Why Sampling Matters
Sampling is the process of selecting a subset of individuals from a population. Since studying an entire population is often impractical, we select a representative sample and generalize findings.
Probability Sampling Methods
Every member of the population has a known chance of being selected:
1. Simple Random Sampling
Every individual has an equal probability of selection. Like drawing names from a hat.
When to use: Population is homogeneous and accessible.
2. Systematic Sampling
Select every k-th individual from a list. For example, every 10th student from an enrollment list.
3. Stratified Random Sampling
Divide the population into subgroups (strata) based on a characteristic, then randomly sample from each stratum.
Example: Stratify by department, then randomly select from each department proportionally.
4. Cluster Sampling
Divide the population into clusters (e.g., schools, districts), randomly select clusters, then survey all members or a random sample within selected clusters.
Non-Probability Sampling Methods
1. Convenience Sampling
Select whoever is most easily accessible. Quick but may introduce bias.
2. Purposive (Judgmental) Sampling
Researcher selects participants based on their expertise or specific characteristics.
3. Snowball Sampling
Existing participants recruit others. Useful for hard-to-reach populations.
Determining Sample Size
Several methods exist for calculating sample size:
Method 1: Using Tables
Krejcie and Morgan (1970) table provides recommended sample sizes:
| Population | Sample Size |
|---|---|
| 100 | 80 |
| 200 | 132 |
| 500 | 217 |
| 1,000 | 278 |
| 5,000 | 357 |
| 10,000 | 370 |
| 100,000 | 384 |
Method 2: Formula
For known population size (N):
n = N / (1 + N × e²)
Where: n = sample size, N = population size, e = margin of error (usually 0.05)
Method 3: Statistical Power
For advanced studies, use G*Power software to calculate sample size based on effect size, significance level, and desired power.
Rule of Thumb: For most survey research, a minimum of 30 respondents per group is recommended for parametric tests. For factor analysis, aim for at least 5 respondents per questionnaire item.