Evaluating measures of campaign advertising exposure on political learning

Scholars employ various methods to measure exposure to televised political advertising but often arrive at conflicting conclusions about its impact on the thoughts and actions of citizens. We attempt to clarify one of these debates while validating a parsimonious measure of political advertising exposure. To do so, we assess the predictive power of six different measurement approaches - from the simple to the complex - on learning about political candidates. Two datasets are used in this inquiry: (1) geo-coded political advertising time-buy data, and (2) a national panel study concerning patterns of media consumption and levels of political knowledge. We conclude that many traditional methods of assessing exposure are flawed. Fortunately, there is a relatively simple measure that predicts knowledge about information featured in ads. This measure involves combining a tally of the volume of advertisements aired in a market with a small number of survey questions about the television viewing habits of geo-coded respondents.