MVE155 / MSG200 Statistical inference

MVE155 / MSG200 Statistical inference (7.5 hp) 2024

Course is offered by the department of Mathematical Sciences

Exams

The June 5 exam with the solutions is here.

The March 12 exam is here, and with solutions here.

Teacher

Lectures: Tony Johansson

Exercises: Tony Johansson

Course purpose

"Statistical Inference" is a second course in mathematical statistics suitable for students with different backgrounds. A main prerequisite is an introductory course in probability and statistics. The course gives a deeper understanding of some traditional topics in mathematical statistics such as methods based on likelihood, aspects of experimental design, non-parametric testing, analysis of variance, introduction to Bayesian inference, chi-squared tests, multiple regression.

Sample exam & Distribution tables

A sample exam is here.

For the final exam, tables of distribution values will be provided for you. The tables can be found here. Do not bring this document with you to the exam.

Topics not included this year

I move through topics at a slower pace than previous lecturers in this course, and have decided to scrap some topics. These are:

  • Bayesian hypothesis tests (the basics of Bayesian statistics are still covered)
  • McNemar's test
  • 2-way ANOVA and Friedman's test

Lecture notes

Please note that these notes may contain mistakes. The correspondence between lectures and notes is not one-to-one (the main difference is some examples are tweaked).

Supplemental material

Under this header I will post any supplemental material that I write. Please let me know if you find any errors.

Preliminary schedule 

The linked notes are from last year, when the course was taught by Aila Särkkä. This year's content will be similar but not identical.

See also 

Mon 15/1, 13:15-15:00, Vasa B

Lecture: Introduction; Parametric models (notes, Chapter 1)

Chapter 1

Tue 16/1, 13.15-15.00, KC

Lecture: Parametric models  Chapter 1

Wed 17/1, 13.15-15.00, HC3

Exercises Chapter 1

Fri 19/1, 13.15-15.00, KA

Lecture: Random sampling (notes, Chapter 2)

Chapter 2
Mon 22/1, 13.15-15.00, KC Exercises Chapter 2
Tue 23/1, 13.15-15.00, KC

Lecture: Random sampling

Chapter 2
Wed 24/1, 13.15-15.00, KA Exercises (R code: Problem 7, Problem 10) Chapter 2
Fri 26/1, 13.15-15.00, KA Lecture: Parameter estimation (notes, Chapter 3)  Chapter 3
Mon 29/1, 13.15-15.00, KA Exercises Chapter 3
Tue 30/1, 13.15-15.00, KC

Lecture: Hypothesis testing (notes, Chapter 4)

Chapter 4
Wed  31/1, 13.15-15.00, KA Exercises Chapter 4
Fri 2/2, 13.15-15.00, KA Lecture: Hypothesis testing Chapter 4
Mon 5/2, 13.15-15.00, Pascal Exercises  Chapter 4
Fri 9/2, 13.15-15.00, KB Lecture: Bayesian inference (notes, Chapter 5) Chapter 5
Mon 12/2, 13.15-15.00, Pascal

Exercises

Chapter 5
Tue 13/2, 13.15-15.00, Pascal Lecture: Summarising data (notes, Chapter 6)

Chapter 6
Wed 14/2, 13.15-15.00, Pascal Exercises (R code) Chapter 6
Fri 16/2, 13.15-15.00, KA Lecture: Summarising data, Comparing two samples (notes, Chapter 7) Chapter 7
Mon 19/2, 13.15-15.00, Pascal Exercises Chapter 7
Tue 20/2, 13.15-15.00, Pascal Lecture: Comparing two samples, Analysis of variance (notes, Chapter 8) Chapter 8
Wed 21/2, 13.15-15.00, Pascal Exercises Chapter 8
Fri 23/2, 13.15-15.00, KA

Lecture: Analysis of variance 

Chapter 8
Mon 26/2, 13.15-15.00, Pascal Exercises Chapter 8
Tue 27/2, 13.15-15.00, Pascal

Lecture: Categorical data analysis (notes, Chapter 9)

Chapter 9
Wed 28/2, 13.15-15.00, Pascal

Exercises

Chapter 9
Fri 1/3, 13.15-15.00, KA Lecture: Multiple regression (notes, Chapter 10) Chapter 9
Mon 4/3, 13.15-15.00, Pascal

Exercises

Chapter 9
Tue 5/3, 13.15-15.00, Pascal Lecture: Multiple regression  Chapter 10
Wed 6/3, 13.15-15.00, Pascal Exercises Chapter 10
Fri 8/3, 13.15-15.00, KA

Questions and answers session

 Chapters 1-10
Tue 12/3, 14:00-18:00 Exam (register before 25.02.2024)  

Course literature

The course is build around the compendium - click and download. The compendium may undergo minor updates during the course.

Recommended additional textbook: Mathematical statistics and data analysis, 3rd edition (2nd edition is also OK), by John Rice (Cremona).

Learning objectives and syllabus

Learning objectives:

- summarize multiple sample data in a meaningful and informative way, 

- recognize several basic types of statistical problems corresponding to various sampling designs, 

- estimate relevant parameters and perform appropriate statistical tests for multiple sample data sets.

Link to the syllabus on Studieportalen: Study plan

Examination form

The grading of the course is based on a written examination. Preparing for the final exam, check Section 12.1 of the Compendium to see the list of the topics that may be addressed by the final exam questions.

You are allowed to use a Chalmers allowed calculator and your own course summary (four A4 pages) during the final exam. Importantly, this summary should not be produced by copying and pasting of different parts of the compendium.

Several old exams with solutions are given in the module "Old exams". Aila's old exams of the course Experimental design: Exam 1 and solutions, Exam 2.

Maximal number of points for the final exam is 30. Passing limits 

  • Chalmers students: 12 points for '3', 18 points for '4', 24 points for '5'
  • GU students: 12 points for 'G', 20 points for 'VG'