Surely this is a problem better analyzed by specifying ex ante a set of coding rules in response to measurable inputs, collecting data for those inputs, and the

Surely this is a problem better analyzed by specifying ex ante a set of coding rules in response to measurable inputs, collecting data for those inputs, and then scaling media outlets. The canonical political science/econ cite for this is Groseclose and Milyo QJE 2005. There are complaints about their approach, but it's certainly better than "I made a chart and just put stuff where I feel like it should be". The website supporting this chart (MediaBiasChart.com) has an interminable dozens of pages long blog post explaining that she wrote down a bunch of policy issues on a grid on a piece of paper and then those magically became this chart. I am unconvinced. This is a problem that a lot of smart and qualified people have explored in data-driven ways. I mention this because the purpose of the chart appears to be to appeal to professional journalism against punditry and poorly informed "citizen journalism", but it does so through an act of citizen journalism which does not appear to itself be all that well informed. I also think the visual symmetry in the graph probably creates a false equivalence between the relative influences of "equally bad" outlets on each side -- the Blaze and Breitbart are enormous media empires, and Infowars is what it is, while the equivalents on the left are no-name blogspam places. What was the inclusion rule?