Writing Your Research Proposal

This guide helps you write a proposal that will move smoothly through the review process.

Proposal Components

Every SafeInsights proposal includes:

1. Research Question

State your primary research question clearly and concisely.

Good example: "How does the frequency of accessing worked examples relate to performance on similar practice problems in college physics?"

Too vague: "What factors affect learning?"

2. Background & Significance

Briefly explain:

Why this question matters (2-3 sentences)
How it connects to existing literature (cite 3-5 key papers)
Expected contribution to the field

3. Target Population

Specify:

Which SafeInsights data organization(s)
Education level (e.g., college STEM courses)
Subject area
Time period
Estimated sample size

4. Variables & Measures

List:

Independent variables: What you're examining (e.g., frequency of video access)
Dependent variables: What you're predicting (e.g., quiz scores)
Covariates: What you're controlling for (e.g., prior achievement)
How measured: Reference data catalog fields

5. Analytic Approach

Describe:

Methods (e.g., multiple regression, survival analysis)
Statistical tests
Any data transformations or preprocessing
Software/packages you'll use

6. Expected Outputs

Specify exactly what tables, statistics, and plots you'll need:

Summary statistics tables (specify dimensions)
Regression coefficients with standard errors
Plots (e.g., scatterplot of X vs Y with n=10,000)

Be specific about aggregation:

✅ "Correlation coefficient aggregated across all students"
✅ "Mean quiz score by quartile of practice frequency (n>50 per cell)"
❌ "Individual student scores" (not releasable)

7. Privacy Considerations

Explain how you'll protect privacy:

Minimum cell sizes (typically n≥10)
Suppression of small counts
No individual-level data requested

8. Timeline

Provide estimated timeline:

Analyses development: X weeks
Expected execution time: X hours
Analysis after receiving outputs: X weeks

Proposal Template [ADD FROM DESIGN]

[Provide downloadable template or link to one]

# [Study Title]

## Research Question
[Your clear, focused question]

## Background & Significance
[2-3 paragraphs]

## Target Population
- Member organization: [OpenStax / ASSISTments / etc.]
- Education level: [e.g., Undergraduate STEM]
- Subject: [e.g., College Physics]
- Time period: [e.g., Fall 2023 - Spring 2024]
- Estimated N: [X students]

## Variables & Measures
### Independent Variables
- [Variable 1]: [How measured, data source]

### Dependent Variables
- [Variable 1]: [How measured, data source]

### Covariates
- [Variable 1]: [How measured, data source]

## Analytic Approach
[Methods, tests, software]

## Expected Outputs
1. [Specific output 1]
2. [Specific output 2]
[etc.]

## Privacy Considerations
[How you'll ensure privacy compliance]

## Timeline
[Your estimated timeline]

Common Mistakes to Avoid

❌ Too Vague

"I want to study learning patterns" → What specific patterns? On what platform? In what context?

❌ Intervention Language

"I will randomly assign students to..." → Post-hoc analysis only in v1

❌ Individual-Level Outputs

"I need individual student scores to..." → Request aggregated outputs instead

❌ Insufficient Detail

"I'll run a regression" → Which variables? What model? What outputs?

❌ Missing Privacy Plan

No mention of aggregation or cell sizes → Explain privacy protections explicitly

Review Criteria

Your proposal will be evaluated on:

Scientific Merit
- Clear research question
- Appropriate methods
- Feasible with available data
Privacy Compliance
- Adequate protections specified
- Outputs appropriately aggregated
- FEAT considerations addressed
Technical Feasibility
- Data exists to answer question
- Methods match data structure
- Realistic timeline
Alignment with Member Priorities
- Member sees value in the research
- Fits with their research agenda

Example Proposal Excerpts

[Include 2-3 short examples of well-written sections]