How can a small difference be statistically significant, while larger results may not be statistically significant?
Results presented in the HSTS are based on samples of students rather than on entire populations. For this reason, there is a margin of error around each score. The determination of whether a difference between two variables is statistically significant is based on these margins of error. When a result has a large standard error, a numerical difference that seems large may not be statistically significant. Margins of error are usually larger for smaller population groups. For example, a difference of 0.05 in grade point average between male and female graduates that has a small margin of error may be statistically significant, while a difference of 0.10 in grade point average between Black and Hispanic graduates that has a large margin of error may not be significant.
Related Questions
- What concerns about protocol and procedural issues have been expressed by critics? How might the statistically significant results be "explained away"?
- My software gives results which disagree with the NIST certified results in the __ significant digit. Should this difference be worrisome?
- Are the Survey Results Statistically Significant?
