251y0731 4/30/07 ECO251 QBA1 THIRD EXAM Apr 18, 2007 Name: __KEY_______________ Student Number: _____________________ Part I. (16 points) Do all the following (2 points each unless noted otherwise). Make Diagrams! Show your work! z has the standardized Normal distribution z ~ N 0, 1 for the first four problems. Material in italics below is a description of the diagrams you were asked to make or a general explanation and will not be part of your written solution. The x and z diagrams should look similar. If you know what you are doing, you only need one diagram for each problem. General comment - I can't give you much credit for an answer with a negative probability or a probability above one, because there is no such x thing!!! In all these problems we must find values of z corresponding to our values of x before we do anything else. A diagram for z will help us because, if areas on both sides of zero are shaded, we will add probabilities, while, if an area on one side of zero is shaded and it does not begin at zero, we will subtract probabilities. Note: All the graphs shown here are missing a vertical line. They are also to scale. A hand drawn graph should exaggerate the distances of the points from the mean. Note also that, because of the rounding error necessary to use a conventional normal table, the results for x will disagree with results for z, especially for small values of z. 1. Pz 1.43 Pz 0 P1.43 z 0 .5 .4236 .0764 This is how we usually find a p-value for a left-sided test when the distribution of the test ratio is Normal. For z make a diagram. Draw a Normal curve with a mean at 0. Indicate the mean by a vertical line! Shade the entire area below -1.43. Because this is entirely on the left side of zero, we must subtract the area between -1.43 and zero from the area below zero. 1 251y0731 4/30/07 2. P0 z 3.21 .4993 For z make a diagram. Draw a Normal curve with a mean at 0. Indicate the mean by a vertical line! Shade the area between zero and 3.21. Because this is entirely on the right side of zero and starts at zero, it is precisely the sort of probability in Table 17. Because 3.21 is larger than the numbers on a conventional Normal table, there is a section below the usual table that says ‘If z 0 is between 3.18 and 3.21 P0 z z 0 .4993 . 3. P1.45 z 1.45 P1.45 z 0 P0 z 1.45 2 P0 z 1.45 =2(.4265) =.8530 For z make a diagram. Draw a Normal curve with a mean at 0. Indicate the mean by a vertical line! Shade the entire area between -1.45 and 1.45. Because this is on both sides of zero, we must add the area between -1.45 and zero to the area between zero and 1.45. 2 251y0731 4/30/07 4. z .14 Solution: z .14 is, by definition, the value of z with a probability of 14% above it. Make a diagram. The diagram for z will show an area with a probability of 86% below z .14 . It is split by a vertical line at zero into two areas. The lower one has a probability of 50% and the upper one a probability of 86% - 50% = 36%. The upper tail of the distribution above z .14 has a probability of 14%, so that the entire area above 0 adds to 36% + 14% = 50%. From the diagram, we want one point z .14 so that Pz z .14 .86 or P0 z z.14 .3600 . If we try to find this point on the Normal table, the closest we can come is P0 z 1.08 .3599 . So we will use z.14 1.08 , though 1.09 might be acceptable. x ~ N 3, 7.2 for problems 5 through 8. 1.43 3 Pz 0.62 Pz 0 P0.62 z 0 .5 .2324 .2676 5. Px 1.43 P z 7.2 For z make a diagram. Draw a Normal curve with a mean at 0. Indicate the mean by a vertical line! Shade the entire area below -1.43. Because this is entirely on the left side of zero, we must subtract the area between -1.43 and zero from the area below zero. If you wish, make a completely separate diagram for x . Draw a Normal curve with a mean at 3. Indicate the mean by a vertical line! Shade the area below -1.43. This area is completely to the left of the mean (3), so we subtract the smaller area between -1.43 and the mean from the larger