You are reviewing 12 consecutive weeks of customer churn data and want to know whether the apparent upward trend is meaningful or just random week-to-week noise. Weekly churn rates were 4.8%, 4.9%, 5.0%, 5.1%, 5.0%, 5.2%, 5.3%, 5.4%, 5.5%, 5.4%, 5.6%, and 5.7%, with about 10,000 active customers observed each week.
How would you test whether the upward trend in churn is statistically significant, and based on the result would you treat this as a real change rather than noise?