Haynes, 1994; LaFollette, 1990). Few of these claims, however, have been subjected to empirical testing. Other studies have explored the production of printed media, focusing on the opportunities and constraints that shape their media content (Burnham, 1987; LaFollette, in press; Lewenstein, in press; Nelkin, 1987).
Particularly in the area of risk communication (the study and development of communicating the health implications of particular behaviors) some studies have examined the effects of particular print presentations of scientific information on individual perceptions of risk (Singer and Endreny, 1993; Walters, Wilkins, and Walters, 1989; Weiss and Singer, 1988; Wilkins and Patterson, 1990) In general, these studies have found that media do influence participants’ perception of risk related to events (hazards, natural disasters) that may have immediate consequences for them. However, individuals’ long-term considerations about these issues remain unaffected. This literature has also demonstrated that the social context in which stories are presented (e.g., the overall patterns of news coverage, the degree of trust that exists between readers and governmental or corporate institutions involved in the risk story) are typically more influential on participants’ perceptions of risk than the genre of individual stories (e.g., whether they are sensational or measured and analytic).
In recent years, political scientists and other scholars concerned about political communication have tried to correlate public opinion about scientific and technological issues with media coverage of such controversies as nuclear power, biotechnology, and nanotechnology (Bauer and Gaskell, 2002; Brossard, Scheufele, Kim, and Lewenstein, 2008; Brossard and Shanahan, 2003; Ten Eyck, 1999, 2005; Ten Eyck and Williment, 2003; Gamson and Modigliani, 1989; Gaskell and Bauer, 2001; Gaskell, Bauer, Durant, and Allum, 1999; Nisbet, Brossard, and Kroepsch, 2003; Priest, 2001; Scheufele and Lewenstein, 2005). Although there is evidence that both demographic and psychological characteristics can influence opinion, and claims have been made about the link between those characteristics and exposure to particular media frames (Nisbet and Goidel, 2007; Nisbet and Huge, 2006), the evidence is not yet sufficiently strong to draw conclusions about the effect of particular print media on either broad public opinion or individuals’ particular knowledge (Strand 2) and attitudes (Strand 6).
Particular science books are sometimes said to have had influence on the interests and career choices of later scientists, particularly Paul de Kruif’s 1929 Microbe Hunters and James Watson’s 1968 Double Helix (Lewenstein, in press), but little empirical evidence exists to show the direct effect of books on any of the strands of learning.