area below the mean. 3 251y0731 4/30/07 3.21 3 0 3 z P 0.42 z 0.03 P0.42 z 0 P0 z 0.03 6. P0 x 3.21 P 7.2 7.2 .1628 .0120 .1748 For z make a diagram. Draw a Normal curve with a mean at 0. Indicate the mean by a vertical line! Shade the entire area between -0.42 and 0.03. Because this area is on both sides of zero, we must add the area between -0.42 and zero to the area from zero to 0.03. If you wish, make a completely separate diagram for x . Draw a Normal curve with a mean at 3. Indicate the mean by a vertical line! Shade the area from 0 to 3.21. This area is on both sides of the mean (3), so we add the area between 0 and the mean to the area between the mean and 3.21. 1.45 3 1.45 3 z P 0.62 z 0.22 7. P1.45 x 1.45 P 7.2 7.2 P0.62 z 0 P0.22 z 0 .2324 .0871 .1453 For z make a diagram. Draw a Normal curve with a mean at 0. Indicate the mean by a vertical line! Shade the entire area between -0.62 and -0.22. Because this area is on the left side of zero, we must subtract the area between -0.22 and zero from the larger area between -0.62 and zero. If you wish, make a completely separate diagram for x . Draw a Normal curve with a mean at 3. Indicate the mean by a vertical line! Shade the area from -1.45 to 1.45. This area is on the left side of the mean (3), so we subtract the area between 1.45 and the mean from the larger area between -1.45 and the mean. 4 251y0731 4/30/07 8. x.14 Solution: Since x ~ N 3, 7.2 , the diagram for x would show 86% probability below x.14 split in two regions on either side of 3 with probabilities of 50% below 3 and 36% above 3 and below x.14 , and with 14% above x.14 . On the first page we found z.14 1.08 , so the value of x can then be written x.14 z.14 3 1.087.2 3 7.776 10.776 . Check: 10 .776 3 Pz 1.08 Pz 0 P0 z 1.08 .5 .3599 .1401 14% Px 10.776 P z 7.2 5 251y0731 4/30/07 Part II: (15+ points) Do all the following: All questions are 2 points each except as marked. Exam is normed on 50 points including take-home. (Showing your work can give partial credit on some problems! In open-ended questions it is expected. Please indicate clearly what sections of the problem you are answering and what formulas you are using. Neatness counts!) Remember that you may not be able to finish this section, so ration your time on each problem. [Numbers in brackets are a cumulative total] 1. Which of the following are correct for a discrete distribution? (You may circle more than one.) a) P7 x 20 Px 20 Px 7 b) Px 0 1 Px 0 Sorry about this! This should have read Px 1 1 Px 0 in which case it would be correct for discrete distributions but not continuous distributions. Because of my error, if you say that this is correct, you will be given credit for the answer. c) Px 2 1 Px 2 But note that Px 2 Px 3 d) * Px 2 1 Px 2 e) * Px 6 Px 5 f) None of the above is correct. (1.5 for 2, 1 for 1, 0.5 off for each one wrong.) Explanation: The rule that you were given for a discrete distribution said Pa x b Px b Px a 1 , so P7 x 20 Px 20 Px 6 . Px 2 1 Px 1 2. Which of the following are correct for a continuous distribution? (You may circle more than one.) a) * P7 x 20 Px 20 Px 7 b) Px 0 1 Px 0 Note that Px 0 and Px 2 are not defined for a continuous distribution. c) Px 2 1 Px 2 d) * Px 2 1 Px 2 But note that Px 2 Px 2 e) Px 6 Px 5 f) None of the above is correct. (1 for 1, 0.5 off for each one wrong.) Explanation: The rule that you were given for a continuous distribution said Pa x b Px b Px a . So P7 x 20 Px 20 Px 7 . Px 0 and Px 2 do not exist in a continuous distribution. where all probabilities are defined over intervals. Px 6 Px 6 . Both of the above questions will have points taken off for wrong choices, though no question will be given a score below zero. 3. Which of the following is not a requirement of a discrete probability distribution? a) * Equally likely probability of a success. b) Sum of the possible outcomes is 1.00. c) The outcomes are mutually exclusive. d) The probability of each outcome is between 0 and 1. e) All of the above are required. f) None of the above is required. 