Survey Sample Size Calculator
Calculate sample size for campus and population surveys with finite population correction and response rates. Pre-configured for realistic survey scenarios—works offline, no signup required.
When you have no prior estimate, use 50% (the assumption that maximizes required sample size, making your calculation conservative).
Realistic for email and online campus surveys. Adjust based on your recruitment method and population.
§2The formula
- n₀
- Initial sample size (without finite population correction)
- z
- Critical z-value for your chosen confidence level
- p
- Expected proportion (0 to 1); use 0.5 if unknown (most conservative)
- e
- Margin of error as a decimal (e.g., 0.05 for 5%)
- Finite population correction:
- n = n₀ / (1 + (n₀ − 1) / N)
- With non-response:
- Recruit = n / (1 − non-response rate)
§3Worked example: Planning a campus survey
Suppose you are the student government running a survey on academic support services usage at a mid-size university with 8,000 enrolled students. You want 95% confidence, a margin of error of 5%, and assume 50% use the services (conservative). You expect a 15% dropout rate via email recruitment.
Step 1: Initial sample size (without correction)
Round up to 385 responses.
Step 2: Apply finite population correction
Your sample of 385 represents 385 / 8000 = 4.8% of the campus population. This is just below the 5% threshold, but we apply the correction anyway since you specified a finite population.
The correction reduces your requirement by about 18 responses (4.7%).
Step 3: Adjust for non-response
Email surveys to students historically achieve 15% response rates. To ensure you collect 367 completed responses, you must invite:
Send invitations to approximately 432 students to expect 367 completions, which after finite population correction gives you the sample size needed to estimate campus opinion with ±5% margin of error at 95% confidence.
Why the finite population correction made a difference here
The campus has 8,000 students, and 385 is 4.8% of that population—close to the 5% threshold. At smaller institutions (under 3,000 students) or larger samples, the FPC becomes more substantial. At very large universities (30,000+), the FPC has minimal effect because your sample remains a tiny fraction of the population.
§4When the finite population correction matters
The 5% rule of thumb
The finite population correction (FPC) adjusts your sample size downward when your sample will be a non-negligible fraction of a finite population. A practical guideline: if your sample size n₀ is greater than 5% of your population N, apply the FPC. When n₀ ≤ 5% of N, the correction has minimal effect (typically less than 2% change) and you can often omit it.
| Scenario | Population (N) | Calculated n₀ | % of N | Use FPC? |
|---|---|---|---|---|
| Large university | 30,000 | 385 | 1.3% | No (negligible effect) |
| Mid-size campus (this example) | 8,000 | 385 | 4.8% | Yes (4.7% reduction) |
| Small college | 2,000 | 385 | 19.3% | Yes (16% reduction) |
| Department survey | 300 | 169 | 56% | Yes (36% reduction) |
If your population is unknown or very large (national survey), leave the population field blank and the calculator will skip the FPC.
§5Survey response rates and realistic dropout assumptions
Typical response rates by method
Email surveys on campus average 15–25% completion. In-person recruitment (tabling on campus, classroom visits) can reach 40–60%. Incentives (raffle entry, course credit) boost response rates by 10–20 percentage points. Online survey links embedded in campus email tend to see lower response than surveys embedded in learning management systems (LMS). Always plan conservatively and consult prior surveys at your institution for a reality check.
The calculator pre-fills the non-response rate at 15%, a realistic default for email and web surveys on campus. If your recruitment method differs—for example, you'll visit residence halls in person—reduce this figure to 10% or lower. If incentives are unavailable and your topic is not mandatory, increase it to 25% or higher.
§6Confidence level and margin of error
What 95% confidence really means
A 95% confidence level does not mean there is a 95% chance your true population proportion is in your calculated interval. Instead: if you ran this survey many times using the same method, about 95% of the confidence intervals you computed would contain the true proportion. It is a statement about the procedure, not the probability after one survey.
