This activity is intended to clarify the relationship between discrete sample spaces, discrete random variables, probability mass functions, and cumulative distribution functions, with a focus on computing probabilities using these concepts. We will use cumulative distribution functions often in this course to organize different approaches to computing probability. The real importance of the cumulative distribution and probability mass function is their ability to describe the behavior of a population from a plot of sample data.
For each value xi in the range of a discrete random variable X we can compute the probability f(xi) of the sample space event that maps into xi. The pairs (xi, f(xi)) define a probability mass function which follows the rules
The cumulative distribution function F(x) associated with f(x) is
defined for all real numbers as F(x) =
f(x1) + f(x2 + ... + f(xi) where
xi <= x < xi+1.
A disk jockey selects 4 records at random from a collection consisting of 4 jazz records, 2 classical records, and 3 polka records. The sample space can be represented as {R1R2R3R4} where Ri is either a jazz, classical or polka record. Let X be the number of classical records chosen. The following table applies:
Possible Values for X: 0 1 2
# ways to choose x 7 6 5 4 2 7 6 5 2 1 7 6
---------- ----------- ----------
classical records 4! 3! 2! 2!
# ways to choose any 9 8 7 6 9 8 7 6 9 8 7 6
---------- ---------- ----------
4 classical records 4! 4! 4!
P(X = x) (I.e. f(x)) 5/18 5/9 1/6
The cumulative distribution function associated with f(x) is:
{ 0 x < 0
| 5/18 0 <= x < 1
F(x) = | 15/18 1 ó x < 2
{ 1 x >= 2
These will be provided in class.
Return to the
List of Activities