MF9375 ¨C Choice modelling in health

Course content

Decision-making in the healthcare sector often involves selecting from a set of alternatives. For example, healthcare providers face discrete choices when deciding which diagnostic test to conduct, which medication to prescribe, or which clinic to refer their patients to. Similarly, patients make discrete choices when selecting their general practitioner, choosing a hospital for childbirth, or deciding between brand-name medications and their generic counterparts.

This course offers a practical introduction to applying discrete choice models to various types of data, including registry data, data from discrete choice experiments, and data from laboratory experiments. The focus is on decision-making processes within the health sector, ensuring a high degree of relevance to real-world healthcare scenarios.

The course starts out with an introduction to the literature that provides the motivation and theoretical foundation for modelling behavior by means of logit models. Participants are introduced to key concepts of Discrete Choice Experiments.

Participants are introduced to examples of data structures from other types of choice experiments such as, for example, "dictator games" and "cournot games".

The course proceeds in a workshop format. Participants will be provided with case material comprising example data sets and code. Participants will be working actively with analyzing example data sets in Stata and learn how to simulate choice behavior. The software used in the course is Stata, Excel with Excel solver add-in, and Gambit. Some code for conducting analysis in R will be made available for R-users upon request.

Participants will learn to estimate several types of choice models. Participants will use Stata for estimating multinomial logit models, mixed logit models, generalized multinomial logit models and quantal response equilibrium choice models.

Learning outcome

Knowledge

You will gain knowledge of:?

• How to generate new knowledge by applying discrete choice models to health settings.
• How knowledge of preferences of patients and health care providers can guide healthcare decision-making and policy development.
• How choice modeling combines economic utility theory and probability theory.
• How to account for heterogeneity of preferences and degrees of rationality in model specification.
• The role of context and framing effects in influencing choice behavior.
• An overview of the latest advancements and applications of discrete choice modeling.

Skills

You will learn how to:

• Design a discrete choice experiment.
• Prepare data for discrete choice analysis.
• Simulate choice behavior using specialized software and tools.
• Estimate multinomial logit models, mixed logit models, and generalized multinomial logit models.
• Analyze data from games using quantal response equilibrium choice models.
• Interpret and communicate the results of discrete choice models to support decision-making processes.
• Specification of discrete choice models that address specific questions in healthcare, such as treatment choices and choice of provider.

General Competence

You will attain:

• Competence in using Stata for discrete choice analysis.
• Competence in estimating tailored discrete choice models to address specific research questions and practical applications.
• The ability to discuss results from discrete choice models.
• The ability to translate discrete choice model findings into actionable healthcare policy recommendations and/or supply side market strategies.

Admission to the course

Applicants admitted to a PhD programme at UiO sign up for classes and exam to this course in StudentWeb.

Applicants who are not admitted to a PhD programme at UiO must apply for a right to study before they can sign up for classes and exam to this course. See information here: How to apply for a right to study and admission to elective PhD courses in medicine and health sciences..

Applicants will upon registration receive an immediate reply in StudentWeb?as to whether a seat at this course is granted or not.

Formal prerequisite knowledge

• Completed Master's degree.
• Background in statistics and economics.

Overlapping courses

• 4 credits overlap with MEDFL5375 ¨C Choice modelling in health.

Teaching

The course is organized over five days; Consisting of four full days of teaching, followed by an exam on the fifth day.?

The course is taught in a workshop format.

Participants will be provided with case material comprising example data sets and code.

Participants must bring their own laptop computer and have access to the recent versions Stata, Excel, and Gambit. Basic experience with using Stata and Excel is recommended. Gambit can be downloaded here: http://www.gambit-project.org/ Kenneth Train¡¯s textbook ?Discrete choice methods with simulation, (Second edition)? is the main textbook reference for the course. It can be downloaded from Train¡¯s own webpage: https://eml.berkeley.edu/books/choice2.html . Additional reading material will be shared in class.

You have to participate in at least 80 % of the teaching to be allowed to take the exam. Attendance will be registered

Examination

School exam on the fifth day of the course.

Language of examination

The examination text is given in English, and you submit your response in English.

Grading scale

Grades are awarded on a pass/fail scale. Read more about?the grading system.

More about examinations at UiO

• Use of sources and citations
• Special exam arrangements due to individual needs
• Withdrawal from an exam
• Illness at exams / postponed exams
• Explanation of grades and appeals
• Resitting an exam
• Cheating/attempted cheating

You will find further guides and resources at the web page on examinations at UiO.