Job - PhD Student on the subject ‘Multivariate logistics regression using the ideal point classification model’

Project description

Multivariate categorical data, with multiple dependent variables and one or more independent variables are often collected in the social sciences. In the Netherlands study of depression and anxiety (NESDA, see http://www.nesda.nl), for example, data are collected on both depression and anxiety disorders (dependent variables) and personality characteristics (independent variables). For the analysis of such data, however, only limited statistical tools are available. The methodology that is currently available makes either unverifiable assumptions or requires the independent variables to be categorized. In this project new methodology is proposed, based on the ideal point classification model, which requires a minimal set of assumptions and takes the data as it is (no categorization required). Essential tools for the evaluation of effects and for the design of empirical studies will also be proposed.

A more detailed project description can be requested from Dr. Mark de Rooij (rooijm@fsw.leidenuniv.nl) and/or Jacqueline Vreriks (jvreriks@fsw.leidenuniv.nl).

We are searching for an enthusiastic and ambitious candidate with a recent MSc in the methodology of psychological research, statistics, or a comparable qualification. The preferred candidate has experience in scientific programming (for example in R and/or MATLAB).

We offer you an appointment for a period of one year with a possibility for prolongation up to four years depending on your performance in the first year. The gross monthly salary is € 2.042 in the first year and increases progressively each year up to € 2.612 in the fourth year in accordance to the Collective Labour Agreement for Dutch Universities (CAO Nederlandse Universiteiten).

The Methodology and Statistics group of the Leiden University Institute for Psychological Research consists of 11 staff members and 4 PhD candidates. The research themes include longitudinal data analysis, latent variable models, multilevel models, and multidimensional scaling and unfolding. For more information see http://www.fsw.leidenuniv.nl/psychologie/organisatie/meth/. The project leaders for this research project are Dr. Mark de Rooij. prof.dr. Willem Heiser, and Prof.dr. Philip Spinhoven.