By C. Brandon Chapman, College of Education
PULLMAN, Wash. – Recent research shows that when a class size becomes large enough, one-on-one computer-based tutoring is more effective than traditional teacher-based instruction.
The instructional programs are called intelligent tutoring systems (ITS). They’ve been studied many times since the first well-recognized ITS in 1970.
But in a comprehensive statistical analysis of more than 100 individual studies, researchers at Washington State University and Simon Fraser University found much more reliability in their results than any one study found individually. Their work was published in the Journal of Educational Psychology in November (http://psycnet.apa.org/index.cfm?fa=search.displayrecord&uid=2014-25074-001).
That meta-analysis shows that with small, more individualized groups, computer programs and human tutors/teachers yield results that are statistically about equal. But as the group size grows, teacher effectiveness shrinks.
Less costly instruction, based on feedback
Olusola Adesope, a researcher and associate professor of educational psychology at WSU’s College of Education, said the genius of ITS is that they have highly individualized task selection and give prompting and response feedback.
“ITS can successfully complement and substitute for other instructional modes, and these situations exist at all educational levels and in many common academic subjects,” he said.
ITS can model the state of the learner by asking questions to determine level of understanding. They then figure out the best way to provide individualized instruction to help improve learning deficiencies, whether those stem from cognitive or motivational gaps.
A teacher can often get to the root of a student’s issues in order to be an effective tutor. But such individualized instruction is costly.
“Back in 1984 an educational psychology researcher, Benjamin Bloom, published a classic paper showing that an average student who receives one-on-one tutorial instruction is ahead of 98 percent of those who receive conventional instruction,” Adesope said. “However, it is practically impossible to have a single teacher for each student, considering the huge cost implications.”
Difficult research worth the effort
The team included Adesope and three researchers from Simon Fraser University: Wenting Ma, John Nesbit and Qing Liu. They examined 107 previous studies conducted over the past 40 years with 14,321 students.
“The most difficult part of this research was dealing with a large number of studies with a limited budget,” Adesope said. “Our initial search returned over 26,000 studies. Although the large number of studies was daunting, we were encouraged to finish the project considering the potential impact.”
Because of the potential impact, the team was able to secure most of its funding through a grant from the Social Sciences and Humanities Research Council of Canada. For additional funding, such as for Adesope’s methodology, the team received a WSU College of Education Faculty Funding Award.
“It was so much fun working together with other colleagues on this project,” Adesope said. “It took a few years to complete, so we all needed to stay together – which we did well in spite of the distance between the collaborators. Everyone contributed intellectual uniqueness to the success of the project.”
Costs and benefits to be examined
Even after a comprehensive meta-analysis, with solid results, Adesope said the work on this project is far from finished.
“We hope this research will engender serious discussions from policy-makers and educational stakeholders,” he said.
Those discussions have already begun. He said scholars have expressed their desire to learn more about how these results will factor into future research and educational technology entrepreneurship.
“Currently, there is limited use of ITS in the classroom,” he said. “It will be interesting for us to work with faculty in computer science, business and economics to examine the cost implications in relation to the benefits.”
Additionally, Adesope said he plans to examine cost-benefit analyses of ITS and their scalability for the classroom.