Expert Elicitation for Latent Growth Curve Models. The Case of Posttraumatic Stress Symptoms Development in Children With Burn Injuries


Experts provide an alternative source of information to classical data collection methods such as surveys. They can provide additional insight into problems, supplement existing data, or provide insights when classical data collection is troublesome. In this paper, we explore the (dis)similarities between expert judgments and data collected by traditional data collection methods regarding the development of posttraumatic stress symptoms (PTSSs) in children with burn injuries. By means of an elicitation procedure, the experts’ domain expertise is formalized and represented in the form of probability distributions. The method is used to obtain beliefs from 14 experts, including nurses and psychologists. Those beliefs are contrasted with questionnaire data collected on the same issue. The individual and aggregated expert judgments are contrasted with the questionnaire data by means of Kullback–Leibler divergences. The aggregated judgments of the group that mainly includes psychologists resemble the questionnaire data more than almost all of the individual expert judgments.

Front. Psychol. 11:1197
Click the Cite button above to demo the feature to enable visitors to import publication metadata into their reference management software.
Duco Veen

I currently hold a post-doctoral position at the department of Methodology & Statistics of Utrecht University in the Netherlands. From December until July I’m consulting as a scientific project manager for Rick Grobbee at the University Medical Center Utrecht. My research interests lie primarily in the area of Bayesian statistics. My docteral thesis, under the supervision of Rens van de Schoot, concerned the use of Bayesian statistics, expert elicitation and information theory in the social sciences. I defended my thesis on March 13 at the acadamy building in Utrecht. In June I joined the Stan development team for my work on shinystan.