Hybridized Online Learning – More (But Could be Even More?) Effective

by ajc

Benjamin S. Bloom has two works from which to draw when interpreting the results of a new study by Carnegie Mellon. The key finding, related to the efficiency of this design, is summarized in the following quotation.

By combining the open-learning software with two weekly 50-minute class sessions in an intro-level statistics course, they found that they could get students to learn the same amount of material in half the time.

Essentially, the inclusion of this intelligent tutoring system allows the professor to discuss more nuanced and/or complex aspects of the content, and do so in one less class period. Bloom might infer that the observed improvements were due to an increase in the amount of time spent analyzing, synthesizing, and/or evaluating, i.e., at the higher levels of “his” taxonomy, while in class. Alternatively, he could point to his work on tutoring, and its corresponding “2 Sigma Problem”, which suggests that one can expect to observe a difference of +2 standard deviations when students work with a teacher in a one-to-one setting.

I’m inclined to agree with both of these hypothetical conclusions. I’m also reminded of Ewan McIntosh’s recent response to Will Richardson’s post commenting on an image of a lecture hall filled with (mostly Apple) computers. McIntosh’s critique, in he differentiates between curriculum and pedagogy, noting that teachers can control pedagogy but not curriculum, culminates with the following assertion.

The reason the picture presents a dubious message is that neither curriculum nor pedagogy have changed an iota in this learning space: it’s about the same layout – with as many apples on laps – as a Victorian classroom would have appeared.

I wonder how instructional design fits, as Carnegie Mellon’s design offloads the mundane, didactic portions of instruction to technological entities, thus freeing up space and time for the instructor/professor to do the sorts of things that are much harder for computers, even “intelligent” systems, to replicate. This is a good start, but can we go further?

The program’s efforts to maximize the contributions of technology are impressive. They’ve applied adaptive algorithms similar to those that are used in the GRE, which monitor and subsequently respond to students understanding. This is one of the first times this technology has been used outside of the assessment arena. (The article refers to its use as “novel”.) What needs to be considered, however, is a design that manufactures or generates time for face-to-face. That is to say that although the tutoring aspect of Bloom’s research is (partially) present in the form of the intelligent tutoring system, the potential of human tutoring is greater.

This is not the first time that I’ve considered using technology to offload the sort of lower level tasks of instruction. My presentation, entitled “A Shift in Focus: Designing for Face-to-Face” can be found here. Comments/critiques are welcome.