Sample size determination
Last updated
In qualitative research, sample size is often determined by a subjective, iterative judgment rather than a preset formula.[1] To achieve a desired margin of error W when estimating a proportion, the minimum sample size n can be found by solving 2√(0.25/n) = W/2, which simplifies to n = 4/W² = 1/B², where B is the error bound; for B = 10% one needs n = 100, for B = 5% n = 400, for B = 3% n ≈ 1000, and for B = 1% n = 10000.[2] A table for a two-sample t-test with equal-sized experimental and control groups, a significance level of 0.05, and a desired statistical power can be used to estimate the total number of individuals needed (twice the number given in the table).[3]
Quick Facts
- Field
- Statistics
Facts from the source article.
Key facts and history
For estimating a proportion, a common 95% confidence interval formula uses 1.96 times the square root of 0.25 divided by n.[2] Solving the equation 2√(0.25/n) = W/2 gives the minimum sample size n needed for a desired margin of error W.[2] For a 5% error bound, n = 400 is required; for a 3% bound, n ≈ 1000; for a 1% bound, n = 10,000.[2] The resource equation method calculates error degrees of freedom as E = N - B - T, where E should be between 10 and 20.[4]
If E exceeds 20, the experiment may be wasting resources.[4] The QuickSize algorithm provides a versatile approach that can give an exact solution for a broad range of problems.[5] In stratified sampling, optimum allocation makes sampling rates proportional to standard deviations within strata and inversely proportional to the square root of sampling cost.[6] The equation E = N - B - T (where E is error degrees of freedom, N is total degrees of freedom, B is blocks degrees of freedom, and T is treatments degrees of freedom) should yield an E between 10 and 20; for example, with four treatment groups (T=3), eight animals per group (32 total, N=31), and no blocks (B=0), E equals 28, which is above 20, suggesting six animals per group might be more appropriate.[7][4]
What the research sources say
The NIST/SEMATECH e-Handbook states that the method for determining sample sizes for testing proportions is similar to the method for testing the mean.[8] Fleiss, Levin, and Paik recommend a continuity correction of N = N' + 1/δ for sample size calculations.[8] The resource equation method is appropriate for experiments analyzable by analysis of variance, such as complex biological experiments.[4] Mead's suggestion is that error degrees of freedom should be between 10 and 20 to give a good error estimate without wasting resources.[4]
In a non-blocked design, the total number of animals minus the number of treatments should be between ten and twenty.[4] Provided the error df is not less than about 6 in a blocked design, the experiment is probably of adequate size.[4]
References
- Wikipedia, "Sample size determination" (CC BY-SA 4.0), which cites: Sandelowski, M. (1995). Sample size in qualitative research. Research in Nursing & Health, 18, 179–183 (original not read by us) en.wikipedia.org
- Wikipedia, "Sample size determination" (CC BY-SA 4.0), which cites: "Inference for Regression", utdallas.edu; "Confidence Interval for a Proportion" (original not read by us) en.wikipedia.org
- Wikipedia, "Sample size determination" (CC BY-SA 4.0), which cites: Kenny, David A., "Statistics for the social and behavioral sciences", Little, Brown, 1987 (we also read that source) en.wikipedia.org
- Isogenic.info > Resource equation by Michael FW Festing. Updated Sept. 2006. facts only, no text reused. isogenic.info
- Wikipedia, "Sample size determination" (CC BY-SA 4.0), which cites: Amaratunga, D. (1999). Searching for the right sample size. The American Statistician, 53(1), 52–55. https://doi.org/10.2307/2685652 (original not read by us) en.wikipedia.org
- Wikipedia, "Sample size determination" (CC BY-SA 4.0), which cites: Kish (1965), p. 93. (original not read by us) en.wikipedia.org
- Wikipedia, "Sample size determination" (CC BY-SA 4.0), which cites: Isogenic.info > Resource equation by Michael FW Festing. Updated Sept. 2006 (we also read that source) en.wikipedia.org
- NIST/SEMATECH, "7.2.4.2. Sample sizes required", e-Handbook of Statistical Methods.. facts only, no text reused. itl.nist.gov
Each numbered sentence above was checked against the source it cites before publication. Sentences citing Wikipedia adapt text from the English Wikipedia article Sample size determination (contributors), used under CC BY-SA 4.0; this entry is available under the same licence. Other sources were used for facts only; no text or images are copied from them.
More in Probability & Statistics
Sources
Compiled from Wikipedia and the sources listed below. Text from Wikipedia is available under CC BY-SA 4.0; this entry is adapted from it.
- Wikipedia: Sample size determination (CC BY-SA 4.0).
- Word definitions: the Codexery glossary, each quoted from its Wikipedia article.
Spotted an error? Know more?
Reader corrections go straight into our review queue. Suggest an edit · How this site is sourced