< Back to previous page

Researcher

Roselinde Kessels

  • Research interest:My research is methodologically oriented and centered around two axes. The first involves my main field of expertise, which is the design and analysis of discrete choice experiments (DCEs) to quantify people’s preferences. The second axis involves the development and application of more general econometric and data science methods in health. Experiments are an attractive tool for inferring monetary value through (in)direct statements of preferences. DCEs are arguably the most popular method currently used in preference and willingness to pay (WTP) elicitation studies, both in hypothetical and non-hypothetical settings. Originally, in the 1970s, the method was developed for marketing and transport studies, but in the last two decennia, it has spread to environmental and resource economics, agricultural and food economics, and health economics. The ever growing body of literature on DCEs emphasizes the increasing role they are playing. In a DCE respondents are generally asked to make choices between multiple alternatives, also called profiles, which are described by a number of attributes with different levels. Consequently, through nonlinear regression models, the utility each attribute (level) contributes to the good or service under study can be quantified and translated into (marginal) WTP.I have been a pioneer in the development of Bayesian D-optimal designs with partial profiles that allow the study of many attributes in a DCE, while at the same time keeping the choice tasks simple for the respondents. Partial profile designs unite these two –at first sight– conflicting goals and are therefore smart designs. Also, I have developed these designs using the Bayesian D-optimal methodology, which I have improved extensively and has become the state of the art. The adjective ‘Bayesian’ signifies that prior information concerning the respondents’ preferences is taken into account when designing a DCE. The adjective ‘optimal’ is used because the alternatives or profiles appearing in the choice situations are selected so that, roughly speaking, the statistical model and quantities such as WTP can be estimated with maximum precision. Results of DCEs that I have executed in collaboration with practitioners have easily found their way to the popular press. For example, a DCE collaboration in health with the Centre for Health Economics Research & Modelling Infectious Diseases (CHERMID) of the University of Antwerp to determine the kinds of medical interventions Belgian citizens like to see reimbursed by the government, received much media attention through the press bulletin “Who is old and sick must suffer” in the Flemish newspaper De Standaard of June 18, 2014 and through the interview “Who is sick through their own fault should not count on compassion” in the program “De Ochtend” on Radio 1 on the same day. Coincidently, around the same time, a DCE similar to ours on the measurement of societal preferences for different criteria for reimbursement of medical interventions by the government had been performed by the Belgian Health Care Knowledge Center (KCE) who invited me to act as an external validator on their scientific report which was broadly published on 22 December 2014.Beyond DCEs I have also developed new indices to measure the correlation between socioeconomic conditions and health outcomes as well as new regression methods to decompose these indices. Recently, for example, I have adopted a distributional regression model that falls into the GAMLSS (Generalized Additive Models for Location, Scale and Shape) framework. Such approach proves useful because the difference in the effect of income on health between a low and a high income is found to be stronger for people on the lower health spectrum than would be portrayed by mean-based regression techniques. Note that any other type of analytical, econometric or data science problem also fits my research interest and is highly welcomed.
  • Keywords:DISCRETE CHOICE EXPERIMENT, EXPERIMENTAL DESIGN, STATISTICAL MODELLING, ECONOMETRICS, Economics and applied economics
  • Disciplines:Applied economics, Business administration and accounting, Economic development, innovation, technological change and growth, Economic history, Macroeconomics and monetary economics, Management, Marketing, Mathematical and quantitative methods, Microeconomics, Tourism, Other economics and business, Economic geography, Human geography, Recreation, leisure and tourism geography, Urban and regional geography, Other social and economic geography, Product development, Other social sciences
  • Research techniques:My specialties lie in discrete choice experiments, experimental design, choice modeling, data science, time series and (health) econometrics. I often make use of JMP statistical software from SAS Institute because it is fun to work with (note that the University of Antwerp has a license for JMP), and of the free programming language and environment for statistical computing and graphics R.
  • Users of research expertise:All researchers in economics and business, both in the public and private sector.