Continuous Random Variables Lecture 22 Section Mon, Feb 25, 2008
Random Variables Random variable Discrete random variable Continuous random variable
Continuous Probability Distribution Functions Continuous Probability Distribution Function (pdf) – For a random variable X, it is a function with the property that the area below the graph of the function between any two points a and b equals the probability that a ≤ X ≤ b. Remember, AREA = PROPORTION = PROBABILITY
Example The TI-83 will return a random number between 0 and 1 if we enter rand and press ENTER. These numbers have a uniform distribution from 0 to 1. Let X be the random number returned by the TI-83.
Example The graph of the pdf of X. x f(x)f(x) 01 1
Example What is the probability that the random number is at least 0.3?
Example What is the probability that the random number is at least 0.3? x f(x)f(x)
Example What is the probability that the random number is at least 0.3? x f(x)f(x)
Example What is the probability that the random number is at least 0.3? x f(x)f(x)
Area = 0.7 Example Probability = 70%. x f(x)f(x)
0.25 Example What is the probability that the random number is between 0.25 and 0.75? x f(x)f(x)
Example What is the probability that the random number is between 0.25 and 0.75? x f(x)f(x)
Example What is the probability that the random number is between 0.25 and 0.75? x f(x)f(x)
Example Probability = 50%. x f(x)f(x) Area = 0.5
Uniform Distributions The uniform distribution from a to b is denoted U(a, b). ab 1/(b – a) x
A Non-Uniform Distribution Consider this distribution. 510 x
A Non-Uniform Distribution What is the height? 510 ? x
A Non-Uniform Distribution The height is x
A Non-Uniform Distribution What is the probability that 6 X 8? x 68
A Non-Uniform Distribution It is the same as the area between 6 and x 68
Uniform Distributions The uniform distribution from a to b is denoted U(a, b). ab 1/(b – a)
Hypothesis Testing (n = 1) An experiment is designed to determine whether a random variable X has the distribution U(0, 1) or U(0.5, 1.5). H 0 : X is U(0, 1). H 1 : X is U(0.5, 1.5). One value of X is sampled (n = 1).
Hypothesis Testing (n = 1) An experiment is designed to determine whether a random variable X has the distribution U(0, 1) or U(0.5, 1.5). H 0 : X is U(0, 1). H 1 : X is U(0.5, 1.5). One value of X is sampled (n = 1). If X is more than 0.75, then H 0 will be rejected.
Hypothesis Testing (n = 1) Distribution of X under H 0 : Distribution of X under H 1 :
Hypothesis Testing (n = 1) What are and ?
Hypothesis Testing (n = 1) What are and ?
Hypothesis Testing (n = 1) What are and ? Acceptance RegionRejection Region
Hypothesis Testing (n = 1) What are and ?
Hypothesis Testing (n = 1) What are and ? = ¼ = 0.25
Hypothesis Testing (n = 1) What are and ? = ¼ = 0.25 = ¼ =
Example Now suppose we use the TI-83 to get two random numbers from 0 to 1, and then add them together. Let X 2 = the average of the two random numbers. What is the pdf of X 2 ?
Example The graph of the pdf of X 2. y f(y)f(y) ?
Example The graph of the pdf of X 2. y f(y)f(y) Area = 1
Example What is the probability that X 2 is between 0.25 and 0.75? y f(y)f(y)
Example What is the probability that X 2 is between 0.25 and 0.75? y f(y)f(y)
Example The probability equals the area under the graph from 0.25 to y f(y)f(y)
Example Cut it into two simple shapes, with areas 0.25 and 0.5. y f(y)f(y) Area = 0.5 Area =
Example The total area is y f(y)f(y) Area = 0.75
Hypothesis Testing (n = 2) An experiment is designed to determine whether a random variable X has the distribution U(0, 1) or U(0.5, 1.5). H 0 : X is U(0, 1). H 1 : X is U(0.5, 1.5). Two values of X are sampled (n = 2). Let X 2 be the average. If X 2 is more than 0.75, then H 0 will be rejected.
Hypothesis Testing (n = 2) Distribution of X 2 under H 0 : Distribution of X 2 under H 1 :
Hypothesis Testing (n = 2) What are and ?
Hypothesis Testing (n = 2) What are and ?
Hypothesis Testing (n = 2) What are and ?
Hypothesis Testing (n = 2) What are and ? = 1/8 = 0.125
Hypothesis Testing (n = 2) What are and ? = 1/8 = = 1/8 = 0.125
Conclusion By increasing the sample size, we can lower both and simultaneously.