The central limit theorem explains why the normal distribution appears across almost every branch of quantitative analysis.
It states that the standardised sum or average of a large number of independent random variables approaches a normal distribution.
Remarkably, this convergence occurs regardless of the shape of the underlying population distribution from which data is drawn.
Whether the original population is uniform, heavily skewed, or discrete, the distribution of sample averages forms a Gaussian curve.
The expected value of the distribution of sample averages equals the true population mean mu.
The standard deviation of the sample averages, called the standard error, equals the population standard deviation divided by the square root of n.
As the number of observations in each sample grows, the standard error shrinks, concentrating sample means tightly around the truth.
The theorem requires that observations are independent and possess both a finite mean and a finite variance.
For moderately symmetric populations, sample counts as small as thirty are often sufficient for the normal approximation to hold.
For severely skewed or heavy-tailed distributions, substantially larger sample counts are required before normality emerges.
The central limit theorem enables data scientists to construct confidence intervals and conduct hypothesis tests without knowing the population distribution.
It also explains why natural measurements like height or blood pressure follow bell curves: they represent the sum of many small genetic and environmental factors.
When random variables are multiplied together rather than added, the central limit theorem produces a log-normal distribution instead.
A critical misunderstanding is believing that the original dataset becomes normal as the observation count grows: only the sample average becomes normal.
Data scientists must remember that the theorem applies strictly to sample sums and means, not to individual raw observations.