251solnJ1 10/16/00 (Open this document in 'Page Layout' view!) J. Random Variables. 1. Definitions. 4.3. 2. Probability Distribution of a Discrete Random Variable. 4.11-4.14, 4.16. 3. Expected Value (Expectation) of a Discrete Random Variable. 4.22a, 4.25a, 4.27. J1. 4. Variance of a Discrete Random Variable. 4.22, 4.25, 4.26. J2-J4. 5. Summary J6, J7 Exercise 4.3: a, b and e are discrete. c, d and f are usually considered continuous. Exercise 4.11: a) By the extended addition rule for mutually exclusive events, we can add together the probabilities that we get the various values. If 2, 3, 5, 8, and 10 are the only values of x , the total P2 P3 P5 P8 P10 1 . Since P2 P3 P8 P10 .15 .10 .25 .25 .75 , P8 1 .75 .25 b) Px 2 x 10 P2 P10 .15 .25 .40 c) F 8 Px 8 P2 P3 P5 P8 1 P10 1 .25 .75 . Question: What is Px 8 ? Exercise 4.12: x Px a) 0 .1 1 2 .3 .3 b) x Px 2 .25 1 .50 0 .25 x Px 4 .3 c) 9 .4 20 .3 d) x 2 Px .15 3 .15 5 .45 To be valid, these must have positive probabilities 3 .2 1.00 .4 6 .35 .9 1.10 and the probabilities must add to 1. Only b) is valid. Exercise 4.13: If we write down the pattern, the value of x and its probability, we get pattern x probabilit y HHH 3 1 HHT 2 1 HTH 2 1 HTT 1 1 THH 2 1 THT 1 1 TTH 1 1 TTT 0 1 8 8 8 8 The probabilities, of course, add to 1, so that, if we add together the combinations 8 8 8 8 that will give us values of 0, 1, 2 or 3, we find P1 1 8 , P2 3 8 , P3 3 8 and P4 1 8 . Exercise 4.16: a) We can consider relative frequencies estimates of probabilities. Of course, they should be all between zero and one and add to 1. c) Px 30 P31 P32 P33 .1273 .1455 .1273 .4000 , Px 40 0 , Px 30 P20 P21 P22 P23 P24 P25 P26 P27 P28 P29 .0182 .0364 .0545 .0727 .0364 .0182 .0727 .0182 .1090 .0727 .5090 d) Px 25 x 26 Px 25 Px 26 .0182 .0727 .0909 251solnJ1 10/16/00 Exercise 4.22: x 1 2 4 10 From this table Px x P x .2 0.2 .4 0.8 .2 0.8 .2 2.0 1.0 3.8 Px 1.00 (valid distribution!), x 2 P x 0.2 1.6 3.2 20.0 25.0 E x x x P x -2.8 -1.8 0.2 6.2 -0.56 -0.72 0.04 1.24 0.00 2 x Px 25 .0 , xPx 3.8 , x 2 P x 1.568 1.296 0.008 7.688 10.560 x Px 0.00 (a check for accuracy!) and x 2 Px 10.560 . Remember! You do not need the last two columns if you use the computational formula. xPx 3.8 a) E x b) Computational Formula (the easy way!): Var x x2 E x 2 2 25 .0 3.82 10 .56 x 2 Px 2 x Px 10.560 . Definitional Formula (the hard way): Varx x2 E x 2 2 c) 10.56 3.2496 d) It is the average value that would be attained if the experiment described by the distribution were repeated many, many times. e) No!!! f) Depends on the distribution - If E x is the same as a value that could be assumed by x , it is possible. Exercise 4.25: Px x P x x 2 P x P y y P y y 2 P y y .3 0.0 0.0 0 .1 0.0 0.0 .4 0.4 0.4 1 .8 0.8 0.8 .3 0.6 1.2 2 .1 0.2 0.4 1.0 1.0 1.6 1.0 1.0 1.2 a) What did you guess? Did you notice that these are symmetrical, unimodal distributions around 1? b) Since x has more probability away from the mean, we would expect it to be more variable. c) Since these have identical means Ex 1.0 , variability is shown by the variance. x 0 1 2 Varx x2 xPx 2 x 1.6 1.02 0.6 , Var y y2 yPy 2 y 1.2 1.02 0.2 . Exercise 4.27: The spread is used to adjust your chance of winning so that the probability is .5. The probability of three wins in a row is .53 .125 . Your probability of losing is thus 1 - .125 = .875. That is, the bookie has a .125 chance of losing $5 and a .875 chance of winning $1. From his viewpoint, the mean x 1 and variance are: 5 Px xPx x 2 Px .875 0.875 .875 .125 0.625 3.125 1.000 4 0.25 3.9375 . 