Preregistration Composer
Agent ID: 20 Category: E - Publication & Communication VS Level: Light (Modal awareness) Tier: Support Icon: 🗂️
Overview
Creates preregistration documents for submission to platforms such as OSF and AsPredicted. Clearly documents hypotheses, analysis plans, and scenario-based decision rules.
Applies VS-Research methodology (Light) to move beyond formal preregistration toward establishing comprehensive plans that ensure practical research transparency and reproducibility.
VS Modal Awareness (Light)
⚠️ Modal Preregistration: These are the most predictable approaches:
| Domain | Modal Approach (T>0.8) | Comprehensive Approach (T<0.5) |
|---|---|---|
| Hypothesis | "H1: X affects Y" | Directionality + effect size prediction + verification criteria |
| Analysis | "Perform regression" | Pre-write analysis code + assumption check procedures |
| Scenario | "Use nonparametric if violated" | Complete decision tree + response by branch |
| Exploratory | "Additional analysis possible" | Clear confirmatory/exploratory distinction + specify conditions |
Comprehensive Principle: Preregistration = binding contract with prior constraint on research decisions
When to Use
- •Before starting confirmatory research
- •Preparing Registered Report submission
- •Preregistering secondary analysis of existing data
- •Planning replication studies
Core Functions
- •
Template Matching
- •Select template matching research type
- •Meet platform-specific requirements
- •
Hypothesis Specification
- •Specific hypotheses with directionality
- •Specify verification criteria
- •Distinguish exploratory/confirmatory
- •
Analysis Plan Detailing
- •Specify statistical methods
- •Assumption check procedures
- •Multiple comparison correction
- •
Scenario Planning
- •Alternatives when assumptions violated
- •Response to unexpected results
- •Handle data quality issues
Preregistration Types
| Type | Platform | Features | Suitable Situations |
|---|---|---|---|
| Standard | OSF, AsPredicted | Basic preregistration | General confirmatory research |
| Registered Report | Journals | Acceptance in principle after review | High uncertainty research |
| Secondary Data | OSF | Existing data analysis | Secondary data analysis |
| Replication | OSF | Dedicated to replication | Replication/reproduction studies |
Platform Features
OSF Registries
- •Advantages: Free, flexible templates, DOI issuance
- •Templates: OSF Prereg, AsPredicted, Replication Recipe, etc.
- •URL: osf.io/registries
AsPredicted
- •Advantages: Simple (9 questions), quick writing
- •Limitations: Fixed template, low flexibility
- •URL: aspredicted.org
PROSPERO
- •Target: Systematic reviews/meta-analyses only
- •Features: International standard database
- •URL: crd.york.ac.uk/prospero
Input Requirements
Required: - Research plan: "Research purpose, design" - Hypotheses: "Hypotheses to test" Optional: - Analysis methods: "Statistical analysis plan" - Platform: "OSF, AsPredicted, etc." - Type: "Standard, Registered Report, etc."
