dr.susan.carroll@snet.net

Sampling question #2: How do I determine sample size?

Many students wonder about their sample size. How large should it be? Some of the decision is based on practical aspects. What population do you have access to? How much time and money are you willing or able to spend on sample acquisition.

*Ask experts what they would suggest for sample size.

*Conduct a power analysis. There are programs on the web that you can access for free. But you must input data in order for them to compute the sample size for you.[ Go to the HYPOTHESES section of this website and review the information on POWER ANALYSIS. This will help if you need to do a power analysis for your sample size.]

*See what the literature recommends.

*Must have at least 10 per cell for statistical analysis.

*Rule of thumb- larger is better.

If you would like to get a feeling for the size of samples that represent entire populations, please refer to Krejcie (1970). The chart below suggests the great benefit that randomly-selected samples afford. (Population size is noted by uppercase āNā and sample size by lower case ānā.)

Random Sample Sizes (n) Required for Population (N) Representation

Population Size (N) ------ Sample Size(n)

50 ------------------ 44

100 ----------------- 80

500 ----------------- 217

1,000 ---------------- 278

1,500 ---------------- 306

3,000 ---------------- 341

5,000 ---------------- 357

10,000 --------------- 375

50,000 --------------- 381

100,000 -------------- 384

Source: Krejcie, R.V. & Morgan, D.W. (1970) Determining sample size for research activities. Educational and Psychological Measurements, 30, 607-610.

Return from sampling question #2: determining sample size to samples and sample size.