Experiment Design
Design rigorous experimental protocols to validate research hypotheses.
Protocol Components
For each hypothesis, produce a complete protocol with:
| Field | Description |
|---|---|
| Design | Experimental design type (RCT, quasi-experimental, longitudinal, in silico, etc.) |
| Independent Variable (IV) | What is manipulated |
| Dependent Variable (DV) | What is measured |
| Controls | Variables held constant |
| Sample Size | N per condition + power analysis (α=0.05, β=0.80, effect size) |
| Timeline | Phase-by-phase schedule |
| Protocol | Step-by-step procedure |
| Expected Outcome | What would confirm vs. refute the hypothesis |
Power Analysis
Always include a power analysis. Standard parameters:
- α (Type I error rate): 0.05
- Power (1−β): 0.80
- Effect size: use domain-specific estimates or Cohen's conventions (small=0.2, medium=0.5, large=0.8)
Example: 30 simulations per condition (90 total, power analysis: α=0.05, β=0.80, η²=0.25)
Output Format
markdown# 🧫 Experiment Design ## Strategy Overview [2–3 sentences: how the experiments collectively test the hypotheses] ## Proposed Experiments ### Experiment 1: [Design Type] **Tests Hypothesis:** [Exact hypothesis being tested] | Parameter | Detail | |-----------|--------| | **Design** | [Design type] | | **Sample Size** | [N per condition with power analysis] | | **Timeline** | [X months: phase breakdown] | **Independent Variables:** [What is manipulated] **Dependent Variables:** [What is measured] **Control Variables:** [What is held constant] **Protocol:** 1. [Step 1] 2. [Step 2] ... **Expected Outcome:** [What confirms the hypothesis. What would refute it.] ---
Feasibility check
Before finalizing a protocol, consider:
- Can the required equipment/resources be co-located?
- Is the timeline realistic for the sample size?
- Are controls sufficient to rule out confounds?
- Does the analysis plan address multiple comparisons if >1 DV?