6 251y0731 4/30/07 4. A company that sells annuities must base the annual payout on the probability distribution of the length of life of the participants in the plan. Suppose the probability distribution of the lifetimes of the participants is approximately a normal distribution with a mean of 68 years and a standard deviation of 3.5 years. What proportion of the plan recipients would receive payments beyond age 75? (2.5) [8.5] Solution: 75 68 Pz 2.00 Pz 0 P0 z 2.00 .5 .4772 .0228 Px 75 P z 3.5 For z make a diagram. Draw a Normal curve with a mean at 0. Indicate the mean by a vertical line! Shade the entire area above 2.00. Because this is entirely on the right side of zero, we must subtract the area between zero and 2.00 from the larger area above zero. 5. What type of probability distribution will the consulting firm most likely employ to analyze the insurance claims in the following problem? An insurance company has called a consulting firm to determine if the company has an unusually high number of false insurance claims. It is known that the industry proportion for false claims is 3%. The consulting firm has decided to randomly and independently sample 100 of the company’s insurance claims. They believe the number of these 100 that are false will yield the information the company desires. a) *Binomial distribution. b) Poisson distribution. c) Normal distribution. d) Hypergeometric distribution. e) Geometric distribution 6. If n = 10 and p = 0.70, then the standard deviation of the binomial distribution of x is a) 0.021 b) 0.07 c) 0.145 d) *1.45 e) 2.10 f) 7.00 g) 14.29 Explanation: The variance of the binomial distribution is 2 npq , where q 1 p . Here n = 10, p = 0.70 and q .30 . 2 10.70.30 2.10 . So 2.10 1.4491 1.45 7 251y0731 4/30/07 7. If n = 10 and p = 0.70, then the standard deviation of the binomial distribution of p a) b) c) d) e) f) g) x is n [14.5] 0.021 0.07 * 0.145 1.45 2.10 7.00 14.29 Explanation: The table in ‘Great Distributions I have Known’ says the following. Distribution Uses Formula Mean Variance Binomial Gives See above. Since probability of pq E p p x Var p p as proportion p n or x pn n of successes in n tries .70 .30 0.021 . So 0.021 0.14491 0.145 Here n = 10, p = 0.70 and q .30 . 2 10 8. What type of probability distribution will most likely be used to analyze warranty repair needs on new cars in the following problem? The service manager for a new automobile dealership reviewed dealership records of the past 20 sales of new cars to determine the number of warranty repairs he will be called on to perform in the next 90 days. Corporate reports indicate that the probability any one of their new cars needs a warranty repair in the first 90 days is 0.05. The manager assumes that calls for warranty repair are independent of one another and is interested in predicting the number of warranty repairs he will be called on to perform in the next 90 days for this batch of 20 new cars sold. [16.5] a) *Binomial distribution. b) Poisson distribution. c) Normal distribution. d) Hypergeometric distribution. e) Geometric distribution 8 251y0731 4/30/07 9. On the average, 3 customers per minute arrive at any one of the checkout counters of a convenience store. Checkout time takes 2 minutes. How many clerks are needed to be sure that the chance of someone having to wait is no more than 10%? Your answer should briefly explain what distribution you are using and how you got the number you give. (2.5) [19] Solution: The average for the 2 minute checkout period is 6. Since we are looking at numbers of successes when the mean is known, we are using the Poisson distribution. Here is the relevant part of the Poisson table. Poisson 6.0 k 0 1 2 3 4 5 6 7 8 9 10 11 12 P(x=k) 0.002479 0.014873 0.044618 0.089235 0.133853 0.160623 0.160623 0.137677 0.103258 0.068838 0.041303 0.022529 0.011264 Poisson 6.5 P(xk) 0.00248 0.01735 0.06197 0.15120 