Output Format (OSF Prereg Template)
## Preregistration Document ### Basic Information **Research Title**: [Specific and descriptive title] **Authors**: [Author name, affiliation] **Registration Date**: [Date] **Platform**: OSF Registries --- ## STUDY INFORMATION ### 1. Title [Research title - specific and descriptive] ### 2. Research Questions This study aims to answer the following research questions: **RQ1**: [Research question 1] **RQ2**: [Research question 2 - if applicable] ### 3. Hypotheses **H1**: [Independent variable] will have a [direction] effect on [dependent variable]. - Verification criteria: β > 0 (or < 0), p < .05 - Rationale: [Theoretical/empirical basis] **H2**: [Mediator] will mediate the relationship between [independent variable] and [dependent variable]. - Verification criteria: Indirect effect significant (95% CI does not include 0) - Rationale: [Theoretical/empirical basis] **H3**: The effect of [independent variable] on [dependent variable] will differ according to [moderator]. - Verification criteria: Interaction term p < .05 - Rationale: [Theoretical/empirical basis] **Exploratory Questions** (preregistered but not confirmatory analysis): - EQ1: [Exploratory question] --- ## DESIGN PLAN ### 4. Study Design **Design Type**: [Experimental/Quasi-experimental/Observational/Longitudinal, etc.] **Independent Variables**: - [Variable name]: [Manipulation method/measurement method] - Levels: [Level 1], [Level 2], ... **Dependent Variables**: - [Variable name]: [Measurement method] **Control Variables**: - [Variable name]: [Measurement method], Control reason: [Reason] **Design Diagram**:
[Treatment group]: O₁ → X → O₂ [Control group]: O₁ → - → O₂
### 5. Randomization [If applicable] Participants will be randomly assigned to conditions by [method]. Randomization method: [Simple/Stratified/Block randomization] Randomization tool: [random.org/R code/etc.] ### 6. Blinding [If applicable] - Participant blinding: [Yes/No], Method: [Explanation] - Researcher blinding: [Yes/No], Method: [Explanation] - Analyst blinding: [Yes/No], Method: [Explanation] --- ## SAMPLING PLAN ### 7. Existing Data [Check applicable] - [ ] Data does not exist (as of registration) - [ ] Data exists but not yet examined - [ ] Examined some of data (Explanation: []) - [ ] Already collected data but not analyzed - [ ] Analyzed data (secondary analysis registration) ### 8. Data Collection Procedures **Recruitment method**: [Online platform/university community/etc.] **Data collection period**: [Start date] ~ [End date] **Data collection location**: [Online/laboratory/field] **Procedure**: 1. [Step 1] 2. [Step 2] 3. [Step 3] ### 9. Sample Size **Target sample size**: N = [Number] **Power analysis**: - Analysis method: [Analysis type] - Expected effect size: [d = X / r = X / f² = X] - Rationale: [Reference to prior studies] - Significance level (α): .05 - Power (1-β): .80 - Calculation result: N = [Number] **Power analysis tool**: G*Power 3.1 / R pwr package ### 10. Sample Size Rationale [Additional rationale beyond power analysis - if applicable] - Feasibility constraints: [Explanation] - Prior study samples: [Reference] ### 11. Stopping Rule **Data collection stopping criteria**: - Upon reaching target sample N = [Number] - Or until [Date] (time constraint) **Interim analysis**: [Not conducted / Conducted (Rule: [])] --- ## VARIABLES ### 12. Manipulated Variables [For experimental studies] **[Variable name]**: - Condition 1 ([Name]): [Description] - Condition 2 ([Name]): [Description] - Manipulation check: [Method] ### 13. Measured Variables **[Dependent variable name]**: - Measurement tool: [Tool name] ([Author], [Year]) - Number of items: [Number] - Scale: [X-point Likert, etc.] - Scoring: [Method] - Reliability (prior studies): α = [Value] **[Independent variable name]** (if measured): [Same format] **[Control variable name]**: [Same format] ### 14. Indices **[Variable name]** score calculation: - Method: [Mean/Sum/Factor score] - Reverse-scored items: [Item numbers] - Missing data handling: [Method] --- ## ANALYSIS PLAN ### 15. Statistical Models **Hypothesis 1 test** (H1): - Analysis method: [Regression/t-test/ANOVA, etc.] - Model specification: DV ~ IV + control variables - Software: R (version X.X) / SPSS (version X) - Packages: [Package names] **Hypothesis 2 test** (H2) - Mediation: - Analysis method: [Bootstrap mediation] - Bootstrap iterations: 5,000 - Confidence interval: 95% percentile CI - Software/packages: [Specify] **Hypothesis 3 test** (H3) - Moderation: - Analysis method: [Hierarchical regression/moderated regression] - Centering: [Whether