6.7
Distribution de probabilité binomiale
Une distribution binomiale est une distribution de probabilité associée à une expérience comportant un nombre fixe d’essais, où chaque essai ne peut a…
La distribution de probabilité binomiale représente les cas qui ont un nombre multiple mais fixe d’essais, comme dans un tirage au sort, avec deux résultats possibles par essai.
Ici, n désigne le nombre d’essais.
Dans chaque essai, la probabilité de succès, face, est notée p, tandis que la probabilité d’échec, queue, est représentée par q. Si l’un est connu, l’autre peut être facilement calculé.
Pour une distribution binomiale, la probabilité de succès ou d’échec doit toujours être la même pour tous les essais.
De plus, l’issue de chaque procès doit être indépendante des autres essais.
Dans cet exemple, le nombre de têtes est la variable aléatoire, x, dont la valeur peut être un nombre entier compris entre 0 et n.
P de x désigne la probabilité de x têtes parmi n essais, calculée à l’aide de la formule de probabilité binomiale.
Ici, le symbole factoriel représente le produit de facteurs décroissants.
Pour chaque valeur de x, on peut obtenir P de x, qui peut être tracé pour obtenir la forme graphique de la distribution binomiale.
View the full transcript and gain access to JoVE Core videos
Q1: What are the key conditions for a binomial experiment?
A binomial experiment requires a fixed number of trials, n, with only two possible outcomes per trial: success or failure. The probability of success, p, and probability of failure, q, must remain constant across all trials. Additionally, each trial must be independent, meaning the outcome of one trial does not affect another. These conditions ensure the experiment fits a binomial probability distribution.
Q2: How do you calculate the probability of a specific number of successes in a binomial distribution?
The binomial probability formula calculates P(x), the probability of exactly x successes in n trials. The formula uses the binomial coefficient, represented by the factorial symbol, multiplied by p raised to the power of x and q raised to the power of (n-x). For each value of x, you can compute P(x) to determine the complete probability distribution.
Q3: What is the relationship between p and q in a binomial distribution?
In a binomial distribution, p represents the probability of success and q represents the probability of failure on a single trial. These probabilities are complementary, meaning p + q = 1. If you know one probability, you can easily calculate the other by subtracting from 1, ensuring the two outcomes account for all possibilities.
Q4: Why must trials be independent in a binomial experiment?
Independence ensures that the outcome of one trial does not influence or help predict the outcome of another trial. This requirement allows the probability of success, p, and probability of failure, q, to remain constant across all trials. Without independence, the probabilities would change based on previous results, violating the fundamental assumptions of a binomial distribution.
Q5: How is a binomial distribution graphically represented?
A binomial distribution is graphically represented using probability histograms, where each value of x (number of successes) is plotted against its corresponding probability P(x). The x-axis shows possible outcomes from 0 to n, and the y-axis shows probabilities. This visual representation helps identify the distribution's shape and which outcomes are most likely.
Q6: What does the random variable x represent in a binomial distribution?
The random variable x represents the number of successes observed in n trials of a binomial experiment. Its value must be a whole number ranging from 0 to n. For example, in coin tosses, x could represent the number of heads obtained, with each possible value having an associated probability calculated using the binomial formula.
Q7: Can you give a real-world example of a binomial experiment?
A true-false statistics test is a practical binomial experiment. If Joe guesses on each question with a constant probability of success p = 0.6 and failure q = 0.4, each question represents an independent trial with two outcomes. The number of correct answers Joe gets follows a binomial distribution, where x can range from 0 to the total number of questions.