can only tell the researcher about surface-level meanings of media text-- what can be readily counted, not the deep, often unconscious effects that a newspaper article or television program could have on a reader or viewer. In politics, it is reasonable to measure manifest content, since that is the content most often prompting policy debate, lobbying, and citizen participation. However, there is a feeling among many communication researchers that visual content of media may have subtle effects on the audience and that these effects are not readily noticed in the evaluation of manifest content of media. Research on how visual imagery represents and affects the public mood is only in its infancy, however, so media scholars have not as yet developed methods for coding latent meanings of visual content.
This chapter has explored four of the most common means for assessing public opinion, but there are others as well--analysis of election results or movement of the stock market, for example. The student of public opinion must decide upon the most effective means for understanding popular moods, and the tools chosen must fit the research question at hand. Do keep in mind that all the methods discussed here assume certain conceptions or dimensions of public opinion. In conducting a poll, a researcher is arguing that public opinion can reasonably be defined as the aggregation of individual opinions, anonymously and scientifically collected. Yet if focus groups are used, the researcher wants to witness the processual nature of public opinion: how people develop opinions and change them, especially when they articulate their opinions in the presence of others. The survey researcher understands that a respondent's opinions have been influenced by others as well (in conversations before the survey), but the focus group researcher wants to watch how those opinions are formed. All of the approaches to understanding public opinion discussed in this chapter are valid, if used with care and rigor. And it is important to note that they are not mutually exclusive. In fact, the best research project would evaluate public opinion using multiple methods. But this is often unnecessary and also quite expensive, so researchers must make the best choices they can, based upon their theoretical concerns, the possibilities for collecting data, and budgetary constraints.