DCED Sample Size Calculator (FPC & Diff-in-Diff)
Calculate exact survey sample sizes for market systems monitoring and endline evaluations using Cochran sampling, Finite Population Correction (FPC), cluster design effect, non-response buffers, and comparison group ratios.
Field Practice Tools & Market Evidence
Search 2,523 evaluated MSD tools, donor programmes, DCED guidance pages, and Social Enterprise publications. Check commercial dealer margins, verify causal attribution, or audit a draft project idea in real time.
Plan DCED survey sample size with Finite Population Correction, cluster effects, partner Will-Skill fit, and AAER change stages.
How the DCED Sample Size Formula Works
Donor audits under the DCED Standard look for defensible sample sizes rather than arbitrary round numbers. This calculator runs four sequential steps in your browser with zero login required.
1. Cochran Base (n0)
Computes initial random sample size from Z-score (90%, 95%, or 99% confidence), expected indicator prevalence p, and margin of error e.
2. Finite Population Correction (FPC)
Adjusts sample size downward when your total population N (for example 1,200 farmers or 5,000 shops) is small enough that sampling >5% changes variance.
3. Cluster Design Effect (Deff) & Buffer
Multiplies by design effect (typically 1.3x to 2.0x for village cluster surveys) and adds a non-response buffer (10% to 15%) for field dropouts.
4. Comparison / Control Sample
Adds a matched difference-in-differences comparison group ratio (for example 50% or 100% of treatment interviews) to isolate causal impact.
Frequently Asked Questions
What confidence level and margin of error does the DCED Standard require?
Most DCED-audited programmes target 95% confidence and a +/-5% margin of error for core enterprise and smallholder income indicators, though small pilot interventions often use 90% confidence with +/-7% or +/-10% precision.
When should I use a cluster design effect (Deff) above 1.0?
Use a design effect of 1.5x to 2.0x whenever enumerators interview groups of farmers or households inside selected villages instead of pulling a pure simple random sample across all villages.
How do I export the sample size table for a donor proposal?
Use the Copy Proposal Annex (Markdown), Download .md Brief, or Print / Save 1-Page PDF Field Sheet buttons right below the results panel.
Run these tools inside Claude, Cursor, or your own scripts via MCP (https://msd.linn.services/mcp, Registry ID: services.linn/msd), browse 2,523 indexed guides, or check client examples on GitHub.