hypothesis-testing

# Hypothesis Testing Workflow

This skill guides you through rigorous hypothesis development and testing methodology.

## Phase 1: Observation and Question

### Starting Point Analysis
- What observation or phenomenon prompted this inquiry?
- What patterns or anomalies are you seeing?
- What existing knowledge is relevant?

### Research Question Formulation
Good research questions are:
- **Focused**: Specific enough to answer
- **Researchable**: Can be investigated empirically
- **Complex**: Requires analysis, not just facts
- **Arguable**: Has multiple possible answers

### Question Types
| Type | Example | Hypothesis Style |
|------|---------|------------------|
| Descriptive | "What is X?" | Not hypothesis-driven |
| Relational | "Is X related to Y?" | Correlation hypothesis |
| Causal | "Does X cause Y?" | Causal hypothesis |
| Comparative | "Is X different from Y?" | Difference hypothesis |

**CHECKPOINT**: Confirm research question with user.

## Phase 2: Hypothesis Construction

### Hypothesis Components
```
If [independent variable/condition]
Then [dependent variable/outcome]
Because [theoretical mechanism]
```

### Null vs Alternative Hypothesis
- **H₀ (Null)**: No effect/relationship exists
- **H₁ (Alternative)**: Effect/relationship exists

Example:
- H₀: Training method has no effect on performance
- H₁: Training method improves performance

### Hypothesis Quality Check
- [ ] Is it testable with available methods?
- [ ] Is it falsifiable (can be proven wrong)?
- [ ] Does it make specific predictions?
- [ ] Is it parsimonious (simplest explanation)?
- [ ] Is it consistent with existing knowledge?
- [ ] Does it specify the mechanism?

## Phase 3: Variable Mapping

### Variable Identification
| Variable | Type | Operationalization |
|----------|------|-------------------|
| [Name] | Independent (IV) | [How measured/manipulated] |
| [Name] | Dependent (DV) | [How measured] |
| [Name] | Control | [How held constant] |
| [Name] | Confound | [Potential interfer