While spatial models of party competition with valence have long been acknowledged as work horse model in electoral studies those empirical analysis is still deficit, in particular concerning valence. Conventionally, many empirical researchers measure valence using certain indicators of candidates'/parties' traits, such as perceived integrity and competence. However, the indicators included in the survey represent only a fraction of all possible valence substances. Further, the perceived traits of political actors might be biased by respondents' ideological orientation. This paper aims to close this gap by providing a simple definition of valence as inherent attractiveness based on psychological research and proposing a statistical model that enables researchers to estimate the ideological positioning of political actors and voters as well as valence. The model requires as input only feeling thermometer scores so that it is applicable to different data and enables researchers more systematic comparable analysis.