2 0.250 4.000 So the mean is $0.25 and the variance is 251solnJ1 10/16/00 Exercise 4.26: For Firm A we can compute: x Px xPx x 2 Px 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 .01 0 0 x 2450 .01 5 2500 Var x x2 E x 2 x2 x 2 Px x2 .01 10 10000 6440000 2450 2 437500 .02 30 45000 x 661 .44 .35 700 1400000 You should be also able to compute for Firm B : .30 750 1875000 y 2450 .25 750 2250000 .02 70 245000 Var y y2 E y 2 y2 y 2 P y y2 .01 40 160000 6495000 2450 2 492500 .01 45 202500 y 701 .78 .01 50 250000 1.00 2450 64400000 Thus Firm B has a larger variance, standard deviation and coefficient of variation and thus faces greater risk. Problem J1: Assume that values of x as shown below represent sales of your product. Sales Probability Except for part a do the parts below in two ways - first, x p(x) by figuring out each value of the function of x and its 0 .50 probability, and, second, by using a formula for a function 1 .10 of x . 2 .10 a. What is Ex ,E x 2 ? 3 .10 b. If costs are C1 10 x 10 , what is E C1 ? 4 .20 c. If costs are C2 0.5x 2 10x 10 , What is E C 2 ? d. If the product sells for $25 per unit, and R represents revenue, what is E R ? e. If Costs are described by C1 , and represents profit, what is E ? f. If Costs are described by C 2 , and represents profit, what is E ? Solution: To do these directly , make a table by first computing the values of the functions of x for each value of x and then multiplying each value by the probability of the corresponding value of x. x 0 1 2 3 4 px xpx .50 .10 .10 .10 .20 1.00 0 .10 .20 .30 .80 1.40 x x2 x 2 px 0 1 4 9 16 0 0.10 0.40 0.90 3.20 4.60 E x2 C1 10 20 30 40 50 a. From the table: Ex 1.40,E x 2 4.60 . b. Using the formula Eax b aEx b, 10 Ex 10 101.4 10 24 . C1 px 5 2 3 4 10 24 E C1 C2 10.0 19.5 28.0 35.5 42.0 EC1 E10 x 10 C 2 px 5.00 1.95 2.80 3.55 8.40 21.70 E C 2 R 0 25 50 75 100 Rpx 0 2.5 5.0 7.5 20.0 35.0 E R 251solnJ1 10/16/00 c. Using the formula E ax aEx , E .5x 2 .5E x 2 . Then, using the fact that you can add and subtract means, EC EC .5x EC E.5x EC .5Ex 2 2 24 .54.60 21.70 . d. The formula for revenue is 1 2 1 2 1 R 25x , so using the formula Eax aEx , ER E 25 x 25 E x 251.4 35 . e. Since we can add or subtract means, and R C , E E R E C1 35 24 11 . f. E E R E C2 35 21 .70 13.30 . Problem J2: If y 200x 50, x 3 and x2 10 , what are y and y2 ? Solution: The formula says that if y ax b, y a x b and y2 a 2 x2 , so, if a 200 and b 50 , then y 200 x 50 2003 50 650 , and y2 2002 x2 2002 10 400000 . Thus y 400000 632.46 . Problem J3:.I have measured the temperature at which successful outcomes of a biochemical process occur. It turns out that the sample mean is 72 degrees Fahrenheit and the sample standard deviation is 10 degrees Fahrenheit. My European affiliate needs this data. What are the mean and standard deviation in Celsius? Remember that if x f is the temperature in Fahrenheit and xc is the temperature in Celsius, x f 9 5 xc 32 Solution: x f 72 ,Var x f 10 2 100 . If we solve the equation above for Celsius temperature, we get xc 5 9 x f 1609 . Using the formula y ax b , and remembering E y aEx b and Var y a 2Varx , let y xc , x x f , a 5 9 , and b 1609 . Also, remember that we can use these formulas for sample means and variances, so that 2500 2 2 . xc ax f b 5 9 72 1609 22.22 and Varxc 5 9 Var x f 5 9 10 2 81 5 9 2 10 2 5 9 10 5.556 . So that the standard