I love Phil Cohen, who runs the blog Family Inequality. He's an outstanding sociologist who's passionately interested in social inequality, but also in the honest, ethical, and nuanced presentation of social science data. Here, he shows how failing to carefully select a range for your data can lead to inaccurate conclusions. In the linked post above, the historical trend in age at first marriage is misrepresented by selecting a range that begins in a decade where people were, for various historical reasons, getting married unusually young, but similar problems occur when, for instance, a pundit describes a rise from 0.4% to 0.8% as "more than doubling" without providing any context. So: do not immediately accept the information displayed graphs, charts, and claims about social trends, especially when presented by people who may be using those claims to their own political ends (as the Heritage Foundation did with the Pew Report). DO ask questions (iI this a reasonable range over which we can see a social trend play out? Would its apparent significance change if we chose a different range?) and seek confirmation from other, trustworthy sources. More on what makes a source trustworthy later.
