Elicitation of expert information: Modelling latent growth models with prior expert information and evaluating predictions
The PhD project has a strong focus on the methodology how to elicit expert information. In specific methods will be developed to elicit beliefs and translate these into probability distributions concerning parameters in a latent growth curve model with distal (long term) outcome. Small sample sizes may make it impossible to estimate the latent growth curve models with distal outcomes in a classical statistical framework. By using the elicited information from experts as prior information in a Bayesian framework, researchers might be able to test their theories and fit their model. The methodology that will be developed will be used in empirical examples to solve the respective data problems and to demonstrate the use of the methodology. Further we will investigate methods for the evaluation of elicited beliefs in the light of new data.
Prof. Herbert Hoijtink & Dr Rens van de Schoot
NWO – Vidi grant Dr. Rens van de Schoot (NWO-VIDI-452-14-006)
1 August 2016 – 31 july 2020