How to Calculate Sample Size for Your Thesis (Yamane and Cochran Formulas)

Few things draw examiner questions faster than a weak sample size justification. Whether you are writing a BSc project, an MSc dissertation, or a PhD thesis at UNILAG, UI, OAU, Covenant, or any other Nigerian university, you must show that your sample is large enough to support your conclusions and that you calculated it with a recognised method. If you are still planning your project, start with our guides on how to write a dissertation proposal and writing your master’s dissertation.

The two formulas Nigerian researchers use most are Yamane (1967) and Cochran (1977). This guide explains how each works, when to choose which, and how to write the result in Chapter Three.

Why Sample Size Matters

  • Too small: your results may not represent the population, and statistical tests lose power.
  • Too large: you waste time and money, and data collection becomes hard to manage.
  • Unjustified: even a good sample draws criticism at your proposal defence if you cannot show how you arrived at it.

Sample size is part of your methodology, so read our top tips for writing a dissertation methodology alongside this guide.

Yamane’s Formula

Best for: a known, finite population, such as staff of an organisation, students in a faculty, or members of an association.

Formula:

n = N / (1 + N·e²)

  • n = sample size
  • N = population size
  • e = margin of error (0.05 for 95% confidence)

Worked Example

A study on employee motivation in a bank with 1,200 staff, at a 5% margin of error:

n = 1200 / (1 + 1200 × 0.05²)
n = 1200 / (1 + 3)
n = 1200 / 4 = 300 respondents

Quick Reference (e = 0.05)

Population (N)Sample size (n)
10080
500222
1,000286
5,000370
10,000385
100,000398

Limitations

  • It assumes a 95% confidence level and maximum variability (p = 0.5). You cannot adjust either.
  • It requires a known population size.
  • It is mathematically the same as the Slovin formula. Critics describe that formula as ad hoc and note that it has no term for how variable the population is. This is why some supervisors prefer Cochran. wolfram

Cochran’s Formula

Best for: large or unknown populations, or when you want control over the confidence level and expected proportion. Background on the method is in this overview of sample size determination.

Step 1: Formula for Large Populations

n₀ = (Z² · p · q) / e²

  • Z = Z-score (1.645 for 90%, 1.96 for 95%, 2.576 for 99%)
  • p = estimated proportion of the attribute in the population (use 0.5 if unknown, as it gives the safest, largest sample)
  • q = 1 − p
  • e = margin of error

Worked Example

95% confidence, p = 0.5, e = 0.05:

n₀ = (1.96² × 0.5 × 0.5) / 0.05²
n₀ = (3.8416 × 0.25) / 0.0025
n₀ = 384.16, rounded up to 385

Step 2: Finite Population Correction

If your population is known and small, adjust n₀:

n = n₀ / (1 + (n₀ − 1) / N)

Example with N = 2,000:

n = 384.16 / (1 + 383.16 / 2000)
n = 384.16 / 1.1916 = 322.4, rounded up to 323

To cross-check your figures, use the Raosoft sample size calculator.

How the Inputs Change the Result (p = 0.5)

Confidence levelMargin of errorSample size (n₀)
90%5%271
95%10%97
95%5%385
99%5%664

Yamane vs Cochran: Which Should You Use?

For the same bank population of 1,200 staff, Yamane gives 300 and Cochran with finite correction gives 292. The two are close for finite populations, so the choice is usually about fit:

SituationBetter choice
Known, finite populationYamane (simple) or Cochran with correction
Large or unknown populationCochran
You need 90% or 99% confidenceCochran
You know the likely proportion (e.g., 30% use a service)Cochran
Your supervisor or department prefers oneUse that one

How to Write It in Chapter Three

Use a short, formal paragraph:

The population of the study comprises 1,200 staff of the selected bank. The sample size was determined using Yamane’s (1967) formula at a 5% margin of error: n = N / (1 + Ne²) = 1200 / (1 + 1200(0.05)²) = 300. Therefore, 300 respondents were sampled.

Add the formula, the substitution, and the result, and cite the source in your references. For the rest of the chapter, see our guide to writing a dissertation methodology.

Common Mistakes to Avoid

  1. Using the wrong population. N must be the population of your study, not the whole institution or state.
  2. Forgetting to round up. Always round to the next whole number.
  3. Ignoring non-response. If you expect 10 to 20% of questionnaires to go unreturned, distribute extra copies. Report the calculated sample, then state how many you administered. A well-built instrument also helps, so see our support for questionnaire design and validation.
  4. Applying these formulas to qualitative research. Interviews and case studies use saturation, not Yamane or Cochran.
  5. Skipping the sampling technique. The formula gives a number. You still need to state how you selected respondents (simple random, stratified, purposive).
  6. Stopping at the sample size. Examiners also ask how you will analyse the data, so link your sample to your tests (regression, ANOVA, chi-square). If SPSS or STATA is a hurdle, Writers.ng offers thesis data analysis support.

For more pitfalls that cost marks, read mistakes to avoid with dissertation writing.

Frequently Asked Questions

Is Yamane’s formula acceptable for a thesis?
Yes, it is widely accepted in Nigerian universities for finite populations, though some supervisors prefer Cochran for its flexibility.

What if my population is unknown?
Use Cochran’s formula with p = 0.5.

Can I use a smaller sample than the formula gives?
You can only do so with justification, such as limited access or a stated lower confidence level. Be ready to defend it.

Should I add extra respondents for non-response?
Yes. Inflate your distribution, for example by 10 to 15%, so your usable responses still meet the calculated sample.

Can Writers.ng help with SPSS analysis after I collect my data?
Yes. Writers.ng supports Nigerian postgraduate students with SPSS, STATA and NVivo analysis and the results chapter write-up. See our thesis writing services in Nigeria guide for details.

Related Guides from Writers.ng

Need Help With Your Thesis Methodology and Data Analysis?

Sample size is only one piece of Chapter Three. At Writers.ng, our team supports Nigerian postgraduate students with research methodology, questionnaire design, data analysis, and full dissertation writing services in Nigeria. If you are stuck on your sample size calculation or the analysis that follows, contact Writers.ng for guidance tailored to your topic and institution.