ST104: Statistical Laboratory I

ST104 ("Statistical Laboratory I") is an introductory statistics course for first-year students, emphasising general principles and healthy scepticism.

Aims

To introduce and explain the important ideas in practical statistics, so that, with full access to textbooks and other resources, students will know when to apply various statistical methods, and will understand the associated problems and pitfalls.

Objectives

After completing this course, students should be able to do the following given a practical problem:

1. Suggest methods to obtain relevant data.
2. Summarise low-dimensional data-sets, both graphically and numerically.
3. Apply simple formal statistical techniques and interpret the results.
4. Criticise the whole process and the applicability of the conclusions, in the light of the practical situation and the actual data, discussing points such as: (sampling) bias, data quality, independence, explanatory variables, distributional assumptions, outliers, prediction.
5. Suggest improvements to the design and analysis.

Prerequisites

A level Mathematics.

Syllabus

The lectures are supported and extended by "laboratory sessions", in which students write short reports on about 25 problems (data sets + some background information), of which about 15 are genuine and the others are simplified to emphasise particular points. The following syllabus refers only to the lectures.

Background

What "statistics" means; overview of ST104.

Data

Examples of practical problems, types of data-set; populations, samples, variation; stages of data analysis; stem-and-leaf plots, boxplots, scattergrams, other types of plot; parameters and statistics.

Probability

Experiments, outcomes, events; mathematical probability; independence; random quantities; probability models; expectation, properties of means and variances; distributions; the central limit theorem.

Statistical inference

Assumptions, model criticism, transformations. Hypothesis testing, chi-squared tests, two-sample and paired t tests; point estimation, interval estimation, likelihood.

Review

Applicability and limitations of the methods used in ST104, important extensions of these methods.

Assessment

30% by assignments,
70% by open book examination at the end of Term 1.

Past Examinations

1993 exam

1994 exam

1995 exam

1996 exam

1997 exam

1998 exam (+specimen solutions)

1999 exam (+specimen solutions) (+comments)

Books

Chatfield (1995) "Problem Solving"

Daly et al. (1995) "Elements of Statistics"

Moore & McCabe (1999) "Introduction to the Practice of Statistics"

Example is the school of mankind, and they will learn at no other.

Edmund Burke, Letters on a Regicide Peace, I.

'Why', said the Dodo, 'the best way to explain it is to do it.'

Lewis Carroll (Charles Lutwidge Dodgson), Alice in Wonderland.