Biostatistics Basics: Data Types, Sampling & Significance

Biostatistics is the application of statistics to medicine and biology: designing studies, summarising data, and deciding whether an observed effect is real or could be chance. Every drug trial, screening test evaluation, and disease outbreak report leans on it.

It matters because modern doctors must read research, not just memorise it. Knowing what a p-value actually means, which test fits which data, and how sampling can mislead is what separates evidence-based practice from anecdote - and it is a heavily tested subject in postgraduate entrance exams.

Key Concepts

• Data types: nominal (named categories like blood group), ordinal (ordered ranks like pain scores), interval and ratio (measured numbers like weight).
• Sampling: simple random, stratified (by subgroups), systematic (every kth person), and cluster sampling; a bad sample biases everything after it.
• Central tendency: mean (average), median (middle value), mode (commonest value); dispersion is described by range and standard deviation.
• p-value: the probability of seeing data this extreme if the null hypothesis were true; p below 0.05 is the conventional significance cut-off.
• Type I error (alpha): falsely claiming an effect exists; Type II error (beta): missing a real effect; power = 1 - beta.
• Common tests: t-test compares two means, chi-square compares categories, ANOVA compares three or more means, correlation measures association.
• Sensitivity and specificity: a test's ability to find the diseased and to clear the healthy.

Formulas

• Mean: x-bar = sum of values / n (sum divided by count).
• Sensitivity = true positives / (true positives + false negatives) x 100.
• Specificity = true negatives / (true negatives + false positives) x 100.
• Chi-square: sum of (observed - expected)^2 / expected, used for categorical data with adequate cell counts.

Worked Example

Five blood-pressure readings: 120, 130, 125, 135, 120 mmHg. Sum = 120 + 130 + 125 + 135 + 120 = 630. Mean = 630 / 5 = 126 mmHg. Ordered, the values are 120, 120, 125, 130, 135, so the median is 125 and the mode is 120. The mean sits slightly above the median because the high value 135 pulls it up - a first glimpse of skew.

Common Misconceptions

• p below 0.05 proves the treatment works. It only means the result would be unusual under the null hypothesis; it says nothing about effect size or clinical importance.
• The mean is always the best average. In skewed data like hospital stay lengths, a few long stays drag the mean up and the median describes the typical patient better.
• Correlation proves causation. Linked variables may share a hidden cause - ice-cream sales and drowning rise together only because both follow summer heat.

Real-World Applications

• Randomised trials use these tests to license new drugs.
• Sensitivity and specificity decide whether a screening test is fit for use.
• Outbreak investigations compare attack rates with chi-square tests.
• Meta-analyses pool trial results to write treatment guidelines.

Biostatistics turns medical data into decisions: right data type, right sample, right test, honest p-value. Remember that significance is not importance and correlation is not causation, and the numbers will inform rather than mislead.