0.28506 0.44568 0.60630 0.74398 0.84724 0.91608 0.95738 0.97991 0.99117 k 0 1 2 3 4 5 6 7 8 9 10 11 12 P(x=k) 0.001503 0.009772 0.031760 0.068814 0.111822 0.145369 0.157483 0.146234 0.118815 0.085811 0.055777 0.032959 0.017853 Poisson 7.0 P(xk) 0.00150 0.01128 0.04304 0.11185 0.22367 0.36904 0.52652 0.67276 0.79157 0.87738 0.93316 0.96612 0.98397 k 0 1 2 3 4 5 6 7 8 9 10 11 12 P(x=k) 0.000912 0.006383 0.022341 0.052129 0.091226 0.127717 0.149003 0.149003 0.130377 0.101405 0.070983 0.045171 0.026350 P(xk) 0.00091 0.00730 0.02964 0.08177 0.17299 0.30071 0.44971 0.59871 0.72909 0.83050 0.90148 0.94665 0.97300 So Px 8 .84724 and Px 9 .91608 . This means Px 8 1 Px 8 1 .84724 .1528 and Px 9 1 Px 9 1 .91608 .0839 . So if we have 9 clerks the probability is below 10%. 10. 13 of a shipment of 15 hard disks are defective. 4 disks are inspected. What is the chance of finding at least one defective? Solution: This is Exercise 5.50 in the text. p 13 , n 4, N 15, M np 15 13 5 10! 1 1 6! 4! 1 10 9 8 7 1 .1538461 .84615 15! 15 14 13 12 11! 4! 11. (Extra Credit) A catalog company that receives the majority of its orders by telephone conducted a study to determine how long customers were willing to wait on hold before ordering a product. The length of time was found to be a random variable best approximated by an exponential distribution with a mean equal to 3 minutes. What proportion of customers will not be on the line after a wait of 4.5 minutes? a) 0.22313 b) 0.48658 c) 0.51342 d) *0.77687 e) The answer is nowhere near any of the above. Here is my answer. 1 1 Solution: F x 1 e cx when x 0 and the mean time to a success is . Let c . c 3 C 5 C 10 Px 1 1 P0 1 0 154 C 4 Px 4.5 F 4.5 1 4.5 1 e 3 1 e 1.5 1 .2231 .7769 12. Assume that there is an average of 6 chocolate chips per cookie. Now assume that an inspector will discard any cookies with fewer than 4 chips. About how many will be discarded from a batch of 100.[23] Solution: This is Exercise 5.35 in the text. The parameter is m 6.0. We discard cookies with fewer than 4 chips. Px 4 Px 3 .15120 . This represents the proportion that have fewer than 4 chips, so the quantity is 100 0.15120 15.120 . Seems like it’s 15 or 16. 9 251y0731 4/30/07 ECO251 QBA1 THIRD EXAM Apr 18, 2005 TAKE HOME SECTION Name: ____KEY____________________ Student Number: _________________________ Throughout this exam show your work! Please indicate clearly what sections of the problem you are answering and what formulas you are using. Part III. Do all the Following (19+ Points) Show your work! Neatness counts! 1. (Webster) Identify the distribution that you are using in each problem. If you have a number like n 20 g , make it very clear what value of n you are using. Look at the solved problems for Section L, the solution to Grass3 and ‘Great Distributions’ (especially the hints on the 3 rd page) before you start. a) The average number of calls that come into a switchboard is 1 each minute. Let g be the last digit of your student number. Assume that the operator is away from her desk for 10 g minutes (so that the time that you use will be between 5 and 9.5 minutes). What is 2 the average number of calls that will come into the switchboard in that time? (0) (i) What is the probability that at least one call is missed while the operator is away? (1) (ii) What is the probability that between 10 and 20 (inclusive) calls are missed while the operator is away? (1) b) You purchase parts from a supplier who has an average defect rate of 1 g units per lot of 100. (Your average for a lot of 100 will be between 1 and 10.) (i) You order 150 units (1.5 lots) , but say that you will not buy from this supplier again if more than 4 units are defective. What is the probability that you will not buy from this supplier again? (1) (Note: If g 9, you order 200 units.) (ii) The next time you order (if there is a next time), you resolve to take a sample of ten items from the thousand in this order. You will return the whole mess if