Mass media effects are small and quantifiable. They do not generally turn good kids into school shooters or launch social movements, but they do have small effects on very large audiences. Due to the amount of people media messages can reach, even a small impact can have significant consequences.
For instance, let's say 10% of X-itis patients die and 10% of Rwandans that don't sleep under ITNs get X-itis. If the message you disseminate through the media reaches 1 million Rwandans and gets 5% of them to start using ITNs, you'd now have half as many cases of X-itis and half as many deaths from X-itis. If the original number was 250,000 cases of X-itis, you now have only 125,000. Rather than 25,000 deaths, you'd have 12,500, a difference of 12,500 lives.
Saving 12,500 lives is nice but how can we maximize this number?
Media interventions don't always outcompete interpersonal and other interventions. There are also multiple ways to use media for social change (entertainment-education, health communication, social marketing, farm radio show, etc.) and not that much known about which methods are the most effective. These methods are also based on divergent models of communication and social behaviour. Usually, many methods are used in unison and usually, this yields pretty good results.
I once made the analogy to a sculptor that can never see the entire sculpture from one vantage point. In order to make an aesthetic evaluation, she must circle the object and synthesize the input from multiple angles. Similarly, no theory simple enough to put into words will ever accurately describe the complex movements of cultural phenomena. It is less productive to try deriving a comprehensive theory than to utilize multiple perspectives.
In his incredible Family Tree of Theories, Methodologies, and Strategies in Development Communication, Waisbord proposes a "tool-kit" conception of strategies:
"Practitioners have realized that a multiplicity of strategies is needed to improve the quality of life of communities in developing countries. Rather than promoting specific theories and methodologies regardless of the problem at stake, there has been an emerging consensus that different techniques are appropriate in different contexts in order to deal with different problems and priorities. Theories and approaches are part of a “tool-kit” that is used according to different diagnoses. There is the belief that the tools that are used to support behaviour change depend on the context in which the program is implemented, the priorities of funders, and the needs of the communities."Waisbord also calls for the integration of top-down and bottom-up theories, personal and environmental approaches, and multimedia and interpersonal communication. Basically, we don't know exactly how to optimize our media use for saving the most lives possible but we know that media interventions are helpful and attacking problems from multiple angles is even more helpful. At the moment, I think it makes more sense to target according to health areas and geographic regions than it does to target according to media strategy.