Home

Statistical tests

• Unpaired (independent) t test
• paired t test
• Tutorial on using t tests
• Chi-squared test

• sample problems
• solutions to sample problems

Tables

• Critical values (chi-square )
• Critical values (t test)

Selected Statistical Methods for the Biosciences

Statistics are designed to draw conclusions from data that are subject to experimental error. When replicate sampling is necessary, a carefully chosen statistical analysis can save the investigator the trouble of performing unnecessary experiments, or the embarrassment of drawing a premature or inappropriate conclusion. Statistics generally yield a probability value for a particular outcome. As a rule, scientists accept a probability of 0.05 or less as convincing evidence that a particular outcome is unlikely.

Information on selected types of statistical analysis is presented on these pages. Do not expect a rigorous explanation of the theory behind each type of analysis. The articles will focus on an understanding of the principles behind the analyses, and on their proper use.