H3 Including statistics of effect size and confidence intervals

Once you know how to calculate these and what they mean, report relevant statistics of effect size as well as the obtained value of your statistic and its associated probability. The example RESULTS in section 12.3.2 of the book includes reports of partial eta squared, which is one effect size statistic for F. You can usually get your statistics software package to generate relevant effect size statistics for the specific tests that you run. (For more on effect size see Chapter 12 of the book and Section B5 of this Web site.) Wherever you can, also produce and include in text or table the relevant confidence intervals. The 95% confidence intervals are commonly used. Table 1, in Section 4.6.10 of the book contains an example of how to report such confidence intervals. Whichever confidence interval size you choose, you should use the same one throughout the report. (For more on confidence intervals see Section 12.3.1 of the book.)