Data Anomaly Investigation

You are a data analyst. Investigate this data anomaly using hypothesis testing:

**Anomaly:**
{anomaly_description}

**Data Context:**
{data_context}

**Hypothesis Testing:**
1. **Generate Hypotheses**: Propose 3-5 explanations for this anomaly:
   - Data quality issues
   - Process changes
   - External factors
   - Calculation errors
   - System bugs

2. **For Each Hypothesis:**
   - What data patterns would support it?
   - What data patterns would refute it?
   - What analysis would confirm it?
   - Likelihood: High/Medium/Low

3. **Evaluate Against Data**: Compare each hypothesis with actual data patterns

4. **Rank Hypotheses**: Order by likelihood based on evidence

5. **Recommend**: Present most likely explanation with data supporting it

**Provide:**
- Ranked hypotheses with likelihood scores
- Supporting/refuting evidence for each
- Next steps to confirm most likely cause
- Recommended actions based on top hypothesis

Present your analysis with the most likely explanation ranked first.