Revised City of South Bend Disparity Study Report

Home Browse Search

City of South Bend Disoarity Study 2020 the null hypothesis. We then calculate a confidence interval to find the proba- bility that the observed relationship (e.g., -35 percent) is between 0 and minus that confidence interval .171 The confidence interval will vary depending upon the level of confidence (statistical significance) we wish to have in our conclu- sion. When a number is statistically significant at the 0.001 level, this indicates that we can be 99.9 percent certain that the number in question (in this exam- ple, -35 percent) lies outside of the confidence interval. When a number is sta- tistically significant at the 0.01 level, this indicates that we can be 99.0 percent certain that the number in question lies outside of the confidence interval. When a number is statistically significant at the 0.05 level, this indicates that we can be 95.0 percent certain that the number in question lies outside of the confidence interval. 171. Because 0 can only be greater than -35%, we only speak of "minus the confidence level". This is a one -tailed hypothesis test. If, In another example, the observed relationship could be above or below the hypothesized value, then we would say "plus or minus the confidence level" and this would be a two -tailed test. 108 0 2020 Colette Holt & Associates, All Rights Reserved.