1. Which statistical test should I use?
Match your test to your design — number of groups, type of outcome, paired or independent, assumptions met or not.
Free tool: Statistical Test Selector
Statistics for research: a plain-English guide
You don’t need to love statistics to use them well — you need to make a handful of decisions correctly: pick the right test, recruit enough participants, and report your results so a reader can judge them. This guide walks through the essentials, each with a free tool to do the maths for you.
The five decisions that matter most
Most analyses in a thesis or first paper come down to five questions. Get these right and the software is just arithmetic:
2. How many participants do I need?
Sample size and statistical power decide whether your study can actually detect the effect you care about.
Free tools: Sample Size & Power calculators
3. What does a p-value actually mean?
The most misunderstood number in research — what it is, what it isn’t, and why “p < 0.05” isn’t the whole story.
Free tool: p-value ↔ CI converter
4. How big is the effect? (effect size)
Significance tells you whether; effect size tells you how much — Cohen’s d, r, and friends.
Free tool: Effect Size Converter
How the pieces fit together
They’re a chain, not a checklist. Before you collect data, a power analysis (built from your expected effect size) tells you the sample size to recruit. After you collect it, you run the right test, then report the result as a p-value with a confidence interval and an effect size — so a reader sees both whether the effect is real and how large it is. Skipping the effect size, or treating p < 0.05 as the finish line, is the most common way good data gets reported badly.
Use the tools as you read
Each guide above pairs with a free, in-browser tool — no signup, nothing stored:
- Statistical Test Selector — answer a few questions, get the right test.
- Sample Size Calculator and Power Calculator.
- Effect Size Converter (d ↔ r ↔ OR ↔ η²).
- p-value ↔ Confidence Interval converter and the Cronbach’s α calculator.
Get the free Statistics toolkit
Test-selection flowcharts, power and effect-size worked examples, and a reporting checklist from the Statistical Test Selection Workbook. We’ll email you the download link.
Frequently asked questions
Do I need to be good at maths to do research statistics?
No. You need to make the right design decisions — test choice, sample size, honest reporting. The arithmetic is done by software and by tools like the ones linked here.
What should I always report?
The test, the effect size, a confidence interval, and the p-value — not the p-value alone. That lets a reader judge both whether an effect is real and how large it is.
Is a significant result a big result?
No. Statistical significance (a small p-value) can come from a large sample even when the effect is tiny. Always read it alongside the effect size.