Precision journalism (1973)

Fourth edition: A Reporter's Introduction to Social Science Methods, by Philip Meyer (Rowman & Littlefield, 2002) The oldest book here and still the one that explains why a pattern in the data is or is not evidence. Meyer argued that reporters should borrow the tools of social scientists, and the fourth edition covers data analysis, statistics, surveys, experiments, databases and election analysis, along with lurking variables and regression. The software chapters have aged; the reasoning about cause, sampling and confounders has not. Worth it for the reporter who wants to defend a data finding against an official's pushback.