Study Guide:    Hypothesis testing

Produced by Graham Currell, University of the West of England, Bristol and:
Royal Society of Chemistry, 'Discover Maths for Chemists' website
, and

Essential Mathematics and Statistics for Science, 2nd Edition
Graham Currell and Antony Dowman, Wiley-Blackwell, 2009

This study unit aims to develop the skills for peforming hypothesis testing of experimental data. This is achieved by references given to sections and pages in the study text, Essential Mathematics and Statistics for Science, associated QVA tutorials (questions with video worked answers), and tutorials for the use of MS excel.
NB: The video tips and QVAs appear in 'pop-up' windows. If your browser blocks 'pop-ups', you can use CTRL+Enter or Right click with the mouse.

Performing a Hypothesis Test. Study Text: Chapter 9 (p243ff)
Hypotheses, p-values & significance level: QVA tutorial

A hypothesis test assesses the probability that a relationship that you observe in your experiment is a true description of the real world, and not just a statistical variation in your experiment.

You will need to state your hypotheses:
• Null Hypothesis, HO, which is usually a statement that there is NO effect or relationship, and, at least one
• Alternative (or Proposed) Hypothesis, H1, that an effect or relationship DOES exist.

The calculated p-value gives the probability that you would be wrong if you decide that H1 is correct.
The term ‘significance’ gives a probability limit, chosen by you, for deciding the observed effect is actually a true effect.

Testing with: (link to Statistical tables for critical values)

t-tests for differences in mean values. Study Text: Sections 10.1,2,3 (p262)
- using the t-statistic: QVA tutorial
- using p-values (Excel and Minitab): QVA tutorial

F-tests for differences in variances. Study Text: Sections 10.4 (p274)
- using the F-statistic, p-values (Excel and Minitab): QVA tutorial

Chi-squared, X2-tests for differences in frequency tables. Study Text: Sections 14.1,2 (p331)
- using the X2 statistic, p-values (Excel and Minitab): QVA tutorial

Tests for differences in proportion. Study Text: Sections 14.3 (p343)
- using Minitab for testing one and two proportions
- using p-values and confidence intervals (Minitab): QVA tutorial

Tests for normality in data distributions. Study Text: Section 9.5.2 (p256)
QVA tutorial  - video feedback in preparation

Excel tutorial files for download: (video commentaries in preparation)
t-tests, F-tests & correlation: Parametric Tests (Excel 2007)
Chi-squared tests: Chi-Squared (Excel 2007)
ANOVAs: ANOVAs (Excel 2007)

This new Study Guide is in the process of development - any comments, corrections or suggestions welcome: graham.currell@uwe.ac.uk