This tutorial teaches use of one-way ANOVA, a statistical technique for testing mean differences betweeen three or more independent groups on a single dependent variable. Practical exercises are based on using SPSS.
With multiple comparisons, if they are all independent (which is unlikley) in a series of 100 tests we would expect to get five Type I errors with a .05 level of significance
By simultaneously computing all possible comparisons in a single significance test, the ANOVA avoids these inflated error rates
However, use of one-way ANOVA limits error rates at the expense of specificity - statistic tells us that there is a significant difference somewhere among the sample means, but does not tell us which means differ significantly (have to use post-hoc and a priori comparison procedures)
Experimental study: Examine reaction time under different levels of alcohol consumption by randomly assigning participants to four conditions (none, low, medium, and high alcohol)
Quasi-experimental study: Examine whether students with behaviour problems behave better in classrooms where teachers have a humanistic philosophy and have control of their classrooms. Classify teachers as (1) humanists with control, (2) strict disciplinarians, and (3) laissez-faire.
Error-bar graph showing mean pulse rates and 95% confidence intervals by exercise level.
Use any dataset
Conduct a one-way ANOVA and graphically present the means and confidence intervals using an Error Bar Graph - is this error bar chart consistent with the statistical results?