Chapter 5: The Normal Distribution

The Normal Distribution limiting distribution allows researchers to calculate the probability that any single measurement falls within a specified range. The normal distribution, or Gaussian curve, emerges as the characteristic distribution for measurements affected by many small random errors without significant systematic bias. The distribution is defined by two parameters: the true value at its center and the standard deviation that determines its width, with smaller standard deviations indicating more precise measurements. A fundamental property established through integration of the Gaussian function shows that approximately 68% of measurements fall within one standard deviation of the true value, 95% within two standard deviations, and 99.7% within three standard deviations, providing quantifiable confidence limits for measurement accuracy. The chapter then justifies why the arithmetic mean of measured values serves as the best estimate of the true value using maximum likelihood principles, and similarly validates using the sample standard deviation as the best estimate of measurement precision. A critical contribution is the mathematical derivation of error propagation rules, demonstrating that when combining independent measurements with random uncertainties, the combined uncertainty equals the square root of the sum of squared individual uncertainties rather than simple addition. The chapter also derives the standard deviation of the mean formula, showing that measurement uncertainty decreases proportionally to the square root of the number of observations, a relationship that quantifies the benefit of repeated measurements. Finally, the chapter provides practical guidance for evaluating whether measured results agree acceptably with theoretical predictions by calculating how many standard deviations separate the values and determining whether such discrepancies could reasonably occur by chance alone, establishing conventional significance thresholds typically at 5% or 1% probability levels.