”Mere Vehicles”: A White Paper

In 1983, Richard Clark called for educational researchers to reconsider their approach and made the case that media are “mere vehicles” and do not influence learning in any form (Clark, 1983). Clark based this statement on the fact that meta-analyses failed to show a significant difference in media and outcomes, noting that any studies that find causal connections between media and achievement are confounded. He called for researchers to focus on instructional method instead of media. Many others, often reported as led by Kozma, called into question Clark’s “mere vehicles” claim, but were unable to refute his assertion that media does not affect achievement with the support of research (Kozma, 1994; Reiser, 1994). Now in the 21st century, technology and media have advanced exponentially beyond what was available 30 years ago when Clark first made his case. Despite the advances in technology and media, no support exists to validate the case of Kozma (Sung & Mayer, 2013). Although, many dispute Clark’s claim, empirical evidence has yet to disprove it (Hastings & Tracey, 2005; Kozma, 1994; Reiser, 1994; Tamim, Bernard, Borokhovski, Abrami, & Schmid, 2011).

In 1983, Richard E. Clark wrote a now famous (or infamous) article calling for educational researchers to cease in the efforts to demonstrate the effectiveness of media in regards to academic achievement (Clark, 1983). He states that “The best current evidence is that media are mere vehicles that deliver instruction but do not influence student achievement any more than the truck that delivers our groceries causes changes in nutrition” (Clark, 1983, p. 445). Clark admits that he is not the first to make the claim of “no learning benefits from media,” but he is the first to state it decisively without leaving the door open for media to have an effect (Clark, 1994). The “mere vehicles” claim was based upon a series of meta-analyses of media comparisons conducted by James Kulik and his colleagues, which demonstrated that the positive effect for media more or less disappears when the other cofounders are accounted for (Kulik, Kulik, & Cohen, 1980). Additionally, he remarks that there is some evidence for selection bias in the fact that journal editors typically select research that find stronger effect sizes for newer media. Clark argues that “All methods required for learning can be delivered by a variety of media and media attributes. It is method which is the ‘active ingredient’ or active independent variable that may or may not be delivered by the medium to influence learning” (Clark, 1994, p. 26). He states that media may influence cost or speed of learning but methods are causal in learning (Clark, 1994). Richard Clark encouraged researchers to focus their efforts on instructional method and not media.

In 1994, Clark resumes his argument by debating his “mere vehicles” claim with other experts in an edition of Educational Technology Research and Development that focused on this debate (Clark, 1994). He defends his position as he has for the past 11 years, and states that although technology and media have advanced, there continues to be no evidence to support the effects of media on achievement. He adds an analogy to his mere vehicles argument, this time, likening media to the way medicine is prescribed. The active ingredient in medicine determines the results in the patient regardless of the method of delivery (Clark, 1994). He expresses that we should stop trying to identify media attributes that are “sufficient” for learning to take place and focus on instructional methods that are “necessary” for learning. Clark takes an even stronger stance in 1994, entitling his article, “Media Will Never Influence Learning.”

Although many refer to this as the Clark and Kozma debate, there were many others who opposed Clark’s claims (Becker, 2010; Hastings & Tracey, 2005; Kozma, 1994; Reiser, 1994; Tamim et al., 2011). Simonson (1987) points out that a “howl of anger” was almost immediately heard from many media professionals. The opposing view to Clark’s is that media definitely influences learning, and that it is so intertwined you cannot effectively separate media from learning (Kozma, 1994). Kozma stated that “Perhaps the appropriate question is not ‘do’ but ‘will’ media influence learning” (Kozma, 1994, p. 2).

The theoretical framework of both Clark’s metaphors is that the learner is a passive recipient of the information, but Kozma argues that learning is an active, constructive, cognitive and social process by which the learner is actively engaged (Kozma, 1994). Reiser points out that successful delivery of the instructional method is dependent upon the attributes of the medium and some mediums make certain instructional methods possible (Reiser, 1994). Additionally, both Kozma and Reiser point out the flaws in Clarks metaphors: arguing that administration of certain medication is less effective or not possible without the appropriate delivery medium (Reiser, 1994).

Richard Clark argued that media attributes are interchangeable and, if they can be accomplished by another medium, these attributes are not specific to the medium itself (Clark, 1983, 1994). Kozma, however, states that a distinction must be made between attributes and variability. Attribute as a property is distinguished by the quality that is present, while the variability represent the degree to which an attribute is present. The attributes of a medium are its capabilities regardless if that capability is used (Kozma, 1994). Although many media share certain attributes it is the cluster of those attributes that define the media, not the individual attributes themselves.

Kozma contends that scientist interested in eliminating something undesirable are concerned with necessary, and scientist interesting in the production of something desirable are concerned with the conditions sufficient to bring it about. Since educational research is concerned with the production of learning in subjects, researchers must look for sufficient conditions in theory, research and design (Kozma, 1994). Kozma argues that media theory and research must reflect both the capabilities of media and the complexities of the social situations within which they are used.

The majority of the Clark-Kozma debate took place in the 1980s and 1990s; since that time technology and media have taken massive leaps and bounds forward. Becker (2010) has described using research from the time of the debate may be equated to discussing today’s traffic issues using data from 1820. Hasting argues that the technological advances have only strengthened Kozma’s position, with decreased cost of media and an everincreasing array of attributes. She states that if we were to still use Clark’s vehicle metaphor our truck would now be a supersonic jet delivering instruction faster, fresher, and less expensively than was ever conceived of in 1983 (Hastings & Tracey, 2005). They argue focus of the debate should be shifted to technologies that are either new or unrecognizable to those in 1983, primarily computers/Internet and serious gaming (Hastings & Tracey, 2005). In regards to computer/Internet media it is argued that there are no replaceable media for having much of the world’s knowledge available to the learner at any time. Additionally, the speed at which learning occurs is at a much faster pace. What used to take someone hours or days to find can be accomplished in seconds with a few key strokes (Hastings & Tracey, 2005). It must be acknowledged that these arguments of ease and speed and likening a car to a supersonic jet do not disprove Clark’s statement. His primary assertion was that media may affect speed or cost but do not affect learning, as that is due to instructional method (Clark, 1983). In the case of serious games or simulations it is argued that learners are able to practice a skill that is not possible in any other medium (due to cost or risk) such as in the space program (Aldrich, 2009). With the media of serious games learners are able to actively learn and the game is able to respond to their learning, this is reported to be not available in other media (Becker, 2010).

Richard Clark advised us to base our conclusions on empirical evidence and not on opinion. Sung and Mayer designed an experiment in 2013 to test Clark’s hypothesis, and found that instructional method determined performance and media had an effect on willingness and engagement, but not on learning (Sung & Mayer, 2013). Tamim performed a second-order metaanalysis of the impact of technology over the past 40 years, and found only a small effect toward media, which is consistent with Clark’s claims in 1983. Despite the vehement disagreement and the massive technological advances that have occurred over the past three decades, no one has been able to disprove Clark’s original claim that media does not affect learning achievement (Kozma, 1994; Means, Toyama, Murphy, Bakia, & Jones, 2009; Sung & Mayer, 2013; Tamim et al., 2011).