Steps and Confidence Intervals in Hypothesis Testing

This section lists four key steps in hypothesis testing and explains the close relationship between confidence intervals and significance tests.

Significance Testing and Confidence Intervals


  1. You cannot reject the null hypothesis because the confidence interval shows that 20 is a plausible value of the population parameter.

  2. This study is testing the difference between means, and significant differences would be either larger or smaller than 0. Thus, confidence intervals that do not contain 0 represent statistically significant findings.

  3. The 99% confidence interval contains all of the values that the 95% confidence interval has, but it extends farther at both ends and has other values, too. If something is not significant at the .05 level, it is also non-significant at the .01 level.

  4. Because the p value was .21, we know that the 95% confidence interval contains the null hypothesis parameter, .5. Thus, both of the confidence intervals that contain .5 are possible confidence intervals that this researcher could have computed.