This chapter outlines the method used to predict the parties’ self-ascribed and unified policy positions at the two stages of the representation process required for the analyses in this book. It compares established methods for detecting party positions and selects the most suitable method and documents for the task at hand. Automated content analysis was chosen for two reasons. First, it enables to closely control the basis for the position prediction, making it possible to clearly connect the estimated position to the two stages. Second, it enables the management of large volumes of text, which is necessary given the sources selected. These sources are manifestos and plenary speeches. They have been selected because they are ideal for estimating the parties’ self-ascribed positions. Both are texts in which parties write and speak about their own positions, thereby offering a perspective on party positions as the parties wish to present them, rather than as they are viewed by other actors, such as media representatives or citizens. The chapter concludes with descriptive information on the specific texts analysed in this book: manifestos and speeches from the German political parties.