All scientific investigation deals with building models for reality. The models will necessarily be simplifications of reality, and so the actual observations we do, i.e., the data we have, will contain variability not captured in the model. When our data are numbers, like counts or measurements, statistics is an indispensable tool for dealing with the variability in a scientific way, so that conclusions about the simplified models can be drawn in spite of the extra variability in the data. When the data is the result of an experiment, the design of the experiment, i.e., how it is performed, determines how firm conclusions can be drawn about the models, in spite of the variability in the data. Good experimental design is essential for a good scientific experiment; without it, a lot of work can be wasted.

The course is given for masters students from a number of different fields within the natural sciences, and we aim to reflect some of their different subject matter in the course. However, the principles of statistical analysis and experimental design are generally the same. We will focus on ideas and methods that should be useful in many masters-projects, as well as in later scientific investigations. The students' differeing interests and backgrounds in statistics will be reflected in their work on a miniproject.


The course gives 7.5 hp.


Latest news
News about the course will be available at the course home pages at GUL: log in at portalen.gu.se and go to GUL. Those who are registered in LADOK as students of this course will have an activity called ”MSA830, V10, Statistical analysis and experimental design” there, which will contain more information.
Examiner and lecturer
Petter Mostad 
Exercises: Malin Östensson 
Course literature
Main text book

Lecture notes, available from GUL.

Douglas C. Montgomery: Design and Analysis of Experiments, 7th edition.
The book is available at the Cremona bookstore. The following parts of Montgomery are recommended reading: Chapter 1, Chapter 2, Chapter 3 (3.1-3.4), Chapter 4 (4.1-4-2), Chapter 5 (5.1-5.3), Chapter 6 (6.1-6.4), C hapter 7 (7.1-7.5), Chapter 8 (8.1-8.2).

Preliminary plan for lectures and classes
Schedule
The shedule is available here.

Computer labs
An important part of the course is working with computers. It is your own responsibility to obtain a username and password for the computers in the computer lab. You can get this at the Mathematical Sciences helpdesk, which is located close to the glass bridge connecting the buildings of our institution and which has office hours 12:30 – 13:30 on Tuesdays and Thursdays, and in addition will be open on MONDAY 18 JANUARY 12:30 – 13:30. NOTE: YOU CAN START USING YOUR ACCOUNT THE DAY AFTER YOU HAVER RECEIVED IT, so you should get your account at least one day before your first computer lab. Before you can get a new account, you need to be registered in Ladok as a student in one of our courses. See below for how you make sure you are registered for MSA830.
To get information about the contents of the computer labs, go to GUL.
Exercise groups
There will be one exercise group each week. At the exercise groups, you will practice exercises similar to those given at the written exam.

Sign-up and registration
Ideally, you should sign up for this course using the studera.nu website. Unfortunately, we are currently not allowed to inform about this course in the English-language part of studera.nu. Thus, if you cannot navigate the Swedish interface of studera.nu, you should sign up by contacting your master program coordinator. He or she will then forward your sign-up to the mathematics department. To avoid administrative problems and delays for yourself and others, please sign up as early as possible.

Course registration is done by personal appearance at the first lecture. If you signed up for the course and then show up for the first lecture, you will be registered for the course in the Ladok database very soon afterwards. If you signed up and would like to take the course, but cannot come to the first lecture, contact Petter Mostad. If you for some reason were unable to sign up for the course but would like to follow it, you should show up at the first lecture and leave your name and personal Swedish ID number, and we will try to register you, but there may be delays, and we make no guarantees.

It is your own responsibility to make sure you are registered for the course. You can check this in the GU student portal. If you have followed one of the procedures above, but are still not registered 3-4 days after the first lecture, you should contact Jeanette Montell to clear things up. It is only after you have been registered that you will be able to access all course material at the GUL website at the GU student portal, get a computer account, and submit project reports. Course-related mails will be sent out to registered students using their GU mail adress. Thus you must check this mail account, or make sure it forwards to a mail account you use.

Taking this course as a compulsory course
This course is gererally compulsory for master students at the Faculty of Natural Science at Göteborg University. Consult your own master program to see if this course is compulsory for you. Even when the course is compulsory, you may be exempt if you have corresponding knowledge from other courses. It is the coordinator of your master program who decides (possibly after consulting with the teacher of this course) if you may be exempted, so you should contact her or him with questions about this.

When this course is a compulsory part of your master degree, it is advisable to pass the exam early in your master program. The course is taught two times each year, in the first part of the autumn and the first part of the spring, and in addition there may be three re-exams, in January, June, and in August. UNDER NO CIRCUMSTANCES will there be additional exams given. Some students delay passing this exam so long that they end up with serious problems with visas, with PhD positions, etc. Make sure you will not be one of these students.

Examination
To pass this course you should pass the written exam and complete the miniproject. The written exam counts as 5.5 higher education points, while the miniproject counts as 2 higher education points. The course grade is based on the written exam, with the additional requirement that the miniproject has been passed.
Written examination
IMPORTANT NOTE: You need to sign up for the written examination, separately from signing up for the course. You do this between three and one weeks before the exam. You sign up for the exam via the Ladok web-service which you should have access to via your student portal. If you want to take a re-examination, you also need to sign up for it the same way.
During the exam the following aids are permitted: A calculator, lecture notes and textbooks, printouts, and your own personal notes. (THIS IS DIFFERENT FROM PREVIOUS YEARS) . Copies of relevant tables from the textbook will be provided. Bring an ID and the receipt for your student union fee. The Chalmers examination room instructions will apply.
The exam will be graded in the weeks following the course. Directly after the grading is finished, suggested solutions will appear at the GUL webpage. Please do not enquire about the grading, as this will only take time away from the grading. After the results have been de-anonymized and registered in Ladok, you will automatically receive a mail from Ladok with your result.
After you have received notification about your result, we will announce a time when you may come and pick up your exam from the teacher. After this time, the exams will then be kept for a year at the students' office in the Mathematical Sciences building, where you may pick up your exam. Check that the number of points and your grade given on the exam and registered in Ladok coincide. Complaints about the marking should be in writing and handed in at the expedition. There is a form you can use; ask at the expedition.
Miniproject
Students should work in groups of 2-3 people. In the projects, the students should apply methods for experimental planning and data-analysis to some problem they find interesting. First, an experimental plan is written and handed in for comments. Then, the actual experiment is performed, and an analysis is written and handed in for comments. Exact instructions will be given out via GUL.

Old exams
There are a number of old exams from this course, which will be made available via GUL. However, as the teaching and the contents of the course is slightly changed from 2010, we recommend that you instead practice on exercises which will be avialable via GUL.