Sometimes political scientists analyze data where distinctions between data points are unobservable above or below a critical limit. In this case study, we will examine how to address the problems related to this kind of data (broadly known as censored data). The reader will be taken through an example of data that (sometimes) exhibit these featureśfeeling thermometers toward candidates in the 2016 U.S. Presidential Election. Particular attention will be paid to the use of Tobit models as a potential solution to censored data
