Course code STAT310

STAT310 Design of Experiments and Analysis of Variance II

There may be changes to the course due to to corona restrictions. See Canvas and StudentWeb for info.

Norsk emneinformasjon

Search for other courses here

Showing course contents for the educational year 2014 - 2015 .

Course responsible: Thore Egeland
Teachers: Trygve Erik Almøy
ECTS credits: 10
Faculty: Department of Chemistry, Biotechnology and Food Science
Teaching language: EN, NO
(NO=norsk, EN=Engelsk)
Limits of class size:
minimum 5, maximum 100
Teaching exam periods:
This course starts in August block. This course has teaching/evaluation in August block, Autumn parallel, .
Course frequency: Annually
First time: 2004H
Last time: 2015H
Course contents:
Fundamental theory for design of experiments and analyses of data from such experiments, replication, randomisation and blocking. Analysis of variance models with fixed, random and mixed effects. Hierarchical models. Orthogonal contrasts. Splitting of sum of squares. Multiple comparisons. Testing equality of variances. Consequences of departure from the assumptions. Transforming data. Analyses of unbalanced data. Some usual experimental plans, such as: Completely randomised design, block design, Latin square design, split-plot design and incomplete block design. Factorial designs interaction. Two- and three-level designs. Fractional factorial designs. Confounding of effects.
Learning outcome:
The students will learn the statistical principles for design of experiments used to compare different groups or treatments, and to analyse data from such experiments, first of all by means of analysis of variance. They also learn the mathematical basis so that they will be able to use their knowledge in new situations that they encounter in their studies and later in their working life. By means of exercises and projects with real problems and data, the students should show that they have reached the learning goals.
Learning activities:
Lectures and exercises with guidance.
Teaching support:
Questions etc. are answered in office hours or by e-mail: thore.egeland@umb.no.
Syllabus:

Reading list is handed out in the first lecture.

Text book Stat210/Stat310 : Douglas C Montgomery, Design and analysis of experiments. 6th edition or later

Prerequisites:
STAT100 or equivalent. The first part of the course is taught in the August block in STAT210 and the students must know the topics covered in STAT210 and also pass the compulsory assignment for STAT210.
Recommended prerequisites:
Knowledge of simple matrix algebra.
Mandatory activity:
Compulsory project assignment.
Assessment:
Written examination, 3.5 hours.
Nominal workload:

Organised teaching: 96 hours.

Individual study: 204 hours.

Entrance requirements:
Special requirements in Science
Reduction of credits:
STAT210, 5 credits (ECTS).
Type of course:

August block (STAT 210 part): 2 hours lectures and 2 hours exercises daily.

Autumn parallel: 2 hours lectures and 2 hours exercises weekly.

Note:
Students taking STAT310 should follow the STAT210 lectures, and take the final exam at the end of the fall term (i.e. these students do not take the august block exam).
Examiner:
Use of external examiner who approves the examination questions and grades the examinations together with the lecturer.
Allowed examination aids: All types of calculators, all other aids.
Examination details: One written exam: A - E / Ikke bestått