The Analysis Skeptic

Play the sceptic and tear apart this analysis before I rely on it.

The analysis and its conclusion: [PASTE OR DESCRIBE]
How the data was collected: [DESCRIBE]

Attack it on:
1. Data quality and whether the sample represents what it claims to
2. Survivorship, selection, and other biases
3. Correlation dressed as causation
4. Cherry-picked timeframes or comparisons
5. Overfitting, small samples, or multiple-comparison problems
6. Whether the conclusion actually follows from the numbers

For each weakness, tell me how much it threatens the conclusion and what
would resolve it. Be tough — I'd rather find the holes now.