Nothing really. It concerns an experiment with identified success and failure probabilities (p and q), or Bernoulli trials, like the conventional binomial distribution. In an negative binomial experiment, the experiment is stopped after "r" successes occur in n trials. Thus, there must be r-1 successes in the first n-1 trials, and the final trial must be a success. This stopping event causes a n-1 and r-1 terms to appear in the factorial expressions of the distribution, which I suspect is the origins of calling this distribution a "negative binomial distribution." I would prefer to call this a Bernoulli experiment distibution with a stopping rule, but that's probably much too long. Some excellent websites provide examples and more discussion: http://mathworld.wolfram.com/NegativeBinomialDistribution.html http://stattrek.com/Lesson2/NegBinomial.aspx http://en.wikipedia.org/wiki/Negative_binomial_distribution Stattrek has very good examples. Note the distribution can be expressed in a number of forms.
Chat with our AI personalities
First i will explain the binomial expansion
It is necessary to use a continuity correction when using a normal distribution to approximate a binomial distribution because the normal distribution contains real observations, while the binomial distribution contains integer observations.
Binomial distribution is the basis for the binomial test of statistical significance. It is frequently used to model the number of successes in a sequence of yes or no experiments.
Poisson and Binomial both the distribution are used for defining discrete events.You can tell that Poisson distribution is a subset of Binomial distribution. Binomial is the most preliminary distribution to encounter probability and statistical problems. On the other hand when any event occurs with a fixed time interval and having a fixed average rate then it is Poisson distribution.
The mean of a binomial probability distribution can be determined by multiplying the sample size times the probability of success.