The mass media can be the major influence on the adoption of an idea after the general public begins to become aware. Initially, an idea spreads through limited social networks, with members exchanging information in formats ranging from face-to-face discussions to blogs on the Internet. Such interactivity means that participants are actively engaged in the topic. The spread will be more rapid if the adopters are more vocal, more influential, and have the other traits that Rogers (2003) ascribes to innovators and to early adopters for the diffusion of innovations.
If an idea has features of interest to the general population, it can enter the mass media with its wider audience of people with diverse interests. This population heterogeneity leads to homogeneity in one important respect: most members of the general public will be disengaged and dispassionate about the idea.
Idea adoption can be modeled using survival analysis if this population homogeneity is further accompanied by homogeneity in mass media penetration, meaning that all members of the public have the same chance of being exposed. Survival analysis is a broad class of mathematical models applicable to phenomena as diverse as radioactive decay and the death of people. In survival analysis, there are two key components: (1) a hazard or force causing change, and (2) a collection of unconverted individuals who respond by undergoing a one-way conversion. For human survival, the hazard is the “force of mortality,” and transition is from life to death. A survival-based model for ideas is “ideodynamics,” in which the force for change is a persuasive force and the responding population is the fraction of the public that does not accept the idea.
The mass media contributors to the persuasive force include media messages with each one assumed to have its maximum force at time of its release, and with the magnitude of that force declining exponentially over time with a rate characterized by a persistence constant. People can convert without learning about the message directly from the media. Instead, they can receive the information through interpersonal communications. The effective result is to prolong the persistence of the media message.
The other constituent of survival analysis is the non-adopters of the idea who are the potential targets of the persuasive force. The persuasive force will convert some fraction of the non-adopters with a probability given by a persuasibility constant. Survival analysis only considers the one-way transition from unadoption to adoption. Ideodynamics extends survival analysis by also including the reverse conversion from adoption to unadoption in response to opposing information (Fan & Cook 2003).
The validity of the ideodynamic model can be tested by its ability to use both favorable and unfavorable mass media messages to predict the time trend of the adoption of an idea. Theory assessment through time-dependent predictions is also the key to astronomy, for which it is also impractical to perform controlled experiments. Empirical studies show that the news media content can indeed capture the dominant forces for behavior change, because the ideodynamic model can use media coverage to predict behavior percentages over time with accuracies close to that of survey error. The behaviors have included not only such actions as cocaine use by teenagers (Fan & Holway 1994) but also opinions, such as the University of Michigan’s index of consumer sentiment (Fan & Cook 2003).
The persistence constant of a media message is typically much less than a single day. Thus a message and its interpersonal repetitions must change a behavior almost immediately or it will be forgotten. In other words, people act on new and not old information. If, instead, old information has a major effect, then behavior change would be more sluggish than actually observed due to the time needed to flush prior messages from memory. Nevertheless, prior persuasion can have a long-lasting effect by bringing an individual to a current behavior. After that behavior has been acquired, the individual might even forget the earlier information that had been the cause, but that does not contradict the prior media effect. If no new information is available, the prevalence of the behavior will stay unchanged. However, if new messages do arrive, then changes from adoption to unadoption and vice versa will occur following the ideodynamic processes based on survival analysis.
The ideodynamic equations can use media values at time t to predict behavior at time t + 1. That predicted behavior can then be used to predict behavior at time t + 2, and so on in a recursive manner. A potential danger in such an iterative calculation is that errors in media measurements will grow with each computation cycle, leading to expanding uncertainty in the predictions. However, formal statistics shows that this is not the case. Instead, the variance stabilizes, thereby justifying the recursive method that does not have the autocorrelation and missing data problems of autoregressive models (Fan & Cook 2003).
The Bass model for the diffusion innovations (Mahajan et al. 1990) is a special case of ideodynamics, with the special assumptions that the persuasive force from the mass media is constant over time and that the persuasive force from the social network is proportional to the adopters. This model is prevalent in marketing and related disciplines and assumes no idea unadoption. Given the satisfactory performance of the ideodynamic model, there is no need for the Rogers (2003) postulates of population and communication heterogeneity.
- Mahajan, V., Muller, E., & Bass, F. M. (1990). New product diffusion models in marketing: A review and directions for research. Journal of Marketing, 54, 1–26.
- Fan, D. P., & Holway, W. B. (1994). Media coverage of cocaine and its impact on usage patterns. International Journal of Public Opinion Research, 6, 139 –162.
- Fan, D. P., & Cook, R. D. (2003). A differential equation model for predicting public opinions and behaviors from persuasive information: Application to the index of consumer sentiment. Journal of Mathematical Sociology, 27, 1–23.
- Rogers, E. M. (2003). Diffusion of innovations, 5th edn. New York: Free Press.