A careful review of the arguments and counter arguments presented by Clark (1983; 1994) and Kozma (1991; 1994), responses published in the past 20 years (Jonassen, Campbell & Davidson, 1994; Morrison, 1994; Reiser, 1994; Shrock, 1994) and existing instructional design literature (Morrison, Ross & Kemp, 2001; Reiser & Dick, 1996; Smith & Ragan, 1999) indicates there is, and always has been, significantly more agreement on this subject than the debate would indicate.
Clark never said that a textbook could deliver an instructional method requiring the use of a 3-dimensional graphic representation as effectively as a computer, nor did Kozma maintain that the computer was the only medium with the capabilities to do so. Both acknowledged that the two instructional components — the instructional methods and the delivery medium — must be aligned to facilitate learning.
The debate is, and always has been, about the ability of more than one medium to support a selected instructional method, whether or not any given medium has capabilities that cannot be replicated by another medium, and the validity of the research. We believe that today, in 2005: • Computers are capable of supporting instructional methods that other media are not • Computers, by means of their unique capabilities, affect learning • Computers are often the most cost-effective, efficient delivery method for any given unit of instruction We also: • Acknowledge the limitations of media comparison studies • Acknowledge the need to align the message, the medium and the learning task • Agree that some media are interchangeable and • Support the use of the most cost-effective, efficient delivery method for any given unit of instruction We believe that after 22 years it is time to reframe the original debate to ask, not if, but how media affects learning. We agree that media comparison studies are inherently flawed and support the argument that we must identify research designs that will provide answers to this question in significantly less time.