deviation of x c is Problem J4:.A package of a product sells for $21.00/lb. plus a $1.00 packing cost, so that A 2 lb. package would sell for $21.00(2)+$1.00 = $43.00. If the mean package size is 37 oz. and the standard deviation of package size is 3 oz. , what are the mean and standard deviation of package price? Solution: If we choose to work in pounds, let x be the weight in pounds, x 3716 and x 316 . The cost is y 21x 1 . Using the formula y ax b , and letting a 21 and b 1 , we can use our formulas 37 for the means and variances of y . y E y aEx b a x b 21 1 49 .56 and 16 2 3 15 .5039 . So that y 15.5039 3.9375 . 16 y2 Var y a 2Var x a 2 x2 212 We could also take the square root of the variance equation and say y a 2 x2 a x 3 21 15 .5039 . 16 251solnJ1 10/16/00 Problem J5: You are investing in a project with the following distribution of profits(in millions of dollars): x P(x) a. What is the mean and variance of the profits? 10 .500 b. Assume that you must pay me 50% of the profits plus a 20 .250 finder's fee of $1(million). What is the mean and variance of 30 .125 your profit? 40 .125 c. Use your answer in part b. to find the mean and variance of your profit if I raise my finder's fee to $2(million)? Solution: a) x 10 20 30 40 Px xPx x 2 Px .500 5.00 50 .0 .250 5.00 100 .0 .125 3.75 112 .5 .125 5.00 200 .0 1.000 18 .75 462 .5 From the table at left xPx 18.75 Var x E x x Px x 2 x 2 2 x 2 2 x 462 .5 18 .75 2 110 .9375 x 110 .9375 10 .53 b) Since x represents the profits, the formula for my share is 0.5x 1 . So your share is y x 0.5x 1 0.5x 1 . Using the formula y ax b , and letting a 0.5 and b 1 , we can use our formulas for the means and variances of y . y E y aEx b a x b 0.518.75 1 8.375 and y2 Var y a 2Varx a 2 x2 0.52 110.9375 27.7344 . (or y 4.892 ) c) Just as if y ax b , y E y aEx b a x b and y2 Var y a 2Var x a 2 x2 , so if w cy d , w Ew cE y d c x d and w2 Var w c 2Var y c 2 y2 . Our new return is w 1y 1 , so that c 1 and d 1 . Therefore w Ew c x d 18.375 1 7.375 and w2 Varw c2 y2 12 27.7344 27.7344. Problem J6: Assume that x is a standardized random variable. If y 5x 3 , find the mean and variance of y . Solution: By definition a standardized random variable has a mean of zero and a standard deviation of 1. x 0 and x 1 . y 5x 3 . Using the formula y ax b , and letting a 5 and b 3 , we can use our formulas for the means and variances of y . y E y aEx b a x b 50 3 3 and y2 Var y a 2Varx a 2 x2 52 12 25 . 251solnJ1 10/16/00 Problem J7: Consider the random variable x in the table below. Make a probability table for a standardized random variable z , created by subtracting x the mean (which is 4) and dividing by the standard deviation P x 0 1/16 (which is 2). Show that the mean and standard deviation of 2 1/4 z are 0 and 1. 4 3/8 6 1/4 8 1/16 x x4 Solution: We standardize a variable by using the formula z , so that x 2 is replaced by 2 2 24 z 1 . First we check the mean and variance of x . 2 x P x xPx x 2 P x From the table at left 0 1 0.00 0.00 16 x xPx 4.00 1 2 0.50 1.00 4 Var x x2 E x 2 x2 x 2 Px x2 3 4 1.50 6.00 8 20 .00 4.00 2 4.00 1 6 1.50 9.00 4 x 4.00 2.00 8 1 0.50 4.00 16 1.00 4.00 20 .00 Now we write out the variable z with its probabilities and compute its mean and standard deviation. z Pz zP z z 2 Pz From the table at left 2 1 0.125 0.25 16 z zP z 0.00 1 1 0.250 0.25 4 Var z z2 E z 2 z2 z 2 Pz z2 3 0 0.000 0.00 8 1.00 0.00 2 1.00 1 1 0.250 0.25 4 x 1.00 1.00 2 1 0.125 0.25 16 1.00 0.000 1.00