If a test shows a statistically significant decrease in impressions, what does this imply?

• A. The impressions will surely rebound
• B. There is a true effect worth noting at the specified confidence level
• C. The test is invalid
• D. The traffic should be redirected

Answer: (C)

When a result is statistically significant, it means the observed change is unlikely to be due to random variation alone, given the null hypothesis and the chosen alpha level. In this case, a significant decrease in impressions indicates there is evidence of a real effect affecting impressions at the specified confidence level. It doesn't say the test is invalid, nor does it guarantee that impressions will rebound or that traffic should be redirected. The key takeaway is that the data support a real decrease, though you’d still assess practical significance and potential causes before making decisions.