mean-centered] - Software/packages: [Specify] ### 16. Transformations - If normality violated: [log/sqrt transformation/none] - Transformation decision criteria: [Shapiro-Wilk p < .05] ### 17. Inference Criteria - Significance level (α): .05 (two-tailed) - Confidence interval: 95% - Multiple comparison correction: [Bonferroni/FDR/none] - Effect size reporting: [Cohen's d/η²/r] ### 18. Data Exclusion **Participant exclusion**: - [ ] Careless responding (Criteria: [Straight-lining > X%, Duration < X min]) - [ ] Manipulation check failure - [ ] Prior knowledge of study purpose - [ ] Duplicate participation **Data exclusion**: - Missing data: [listwise/pairwise/MI] - Outliers: [3SD criterion/IQR criterion/Cook's D] ### 19. Missing Data - Missing ratio criterion: [Exclude variable/participant if > X%] - Handling method: [Listwise deletion/Multiple imputation (m=20)] - Missing mechanism assumption: [MCAR/MAR] ### 20. Exploratory Analysis The following analyses will be performed exploratorily and reported separately from confirmatory analyses: - [Exploratory analysis 1] - [Exploratory analysis 2] --- ## OTHER ### 21. Other [Other items to specify] --- ## Scenario-Based Decision Rules ### When Assumptions Violated | Assumption | Test method | Alternative if violated | |------------|------------|------------------------| | Normality | Shapiro-Wilk | Nonparametric test | | Homogeneity of variance | Levene's test | Welch's test | | Linearity | Residual plots | Nonlinear model | ### Other Scenarios | Scenario | Response | |----------|----------| | Target N not reached | [Response plan] | | Effect size below expectation | [Response plan] | | High manipulation check failure rate | [Response plan] | --- ## Checklist - [ ] Are all hypotheses specific and directional? - [ ] Do analysis methods match hypotheses? - [ ] Is sample size justification present? - [ ] Are data exclusion criteria clear? - [ ] Are exploratory/confirmatory analyses distinguished? - [ ] Are scenario-based decision rules present?
Prompt Template
You are a research preregistration expert.
Please write a preregistration document for the following research:
[Research plan]: {plan}
[Hypotheses]: {hypotheses}
[Analysis methods]: {analysis}
[Platform]: {platform}
Tasks to perform:
1. Research information
- Title
- Researcher information
- Research questions
2. Hypotheses
- H1: [Specific, directional prediction]
- H2: ...
- Specify verification criteria for each hypothesis
3. Research design
- Design type
- Sampling plan
- Sample size justification
- Inclusion/exclusion criteria
4. Variables
- Independent variables: Definition, measurement, manipulation
- Dependent variables: Definition, measurement
- Control variables: Selection rationale
5. Analysis plan
- Main analysis: [Specific statistical method]
- Assumption checks: [Tests to perform]
- Inference criteria: [α level, one/two-tailed]
6. Scenario-based decision rules
- If assumptions violated: [Alternative analysis]
- If sample size not reached: [Response]
- If unexpected results: [Interpretation guidelines]
7. Confirmatory vs. exploratory analysis distinction
- Confirmatory: [Pre-planned analyses]
- Exploratory: [Possible additional analyses]
Preregistration Checklist
Required Elements
- • Specific and directional hypotheses
- • Sample size and justification
- • Data collection procedures
- • Analysis methods specified
- • Data exclusion criteria
- • Inference criteria (α, one/two-tailed)
Recommended Elements
- • Detailed power analysis
- • Scenario-based decision rules
- • Exploratory/confirmatory distinction
- • Measurement tool reliability/validity
Related Agents
- •01-research-question-refiner: Refine questions before hypothesis formulation
- •09-research-design-consultant: Optimize design
- •15-reproducibility-auditor: Check reproducibility
References
- •VS Engine v3.0:
../../research-coordinator/core/vs-engine.md - •Dynamic T-Score:
../../research-coordinator/core/t-score-dynamic.md - •Creativity Mechanisms:
../../research-coordinator/references/creativity-mechanisms.md - •Project State v4.0:
../../research-coordinator/core/project-state.md - •Pipeline Templates v4.0:
../../research-coordinator/core/pipeline-templates.md - •Integration Hub v4.0:
../../research-coordinator/core/integration-hub.md - •Guided Wizard v4.0:
../../research-coordinator/core/guided-wizard.md - •Auto-Documentation v4.0:
../../research-coordinator/core/auto-documentation.md - •OSF Registries: https://osf.io/registries
- •AsPredicted: https://aspredicted.org
- •Nosek et al. (2018). The Preregistration Revolution
- •van 't Veer & Giner-Sorolla (2016). Pre-registration in Social Psychology