Chapter 10: The Binomial Distribution

The binomial distribution models experiments consisting of independent trials where each trial yields one of two outcomes, with probability p representing success and q equal to one minus p representing failure. The core formula quantifies the probability of obtaining exactly ν successes in n trials through the binomial coefficient multiplied by the respective probability terms, where the binomial coefficient accounts for all possible orderings of successes. The distribution exhibits critical statistical properties including a mean equal to np, standard deviation of the square root of np times q, and symmetry exclusively when p equals one half. As the number of trials increases substantially, the discrete binomial distribution converges to a continuous normal or Gaussian distribution, enabling physicists and statisticians to simplify calculations by approximating binomial probabilities using Gaussian functions centered at np with width determined by the standard deviation. This chapter demonstrates how the binomial framework proves that measurements subjected to numerous independent sources of random error approach normal distribution when error sources approach infinity while individual error magnitudes approach zero. The practical applications section addresses statistical hypothesis testing methodology, which involves formulating a null hypothesis specifying an event probability, calculating expected successes under that hypothesis, comparing actual results against expectations, and determining statistical significance using conventional thresholds of five percent and one percent probability levels. The discussion of one-tailed versus two-tailed probability calculations emphasizes how the direction of the alternative hypothesis influences which deviations from the mean are considered relevant to the analysis.