Excel Functions for Discrete Distributions - Copy
Excel Functions for Discrete Distributions - Copy
BINOMIAL DISTRIBUTION
Example: The area of to the left of and including 2 below the binomial distribution with 8 trials and
probability of success 0.5 is: BINOM.DIST(2,8,0.5,TRUE) = 0.144
BINOM.INV(trials,probability_s,alpha) – Returns the smallest value for which the cumulative binomial
distribution is greater than or equal to alpha. Outputs the x-value (number of successes) that has at
least alpha proportion to the left of it based on the binomial distribution with trials number of trials and
probability of success probability_s.
For example, BINOM.INV(8,0.5,0.05) = 2. I’ll attempt to explain this a little bit here. The cumulative
binomial distribution up to 1 success (out of 8) is equal to: BINOM.DIST(1,8,0.5,TRUE) = 0.035. The
cumulative binomial distribution up to 2 successes (out of 8) is equal to (see above):
BINOM.DIST(2,8,0.5,TRUE) = 0.144. The BINOM.INV function outputs the minimum number of trials
whose cumulative distribution is at least alpha. Since alpha = 0.05 lies between the cumulative
probabilities for 1 and 2 successes, we must round up to 2, which is the output of the function.
Example: If we need 3 successes (r = 3), the area up to and including x = 6 (3 rd success in at most 6 total
trials) of the negative binomial distribution with probability of success = 0.5 is:
NEGBINOM.DIST(3,3,0.5,TRUE) = 0.656
There is no .INV version of the negative binomial function in Excel (i.e., there is no NEGBINOM.INV
function).
GEOMETRIC DISTRIBUTION
There is no “GEOM.DIST” function in Excel. However, since the geometric distribution is a special case
of the negative binomial distribution with only 1 success in x number of trials, we can use the
NEGBINOM.DIST function with the number of successes equal to 1:
NEGBINOM.DIST(number_f,1,probability_s,cumulative)
Example: The probability of obtaining the first success in at most 5 trials (area up to and including x = 8
in the negative binomial distribution with r = 1) with probability of success 0.3 is given by:
NEGBINOM.DIST(4,1,0.3,TRUE) = 0.832
HYPERGEOMETRIC DISTRIBUTION
Example: The probability of obtaining 0 or 1 successes in a sample of size 4 drawn from a population of
size 20 that contains 5 successes is:
HYPGEOM.DIST(1,4,5,20,TRUE) = 0.751
Note that it wouldn’t make sense for sample_s to be greater than population_s.
POISSON DISTRIBUTION
Example: The probability that x will be less than or equal to 3 given a mean of 4 is:
POISSON.DIST(3,4,TRUE) = 0.433