notes/scaling.md

In ManifestoProject/manifestoR: Access and Process Data and Documents of the Manifesto Project

output: pdf_document

Things / Steps to consider when using the Manifesto Data

for most of these steps, there is no right or wrong, but these make different assumptions the data generating process and the nature of issues and dimensions.

selecting cases, preprocessing

exclude estimates(progtype)

match successor parties

coding scheme aggregation

aggregation of cee categories

base value

dimensionality

measuring one issue, one latent dimensions,...

selecting or identification of categories

theoretically driven

data-driven - prosser - mölder

weighting of categories

deductive (mostly no weighting)

data driven weighting

scaling function

linear scaling ratio log scaling

smoothing procedures

uncertainty estimates

match with other data / rescaling

empirical or theoretical ends on different scales

ManifestoProject/manifestoR documentation built on May 7, 2019, 2:25 p.m.