SFKKA9041U Statistical Design and Analysis of Experiments

Volume 2013/2014
Education
MSc in Pharmaceutical Sciences - elective, MSc in Medicinal Chemistry - elective, Cand.Scient.Pharm. - elective, Cand.Pharm. - elective
Content
  • Principles of confounding, randomization, blocking and balancing.
  • Choice of necessary sample size.
  • Estimation and testing of treatment effects and components of variance.
  • Experiments with many factors.
  • Restrictions on randomization and elementary split plot and repeated measures designs.
  • Some insight into the design and analysis of experiments with a dichotomous outcome (binomial trial).
The syllabus is presented by three parallel activities. The theoretical and methodological content is covered in about 12 double lectures based on the textbook and lecture notes. Pharmaceutical applications and case studies related to the most important activities involving discovery and development of pharmaceutical products are presented in a series of about six double lectures. Training in the practical design and analysis of experiments is provided in a series of nine double classroom sessions, in which the different methods are discussed and practised using manual calculations and statistical software.
Learning Outcome

Formål / Objective

To introduce a general statistical approach to the design of laboratory and similar experiments and to analyse the resulting data.Along with the statistical and methodological content of the course, a number of concrete and frequently used pharmaceutical applications (designed experiments) are presented. Examples are clinical trials (including e.g. crossover and repeated measures designs), toxicity testing, bio-equivalence analyses, assay validation, design and analysis of epidemiological surveys, etc.Regarding the statistical design of experiments, the main objective is to assure the accuracy and precision of the data so that reliable and reproducible conclusions can be drawn concerning the relations being studied.

Målbeskrivelse / Course outcome

After completing the course, the student is expected to know how to design relevant experimental work to support the lifecycle phases of a new pharmaceutical product: discovery, development, production and quality control. The student can judge the necessary amount of experimentation, take into account practical restrictions and anticipate and prevent/mitigate sources of the types of bias often encountered during these phases.The student will learn to carry out statistical analysis of data obtained from a given experimental design, most often using variance and regression analysis techniques. The student will also learn to present and interpret the results obtained.
The student will become familiar with using and understanding output from modern statistical computer software.
Literature
  • Design and Analysis of Experiments, D. C. Montgomery, 8. ed., 2012
  • Lecture notes are available from the course homepage.

 

Necessary statistical prerequisite for the course is good knowledge of elementary statistics.
• - DLectures/seminars: 33 hours
•Classroom exercises: 18 hours
  • Category
  • Hours
  • Colloquia
  • 18
  • Exam
  • 4
  • Lectures
  • 33
  • Preparation
  • 151
  • Total
  • 206
Credit
7,5 ECTS
Type of assessment
Written examination, 4 timer under invigilation
Prøveform / Examination type:
Four-hour written examination

Prøvedesign / Examination design
The written examination is a two-part questionnaire. The first part is an essay-type, practical question asking the student to partially design an experiment and/or do a statistical analysis. The second part is a multiple-choice test with questions generally related to important issues discussed in the course.
Aid
Written aids allowed
All written aids.

Apart from the standard programs and IT tools listed under The Faculty of Health and Medical Sciences at http://pc-eksamen.ku.dk/pc_exam students will at this exam also have access to MathType and the possibility to use a USB stick (for notes etc.)
Marking scale
7-point grading scale
Censorship form
External censorship
Re-exam
Please note that the registration period for participation in the re-exam (resit) is different from the registration period stipulated in the general curriculum provisions. The exact registration period can be found in the exam timetable.
Criteria for exam assesment

Beskrivelse af prøven og bedømmelse / Description of examination and assessment criteria

Beståkrav / Pass criteria

In general 50% of the examination questions must be answered satisfactorily in order to pass (grade 02). The two parts of the examination are weighted equally.

Karakterbeskrivelse / Description of grades

12 - Excellent performance
Is given for an excellent solution, and answers must contain almost no misunderstandings or errors related to important concepts or methods covered in the course syllabus or related to subjects that are part of the prerequisites required for attending the course. 7 - Good performance
Is given for a reasonable-to-good solution. The student must demonstrate a good knowledge of many of the basic and important concepts and methods presented in the course. 02 - Acceptable performance
Is given for a solution just above the limit for passing the course. The student must as a minimum demonstrate elementary knowledge of a number of the basic and important concepts of the syllabus of the course.