Margin of error
A ±5% margin of error means if you estimate 60% of students use tutoring services, the true proportion is likely between 55% and 65%. Smaller margins require larger samples—the relationship is inverse. To cut your margin in half (from 5% to 2.5%), you need roughly 4 times as many participants.
§7Assumptions and pitfalls
Simple random sampling
These formulas assume every person in your population has an equal chance of being selected. If your survey targets only residential students, or only those who read campus email, you have a biased sampling frame. The resulting confidence interval may not represent the whole campus even if your sample size is correct.
Response bias
Achieving your target sample size does not guarantee accuracy if respondents differ systematically from non-respondents. Students motivated to fill out surveys may differ in their views from those who ignore the invitation. Use follow-up reminders and consider open-ended questions to understand who did not respond.
Question wording
Your sample size calculation is based on the assumption that your survey questions are clear and unbiased. Poorly worded questions will generate noise that no sample size can overcome. Pre-test your questions with 5–10 students before the full launch.
§8FAQ
What response rate should I plan for in an email survey?
Email surveys to campus populations typically see 15–25% response rates, depending on the topic, time of year, and whether incentives are offered. Mandatory surveys (e.g., embedded in an LMS assignment) may reach 60–80%. Always collect baseline data: ask a colleague who has run a similar survey on your campus what response rate they achieved, then plan conservatively (assume 5 percentage points lower). If your institution has no baseline, use 15% as the default.
Do I really need to collect 432 invitations if I only need 367 responses?
Yes, if you expect 15% completion. That accounts for unopened emails, deleted messages, and busy students who see the invitation but don't respond. In practice, your actual response rate may be higher or lower than 15%—keep the calculator open during your survey period and monitor your response count in real time. If by the deadline you have only 300 responses and need 367, you can either extend the survey period or accept a slightly larger margin of error.
Why does finite population correction reduce my sample size? Is that a mistake?
It is not a mistake. The finite population correction reflects a real reduction in sampling error when your sample is a large fraction of a small population. If a campus has exactly 500 students and you survey all 500, you have zero sampling error (you have the whole population). If you survey 250, you have already captured so much of the population's diversity that each additional response contributes less new information. The correction accounts for this diminishing return.
What if my survey targets a subgroup (e.g., only engineering majors)?
Use the size of that subgroup as your population. If your campus has 8,000 students but only 600 are engineering majors and you are surveying only them, enter N = 600. Your sample size will be calculated relative to that smaller population, and the finite population correction will have a larger effect.
Can I use this calculator if I am surveying two groups and comparing them?
This calculator estimates population proportions. If your goal is to compare two groups (e.g., first-years vs. seniors), use the sample size calculator for comparing proportions instead. That calculator is designed for hypothesis testing with two independent samples.
What if my survey is online (through Qualtrics or Google Forms)?
The principles are the same. Response rates for web-based surveys depend on how you recruit. If you embed it in an LMS, expect 40–70% completion. If you send a mass email with a link, expect 15–25%. The dropout/non-response field in the calculator accounts for this.
§9See also
This calculator is specialized for surveys. For other study designs, see:
- Sample Size Calculator (base) — for means, comparing two means, and comparing two proportions.
- Sample Size Calculator for Experiments — for hypothesis testing and power analysis.
- Which Statistical Test Should I Use? — to determine your test before calculating sample size.
§10Sources
- OpenStax Introductory Statistics, 2e — Open-access textbook covering sampling distributions, confidence intervals, finite population corrections, and survey design principles.
- Cochran, W. G. (1977). Sampling Techniques (3rd ed.). Wiley. — Classical reference for survey sampling formulas and finite population corrections in small-population settings.
- UCLA OARC: Sample Size — University of California tutorial on sample size calculation and response rate planning.
- SurveyMonkey: Choosing a Survey Sample Size — Practical guidance on sample size and response rates in real-world surveys.