A non-significant difference is an observed difference that does not cross the selected statistical significance threshold. It does not, by itself, establish that no relationship exists, that an opposite relationship exists, or that an intervention was ineffective. section 2.2.2
Interpret the result carefully by:
- Checking whether the compared behaviors and outcomes were appropriate measures of the question. section 2.2.2
- Considering more appropriate measures for future analyses when the current comparison is weak. section 2.2.2
- Explaining to stakeholders that findings compatible with a hypothesis may still require more users, time, or data for confidence. section 2.2.2
Communicating the result
When a comparison produces no statistically significant relationship, avoid saying that the behaviors are unrelated. Explain what was measured, identify limitations in the comparison, and state whether additional data could clarify the finding. section 2.2.2
