Biostatistics for Medical Students: A Beginner's Guide
Core biostatistics for MBBS and postgraduate students, from types of variables to confidence intervals, in plain language with clinical examples.
Biostatistics is the use of statistics to answer questions about health and disease. You do not need to be good at mathematics to understand it. You need a handful of core ideas, used consistently. This guide covers the ones you will meet in almost every thesis and journal article.
1. Know your variable types
Everything in statistics starts with the type of variable, because it decides how you summarise and test it.
| Type | Meaning | Examples |
|---|---|---|
| Nominal (categorical) | Categories with no order | Sex, blood group, diabetic yes/no |
| Ordinal | Ordered categories | Pain score (mild/moderate/severe), NYHA class |
| Discrete numerical | Counts | Number of pregnancies, number of admissions |
| Continuous numerical | Measurements on a scale | Haemoglobin, blood pressure, age |
Numerical variables can be grouped into categories (age groups), but you lose information when you do. Keep the original values in your dataset.
2. Population and sample
The population is everyone you want to draw conclusions about. The sample is the group you actually study. Statistics lets you use a sample to estimate what is true in the population, and tells you how uncertain that estimate is.
3. Describing data
Descriptive statistics summarise your sample:
- Categorical variables: counts and percentages.
- Numerical variables: a measure of the centre (mean or median) and a measure of spread (standard deviation, or interquartile range).
Use the mean and standard deviation when data are roughly symmetrical, and the median and interquartile range when data are skewed (for example, hospital stay, which often has a few very long stays).
4. The normal distribution
Many biological measurements form a bell-shaped curve. In a normal distribution, about 68% of values lie within one standard deviation of the mean and about 95% within two (more precisely, 1.96). This property is the basis of many tests and of confidence intervals. Check whether your data are approximately normal with a histogram, a Q–Q plot or a normality test before choosing your analysis.
5. Estimating with confidence
A sample mean or proportion is an estimate. The 95% confidence interval gives a range of plausible values for the population. Example: "The prevalence of anaemia was 48% (95% CI 42% to 54%)." Narrower intervals come from larger samples.
6. Testing hypotheses
When you compare groups, you set up a null hypothesis (no difference) and use a test to calculate a p-value: the probability of results at least as extreme as yours if the null hypothesis were true. A small p-value suggests the data are not very compatible with "no difference". It does not tell you how big or important the difference is. Our article on p-values and confidence intervals explains this in more detail.
7. Errors and power
- Type I error (α): concluding there is a difference when there is none. Usually set at 5%.
- Type II error (β): missing a real difference.
- Power (1 − β): the chance of detecting a real difference of a given size. Studies usually aim for 80% or 90% power, which is why sample size matters.
Common mistakes
- Using the mean for heavily skewed data.
- Reporting only p-values without the size of the effect or a confidence interval.
- Treating ordinal scores as if the gaps between categories were equal, without justification.
- Converting continuous data into categories too early and losing information.
Key takeaways
- The variable type decides the summary and the test.
- Use mean (SD) for symmetrical data and median (IQR) for skewed data.
- Confidence intervals show how precise an estimate is; p-values alone are not enough.
- Power and sample size are planned before the study, not after.
References
- Kirkwood BR, Sterne JAC. Essential Medical Statistics. 2nd ed. Oxford: Blackwell Science; 2003.
- Altman DG. Practical Statistics for Medical Research. London: Chapman & Hall; 1991.
- Mishra P, Pandey CM, Singh U, Gupta A, Sahu C, Keshri A. Descriptive statistics and normality tests for statistical data. Ann Card Anaesth. 2019;22(1):67–72. PubMed
This article is for education. Always follow the requirements of your university, guide, ethics committee and target journal.