De Moivre’s Normal Approximation and Sampling Error
Summary
This historical account explains how Abraham de Moivre used probability, calculus, and the binomial structure later associated with Pascal’s triangle to study variation in repeated samples. Its example is sampling stones from a jar: repeated groups produce proportions that cluster around the true proportion, with the normal, bell-shaped curve describing how likely different deviations are. The article connects this distribution to the idea of standard deviation as a measure of distance from the mean, and gives the familiar approximate coverage of one and two standard deviations.
It also illustrates using a probability model to assess an observed count of defective products against an assumed average defect rate. The article cautions that such a calculation depends on knowing or assuming the underlying rate; questions about uncertainty in that rate and how to use other sample sizes motivate its transition to Bayesian methods. This is a conceptual introduction to probability and statistical uncertainty, not a trading strategy or a worked market application.
Key ideas
- Repeated samples can produce proportions distributed around an underlying population proportion.
- The normal curve describes a symmetric concentration of observations near their mean, with more distant values less likely.
- Standard deviation measures typical distance from the mean and helps express how much of a normal distribution lies near its center.
- Probability calculations for observed defect counts depend on an assumed underlying defect rate.
- Uncertainty about the underlying rate raises questions that this account leaves for Bayesian analysis.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.