and usable. Teachers can be trained to use REDEEM in around 90 minutes. They create ITSs from existing CBT at around 3 hours per hour of instruction.

Two further projects are now exploring the use of REDEEM in schools, universities and naval training.

REDEEM's use in the school classroom has addressed issues such as the role of REDEEM in teacher professional development, the similarities and differences in the way that four different educators authored the same courseware and currently is exploring the additional educational benefits that REDEEM provides over underlying CBT. Details on this research can be found here.

REDEEM can also be used in adult training. The Office of Naval Research has funded a three year project to explore the use of REDEEM with Royal Navy trainers. This research looks at the use of REDEEM with both reservists and full time personnel. Further details can be found here.

In five studies, learning outcomes with students learning with REDEEM have been compared to CBT. On average, REDEEM has led to a 29.4% increases whereas CBT has led to 20.4% increase from pre to post-test. An overall effect size analysis suggests that REDEEM improves learning by about the same amount as non-expert human tutors do compared to classroom teaching (REDEEM/CBT = 0.59 sigmas, human tutor = 0.4 (non expert) to 2.0 (expert))

Sadly, REDEEM is not yet available for public use. If you would like to be informed if it is made available, you can email Shaaron Ainsworth.

Selected Publications

Ainsworth, S.E. & Grimshaw, S.K. (2004) Evaluating the REDEEM authoring tool: Can teachers create effective learning environments? International Journal of Artificial Intelligence in Education 14(3/4), 279-312.
Ainsworth, S.E., Major, N., Grimshaw, S.K., Hayes, M., Underwood, J.D., Williams, B. & Wood, D.J. (2003). REDEEM: Simple Intelligent Tutoring Systems From Usable Tools in T. Murray, S. Blessing & S.E. Ainsworth (2003). Advanced Tools for Advanced Technology Learning Environments. pp. 205-232. Amsterdam: Kluwer Academic Publishers
Ainsworth, S.E., Underwood, J.D. & Grimshaw, S.K. (1999) Formatively Evaluating REDEEM: An authoring environment for Intelligent Tutoring Systems. In Lajoi, S. & Vivet, M. (Eds) Artificial Intelligence in Education Open Learning Environments: New Computational technologies to support learning, exploration and collaboration, pp93-100. Amersterdam: IOS press.
Major, N., Ainsworth, S.E., & Wood, D.J. (1997) REDEEM: Exploiting Symbiosis Between Psychology and Authoring Environments, International Journal of Artificial Intelligence in Education 8(3/4), 317-340.
Major, N & Ainsworth, S.E. (1997) Developing ITSs using a psychologically motivated authoring environment, AAAI Fall Symposium, (pp 53-59), AAAI Press: Menlo Park, CA.