A good study design is very important because poorly designed studies may yield imprecise or biased and misleading results. In this course study designs for experimental studies (where an investigator is actively intervening) and for observational studies are considered. Applications are in agricultural, social and bio-medical sciences. A selection of the topics covered is: experimental designs, optimal designs, sample size calculation, randomized clinical trials, follow-up studies, case-control studies, confounding, sources of bias, survey sampling.
Objectives for the medical part:
Familiar with different types of trial designs and can determine the best design in a given situation
Know the basic principles of clinical trials (blinding, randomisation, the trial protocol, the intention to treat principle )
Can perform sample size calculations for different designs covered in this course.
Familiar with different types of sampling designs in observational studies (cohort and case control studies) and can determine the best design in a given situation
Familiar with different effect estimators and know which one can be estimated in the different designs.
Know the pitfall of observational studies (bias and confounding) and know how to prevent or to adjust for these factors.
Objectives for the agricultural part:
Know the structural characteristics of common agricultural and biological experimental designs (CRD, RCBD, BIB, split-plot, LS, cyclic designs)
Know common treatment designs (contrasts, factorial, crossed, nested)
Choose and motivate a suitable experimental and treatment design for a particular situation
Define relevant treatment contrasts and corresponding standard errors as functions of experimental design and replication
Define ANOVA and mixed models corresponding to experimental and treatment designs and perform calculations
Objectives for the social sciences part:
Have a general knowledge about what surveys are, why they are used, and what the challenges are;
Understand the paradigm of Total Survey Error;-
Can critically look at survey questions from the respondent-centered perspective;
Mode of instruction
Class time will consist of lectures and class discussion of assigned readings and homework problems.
Written Exam, plus obligatory assignments, which are graded as pass/fail. A student needs to pass all assignments. The final grade will be the grade of the written exam.
Welham SJ, Gezan SA, Clark SJ, Mead A. Statistical Methods in Biology: Design and Analysis of Experiments and Regression. Chapman and Hall/CRC. 2014
e-books which are available at the University library. A list will be provided during the course
Internet material provided during the course.
Detailed contact information is on Brightspace.
The coordinator can be reached via e-mail: prof. dr. S. le Cessie: S.le_Cessie@lumc.nl