R&D Performance Lab
A full R&D performance-evaluation toolkit built on the logic model — from portfolio and inputs through selection, output, and outcome, to a final program scorecard. Upload one program dataset and run any of 21 analyses.
Overview
21 R&D evaluation analyses
The full logic model — inputs, selection, output, outcome, and a final scorecard.
Upload once, analyze many
Load one program dataset and run any analysis without re-uploading.
Export & AI Chat
Download results as CSV, PNG, or Word — or ask AI Chat to interpret your findings.
What you need
What you can do
How it works
One upload, every analysis
Load a single program dataset and it stays available across all 21 analyses — no need to re-upload as you move between portfolio, selection, output, outcome, and scorecard.
Select an analysis
Pick an analysis from the sidebar — grouped into 5 sections. Before any data is loaded, the page opens on an intro screen for that analysis.
Check what it needs
The intro splits inputs into REQUIRED (needed to run) and OPTIONAL ones that refine the result. Use the Guide or load the sample dataset to see the format.
Upload your data
Drag a CSV or Excel file anywhere onto the page — a drop zone appears; release to upload. Multi-sheet Excel is supported, and one upload works across every analysis.
Map your columns
Open the Variables dropdown and match your columns to the fields each analysis needs. Most auto-detect from the names; the result renders as soon as they’re mapped — no Run button.
On an analysis page
At the top of every analysis page you'll find a Variables dropdown — assign your columns to the fields each analysis needs (most fields auto-detect from the column names) — and a Guide button that explains the method. Both stay at the top while you work.
Reading your results
Once it renders, the result comes in two parts: a plain-language Summary with the headline finding and key findings, followed by the detailed output — metric cards, tables, and charts.
AI Chat — ask anything about your result
Once an analysis runs, a circular toggle button appears at the bottom of the results panel. Click it to open AI Chat — the AI already knows which analysis was run and what the result was, so you can ask follow-up questions without re-explaining the context.
Explain results in plain language
Translates output into plain sentences — useful for reviewers and decision-makers who don't read statistics.
Interpret key numbers
Ask about any metric in the result — what it means, whether it's significant, and why it matters.
Suggest next steps
Ask what the result implies for the program and what to consider doing next.
Draft a summary
Request a written paragraph based on the result, ready to copy into an evaluation report.
Analysis sections
R&D Performance Lab organizes its 21 analyses into 5 sections that follow the R&D logic model — inputs, selection, output, outcome, and evaluation.
Portfolio & Inputs
What is in the portfolio and how is funding allocated?
Composition
Portfolio Composition
Break the portfolio down by field, stage, agency, or project type to see its overall shape.
Budget Allocation
Track how the budget is distributed across programs, themes, and time.
Concentration
Funding Concentration
Measure how concentrated funding is across recipients (Gini / HHI) to flag over-reliance.
Leverage & Matching Funds
Assess how much external or matching investment each public dollar mobilizes.
Selection & Process
How are proposals selected and projects executed?
Selection
Proposal Success Rate
Compute award rates by call, field, and applicant type, with trends over time.
Review Score Reliability
Test inter-reviewer agreement and scoring consistency (ICC, Krippendorff’s alpha).
Duplication / Overlap Check
Detect overlapping or redundant projects across the portfolio by similarity.
Process
Execution & Schedule
Compare planned vs actual milestones and budget execution to spot delays and slippage.
Output & Performance
What did the funded research produce?
Research Output
Publication Output
Count and weight publications by volume, venue quality, and citation impact.
Patents & IP
Track patent filings, grants, and other IP generated by funded projects.
Impact Output
Commercialization
Measure licensing, startups, and revenue arising from funded research.
Target Achievement
Compare delivered outputs against the program’s stated targets and KPIs.
Outcome & Effectiveness
Was the spending efficient and did it make a difference?
Efficiency
Cost-Effectiveness
Relate cost to output — cost per publication, patent, or unit of outcome.
DEA Efficiency
Data Envelopment Analysis — benchmark each project against the best-practice efficiency frontier.
Effect
Supported vs Non-supported
Compare funded against comparable unfunded units to isolate the program’s effect.
Economic Impact
Estimate downstream economic returns — output, jobs, and value added.
Evaluation & Scorecard
How does it benchmark, and what is the overall score?
Comparison
Benchmarking
Compare the program against peers or prior periods on key performance metrics.
Logic Model Linkage
Trace the input → output → outcome chain to test whether the logic model holds.
Attribution & Time-Lag
Separate program effects from external factors and account for delayed impact.
Scorecard
Risk / Outlier Flags
Flag projects that are off-track, underperforming, or statistical outliers.
Program Scorecard
Roll up inputs, outputs, and outcomes into one weighted program score.