Chapter 6 The Normal Probability Distribution

Homework #6 (Week 9): Chapter 6, Exercises 6, 8, 14, 16, 18, 20, 24 & 28.

Probability Distributions

Definition: "A probability distribution lists, in some form, all the possible outcomes of an "experiment" and the probability associated with each one."

Continuous Random Variables

A continuous random variable is assumed to be able to take any value in an interval, e.g. height measured with complete accuracy, time measured with complete accuracy, temperature measured with complete accuracy, ...

Continuous Probability Distributions

The total area under the probability distribution curve bounded by the x-axis is 1 (=100%). The area under the curve between the lines x = a and x = b gives the probability that x lies between a and b.

Note: The height of the curve over a point on the x-axis has no direct probability meaning.

i.e. P(X = a) = P(X = b) = 0

̃ P(a<X<b) = P(a£X£b) = P(a<X£b) = P(a£x<b)

The Normal (Gaussian) Distribution

Intuition: The normal distribution is given by a symmetrical, bell-shaped curve. The position of the curve is determined by the expected value (m) and variance (s2) of the distribution.

X ~ N(m, s2)

Note: If we fix m and let s2 vary we get a family of curves with the same expected values but different variances (i.e. different levels/degrees of "spreadoutness").

Note: If we fix s2 and let m vary we get a family of curves with the same shape but different locations along the horizontal axis.

Examples: Men's heights, Women's heights, Men's weights, Women's weights, Number of goals scored by Premiership teams in a season, Introduction to Sociology grades, Mathematics & Statistics grades (??), …

The Standard Normal Distribution

Let Z be a standard normal variable, that is, a normal variable for which m = 0 and s2 = 1. Statistical tables allow us to calculate the area under any portion of the standard normal curve.

Z ~ N(0, 1) Examples:

(a) P(Z > 1.60) = 0.055 (interpretation?)

(b) P(1.60 < Z < 2.30) = P(Z > 1.60) - P(Z > 2.30) = 0.055 - 0.011 = 0.044 (interpretation?)

(c) P(Z < 1.64) = 1 - P(Z > 1.64) = 1 - 0.051 = 0.949 (» 0.95) (interpretation?)

(d) P(-1.64 < Z < -1.02) = P(1.02 < Z < 1.64) = P(Z > 1.02) - P(Z > 1.64) = 0.154 - 0.051 = 0.103 (interpretation?)

(e) P(0 < Z < 1.96) = 0.5 - P(Z > 1.96) = 0.5 - 0.025 = 0.475 (interpretation?)

(f) P(-1.96 < Z < 1.96) = 2xP(0 < Z < 1.96) = 0.95 (interpretation?)

(g) P(-1.50 < Z < 0.67) = 1 - P(Z > 0.67) - P(Z < -1.50) = 1 - P(Z > 0.67) - P(Z > 1.50) = 1 - 0.251 - 0.067 = 0.682 (interpretation?)

(h) P(Z < -2.50) = P(Z > 2.50) = .006 (interpretation?)

The General Normal Distribution

E(X) = m and Var (X) = s2 i.e. X ~ N(m, s2)

Let Z = (X - m)/s

̃ E(Z) = 0 and Var (Z) = 1 (see previous notes if unconvinced) i.e. Z ~ N(0, 1)

Examples:

E(X) = 25, Var (X) = 25 [Standard Deviation (X) = 5]

X ~ N(25, 25)

(a) P(X > 20) = P((X - m)/s > (20 - 25)/5) = P(Z > -1.0) = 1 - P(Z > 1.0) = 1 - 0.159 = 0.841 (interpretation?)

(b) P(X < 40) = P((X - m)/s < (40 - 25)/5) = P(Z < 3.0) = 1 - P(Z > 3.0) » 1.0 (interpretation?)

(c) P(21 < X < 30) = P(-0.8 < Z < 1.0) = 1 - P(Z < -0.8) - P(Z > 1.0) = 1 - 0.212 - 0.159 = 0.629

(interpretation?)

(d) P(18 < X < 23) = P(-1.4 < Z < -0.4) = P(0.4 < Z < 1.4) = P(Z > 0.4) - P(Z > 1.4) = 0.345 - 0.081 = 0.264

(interpretation?)

IQ Example

IQ (the intelligence quotient) is Normally distributed with (population) mean 100 and (population) standard deviation 16.

IQ ~ N(100, 162) ̃ IQ – 100 ~ N(0, 1)

16

(a) What proportion of the population has an IQ above 120?

P(IQ > 120) = P(Z > [120-100]/16) = P(Z > 1.25) = .1056

(b) What proportion of the population has IQ between 90 and 110?

P(90 < IQ < 110) = P([90–100]/16 < Z < [110–100]/16) = P(-.625 < Z < .625) = 1 – {2 x P(Z > .625)} = 1 – {2 x .266} = .47

(c) In the past, about 10% of the population went to university. Now the proportion is about 30%. What was the IQ of the ‘marginal' student in the past? What is it now?

Note: Offensive assumption.

(i) P(IQ > ?) = .10 ̃ P(Z > [?-100]/16) = .10

̃ [?-100]/16 = 1.28 ̃ ? = 120.5

(ii) P(IQ > ?) = .30 ̃ P(Z > [?-100]/16) = .30

̃ [?-100]/16 = .525 ̃ ? = 108.4