What Is a Good Sample Size for Quantitative Research? (A Practical Guide)

What Is a Good Sample Size for Quantitative Research? (A Practical Guide)

One of the most common questions in quantitative research design is: “How many participants do I need?” There is no single universal number, but there are clear principles that determine what counts as a good sample size. The right sample depends on your research design, statistical methods, and the strength of effect you expect to detect.

This article explains how sample size is determined in quantitative research and provides practical benchmarks you can use in psychology, education, and social science studies.

What Does “Good Sample Size” Mean?

A “good” sample size is one that is large enough to:

• Detect statistically meaningful effects (if they exist)

• Produce reliable and stable results

• Allow generalization to a wider population

• Reduce sampling error

In other words, it’s not about being “as large as possible,” but about being large enough for valid statistical inference.

Key Factors That Influence Sample Size

1. Research Design

Different designs require different sample sizes:

• Correlational studies → moderate to large samples

• Experimental designs → often smaller but more controlled

• Survey research → typically larger samples

• Regression / multivariate models → larger samples needed as variables increase

2. Effect Size

Effect size refers to how strong the relationship or difference is.

• Large effects → smaller samples needed

• Small effects → much larger samples needed

In psychology, effects are often small to medium, meaning larger samples are usually required.

3. Statistical Power

Statistical power is the probability of detecting a real effect.

A commonly accepted standard is:

• 80% power (0.80) = acceptable minimum

• 90% power (0.90) = stronger, more rigorous studies

Higher power requires larger sample sizes.

4. Significance Level (Alpha)

Most studies use:

• α = 0.05 (5% risk of false positives)

A stricter alpha (e.g., 0.01) increases the required sample size.

5. Number of Variables

The more variables you include:

• The larger your sample needs to be

• Especially in regression or structural equation modelling (SEM)

General Sample Size Guidelines (Rules of Thumb)

While exact calculations are best, here are widely used benchmarks:

Small studies (pilot or exploratory)

• 30–50 participants minimum

• Useful for feasibility or preliminary testing

Basic group comparisons (t-tests / ANOVA)

• 30–50 per group (minimum)

• 100–200 total is often more reliable

Correlational studies

• 100+ participants (minimum acceptable)

• 200–300+ preferred for stability

Regression analysis

A common guideline:

• 10–20 participants per predictor variable

• Minimum often around 100–150 total

Survey research

• 200–400+ is common

• Larger samples improve generalizability

Strong large-scale studies

• 500–1,000+ participants

• Used for population-level conclusions

Sample Size in Psychology Research

In fields like psychology, sample sizes are often influenced by:

• Ethical constraints

• Recruitment difficulty

• Student dissertation limits

• Effect size expectations (often small to medium)

For psychology dissertations specifically:

• Minimum acceptable: ~100 participants (many institutions expect this for quantitative work)

• Strong standard: 150–300 participants

• Excellent large sample: 300+ participants

Why Bigger Isn’t Always Better

While large samples improve reliability, they are not automatically better because:

• Very large samples can detect trivial effects as “statistically significant”

• They may increase cost and time demands

• Poor measurement quality cannot be fixed by more participants

Quality of design still matters more than size alone.

How to Calculate Sample Size Properly

Instead of guessing, researchers often use:

• Power analysis software (e.g., G*Power)

• Prior research effect sizes

• Pilot studies

• Institutional guidelines

A basic power analysis typically requires:

• Expected effect size

• Alpha level (usually 0.05)

• Desired power (usually 0.80)

• Statistical test type

Common Mistakes in Choosing Sample Size

• Choosing a number based on convenience only

• Ignoring effect size

• Not accounting for dropouts (attrition)

• Using too many variables for a small sample

• Assuming “100 is always enough”

A good practice is to overestimate slightly to account for missing or unusable data.

Final Thoughts on What Is a Good Sample Size for Quantitative Research? (A Practical Guide)

There is no single “perfect” sample size for quantitative research. A good sample size depends on your design, expected effect size, and statistical requirements. However, as a general guide, most psychology and education studies benefit from at least 100–300 participants, with larger samples improving reliability and generalizability.

Ultimately, the best approach is to use power analysis combined with methodological judgement, rather than relying on fixed numbers.