Precision Journalism

Precision journalism is the application of social and behavioral science research methods to the practice of journalism. The term was coined by Everette E. Dennis for a “Seminar in the New Journalism” that he taught in the winter of 1971 at the University of Oregon. The concept was explicated by one of his students, Neil Felgenhauer, in a term paper that became a chapter in a book based on the seminar's work.

Most of the “new journalism” that inspired the seminar was the creation of talented writers such as Tom Wolfe, Gay Talese, and Norman Mailer, who used fiction techniques to construct powerful narratives about real people and events. The class discussion contrasted their work with precision journalism. “In essence, all the other new journalists push reporting toward art. Precision journalists push it toward science,” said Dennis and William L. Rivers in a 1974 report.

Within the field of investigative reporting, there is also a contrast of method. Some follow paper trails, and a subset uses scientific method in the analysis. Others base their investigations on anonymous sources whose motivation, as well as identity, is veiled or not revealed at all. The work of the former can be verified because its methods are transparent and replicable. Like a scientist, a precision journalist describes the data collection and the methodology in sufficient detail that another investigator could retrace the steps and arrive at the same conclusions. Precision journalists can also build confidence by posting their data in university archives for secondary analysis.

The genesis of precision journalism can be traced to the first public opinion polls that used scientific sampling instead of casual person-on-the-street interviewing. Dr. George Gallup based a newspaper column on national polls and scored a coup by using more rigorous methods and getting a better result than the more famous Literary Digest poll, which failed to predict the outcome of the presidential election of 1936. However, Gallup and other national pollsters did not adopt probability sampling, a basic tool of modern social science, until after their own error in calling the 1948 election for Thomas E. Dewey instead of Harry S. Truman.

Other pioneers in the use of precision journalism were the broadcast networks striving to be first to announce the winner in presidential elections. In 1960, CBS used a computer-driven statistical model based on the timing of the results. It captured the Republican and Democratic standings at given points in time and compared them to their standings at the same times in the 1956 election. The initial extrapolation proved flawed, however, because Kansas had adopted a faster method of vote counting. The resulting Republican bias produced an early and erroneous call for Richard Nixon. Later network projections were based on geographic sampling and proved more reliable.

The civil rights and antiwar movements of the 1960s created further demand for new reporting [Page 1101]techniques. Standard journalism was drawn to the loudest and most visible spokespersons for their respective movements and tended to overgeneral-ize from them. Newsweek magazine recognized this problem when it commissioned Louis Harris to do special polls among black Americans to detect a broader spectrum of attitudes.

...