more than two items are defective. If an average of 1 g units per lot of 100 means that the probability of an item being defective is 1 g % , what is the probability that you will return the order? (2) Note: To keep us on a level playing field, you must work out this answer by hand. Though it is strongly recommended that you check your answer against Table 14 and/or Table 15, no answer using tables will be accepted! [5] c) Lawn fertilizer comes in bags whose weight follows the continuous uniform distribution with a range between 23.8 and 26 0.2g pounds. You need 24.8 pounds to cover your lawn and buy only one bag. (Your upper number will be between 26 and 28.9 pounds.) (i) What is the probability that you will not have enough fertilizer in the bag? (1) (ii) If you buy two bags, what is the mean and standard deviation of their total weight? (1) [7] (iii) If the bags are distributed with a mean of 25.5 pounds and a range of 2 0.1g pounds, what is the probability that a bag will be between 20 and 24 pounds? (Your range will be between 2 and 2.9 pounds) (1) d) There are only 10 2g officers in your firm that are eligible to be on the Compensation Committee. Of these half make over $500,000.00 a year. (i) If you pick 4 officers at random, what is the probability that exactly three make over $500,000? (2) 10 251y0731 4/30/07 (ii) The shareholders will be very unhappy if they see that more than half of the members of the Compensation Committee make over $500,000. If you pick 4 officers at random, what is the probability that more than half make over $500,000? (Note that the answer is not the same as the answer to (i).) (1) [11] e) Only 65% of the many workers in a high security scientific installation seem to be capable of remembering to carry their badges. If they are among the 35% that forget they will be sent home to get them! (i) If the installation opens at 6am, what is the probability that the second employee to arrive is the first one sent home? (1) (ii) On the average, if we number the employees by their arrival order, what is the average number of the first employee to be sent home? (The third? The 4th? The 3.74th?) (1) (iii) If a group of 6 g employees arrive at 9am, what is the probability that at least one will not be sent home? (1) (iv) In the same group, what is the probability that at least one will be sent home? (1) (v) In the same group, what is the probability that more than half will not be sent home? (1) [16] (vi) (Extra Credit) By late morning, the employees are trickling in at only 12 per hour. The guard decides that if the probability of someone arriving in the next 5 minutes after you come in is less than 50%, he will hide behind the warning sign and take a few nips from his bottle. He asks you whether the probability is less than 50%. Do not use the Poisson distribution for this problem. (2) 2. As everyone knows, a jorcillator has two components, a phillinx and a flubberall. It seems that the jorcillator only works as long as both components work (so that it fails in the first month if either component fails). Note change. The probability of the phillinx failing is given by a Normal distribution with 2 .1h and 0.5 .01h , where h is the second-to-last digit of your student number. For example, if the life of the phillinx is represented by x1 , the chance of the phillinx failing in the 3rd month is P2 x1 3 and the probability of it failing after the third month is Px1 3 . For example: Ima Badrisk has the number 375292, so her distribution has a mean of 2 .1h 2.9 and a standard deviation of 0.5 .019 0.59 . Your mean will be between 2.0 and 2.9 and your standard deviation will be between 0.50 and 0.59. The probability of the flubberall failing is: in the first month 40%; in the second month 30%; in the third month 20% and after the third month 10%. , We will divide time into four periods with the last period being ‘beyond the third month.’ The jorcillator is guaranteed to fail in one of the four periods. Failure of components is assumed to be independent, so if the probability of the phillinx failing in the first month is .1, and the probability of the flubberall failing in the first month is .4, the probability of both components failing in the first month is (.1) (.4) =.04 (This is not the necessarily the probability that the jorcillator will fail in the first month!) In order to maintain my sanity, use the following events. Failure of the phillinx in period 1, 2, 3, 4 are events A1, A2 , A3 , and A4 . Failure of the flubberall in period 1, 2, 3, 4 are events B1, B2 , B3 , and B4 . Failure of the jorcillator in period 1, 2, 3, 4 are events C1, C2 , C3 , and C 4 . a) What is the probability that the phillinx will fail in month 1? Month 2? Month 3? After Month 3? (2) b) What is the probability that the jorcillator will fail in the first month? (2) c) What is the probability that the jorcillator will fail in the second month? (1) d) What is the probability that the jorcillator will fail in the third month? (1) 11 251y0731 4/30/07 e) What is the probability that the jorcillator will last beyond 3 months? (1) [21] If you haven’t figured it out already, one of the easiest ways to do this is to make a joint probability table. Put the A events across the top. Put the B events down the side. Figure out what the probability of the joint events must be if they are independent. Now make a similar table. This time, instead of probabilities, fill in the period in which the jorcillator fails. f) (Extra Credit) Find the probability that the jorcillator and the Phillinx both fail in the third month (1) g) (Extra Credit) Find the probability that the phillinx fails in the third month, given that the jorcillator fails in the third month i.e. P A3 C3 (1) h) (Extra Credit) Demonstrate Bayes’ rule by finding showing how to get the probability that the jorcillator fails in the third month, given that the phillinx fails in the third month, two ways. The first would be directly by working with joint probabilities that both the jorcillator and the phillinx fail. The second would be by using Bayes’ rule and your result in g). (1) 3. (Extra Credit) Do the following Binomial problems. If you substitute another distribution for the Binomial, justify the substitution by showing that a substitution condition is satisfied. (6) a) 45% of our many management employees have MBAs. If we select 150 at random, what is the probability that exactly 72 have college degrees? b) 45% of our many management employees have MBAs. If we select 150 at random, what is the probability that more than half have college degrees? c) 5% of our many management employees have MBAs. If we select 90 at random, what is the probability that more than 6 have MBAs? 12 251y0731 4/30/07 1. (Webster) Identify the distribution that you are using in each problem. If you have a number like n 20 g , make it very clear what value of n you are using. Look at the solved problems for Section L, the solution to Grass3 and ‘Great Distributions’ (especially the hints on the 3 rd page) before you start. a) The average number of calls that come into a switchboard is 1 each minute. Let g be the last digit of your student number. Assume that the operator is away from her desk for 10 g minutes. What is the average number of calls that will come into the switchboard in 2 that time? (0) (i) What is the probability that at least one call is missed while the operator is away? (1) (ii) What is the probability that between 10 and 20 (inclusive) calls are missed while the operator is away? (1) Solution: The number of minutes she will be away is between 5 and 8.5. The average number of calls that come in while she is away will be between 5 and 8.5. This is a Poisson problem because we are trying to track the number of successes when the mean is given, (i) Px 1 1 P0 If m 5 , Px 1 1 .00674 .9933 If m 8.5 , Px 1 1 .00020 .9998 (ii) P10 x 20 Px 20 Px 9 If m 5 , P10 x 20 1.00000 .96817 .0318 If m 8.5 , P10 x 20 .99979 .65297 .3468 b) You purchase parts from a supplier who has an average defect rate of 1 g units per lot of 100. (i) You order 150 units (1.5 lots), but say that you will not buy from this supplier again if more than 4 units are defective. What is the probability that you will not buy from this supplier again? (1) (Note: If g 9, you order 200 units.) (ii) The next time you order (if there is a next time), you resolve to take a sample of ten items from the thousand in this order. You will return the whole mess if more than two items are defective. If an average of 1 g units per lot of 100 means that the probability of an item being defective is 1 g % , what is the probability that you will return the order? (2) Note: To keep us on a level playing field, you must work out this answer by hand. Though it is strongly recommended that you check your answer against Table 14 and/or Table 15, no answer using tables will be accepted! [5] Solution: (i) The average defects per lot are between 1 and 9 per lot if you buy 1.5 lots. This means that the mean is between 1.5 and 13.5. There is an error here – the 200 should have been for 1 g 9 . Using a Poisson distribution Px 4 1 Px 4 The average defects per lot are 10 per lot if you buy 2 lots. This means that the mean is 20. If m 1.5 , Px 4 1 .98142 .01858 If m 12 .5 , Px 4 1 .00535 .9465 If m 20 , Px 4 1 .00002 .99998 (ii) p will be between 1% and 10%. You will reject the shipment if x, the number of defective items is above 2. Px 2 1 Px 2 1 P0 P1 P2 . This is a binomial problem where n 10 , Px C xn p x q n x . P0 C010 p 0 q10 q10 , P1 C110 p1q 9 10 p1q 9 and P2 C 210 p 2 q 8 10 9 2 8 p q 45 p 2 q 8 . Computations 2 1 follow for all values of p . 13 251y0731 4/30/07 p .01 .02 .03 .04 .05 .06 .07 .08 .09 .10 q 0.99 0.98 0.97 0.96 0.95 0.94 0.93 0.92 0.91 0.90 P0 0.904382 0.817073 0.737424 0.664833 0.598737 0.538615 0.483982 0.434388 0.389416 0.348678 P 1 P2 0.091352 0.166750 0.228069 0.277014 0.315125 0.343797 0.364288 0.377729 0.385137 0.387420 0.004152 0.015314 0.031742 0.051940 0.074635 0.098750 0.123388 0.147807 0.171407 0.193710 Px 2 0.999886 0.999136 0.997235 0.993786 0.988496 0.981162 0.971658 0.959925 0.945960 0.929809 Px 2 0.0001138 0.0008639 0.0027649 0.0062137 0.0115036 0.0188378 0.0283421 0.0400754 0.0540400 0.0701908 c) Lawn fertilizer comes in bags whose weight follows the continuous uniform distribution with a range between 23.8 and 26 0.2g pounds. You need 24.8 pounds to cover your lawn and buy only one bag. (i) What is the probability that you will not have enough fertilizer in the bag? (1) Solution: If g 0, the distribution goes from 23.8 to 26 pounds. 24 .8 23 .8 .45 26 23 .8 If g 9, the distribution goes from 23.8 to 27.8 pounds. Px 24 .8 24 .8 23 .8 .25 27 .8 23 .8 Make a diagram! You should show a box between 23.8 and a number that is between 26 and 27.8 pounds. If this number is 26 + .2g, the height of the box will be 1 1 . Shade the area in the box between 23.8 and 24.8. Since the 26 .2 g 23 .8 2.2 .2 g 1 length of this area is 1, the shaded area is 2 .2 .2 g Px 24 .8 (ii) If you buy two bags, what is the mean and standard deviation of their total weight? (1) d c . For two bags, if x and x are independent cd and 2 1 2 2 12 and identical. Ex1 x 2 Ex1 Ex 2 2 and Var x1 x 2 Solution: 2 Varx1 Varx 2 2 2 . So StDevx1 x 2 2 2 If g 0, the distribution goes from 23.8 to 26 pounds. 26 23.8 0.403333 23 .8 26 24 .9 and 2 2 12 Ex1 x 2 224.9 49.8 and StDevx1 x 2 20.403333 0.8981 2 If g 9, the distribution goes from 23.8 to 27.8 pounds. 27.8 23.82 1.333333 23 .8 27 .8 25 .8 and 2 2 12 Ex1 x 2 225.8 51.6 and StDevx1 x 2 21.33333 1.6330 [7] (iii) If the bags are distributed with a mean of 25.5 pounds and a range of 2 0.1g pounds, what is the probability that a bag will be between 20 and 24 pounds? (1) Solution: If g 0, 2 0.1g 2 and half the range is 1, the distribution goes from 25.5 – 1 = 24.5 to 25.5 + 1 = 26.5 pounds. Make a diagram! The box for this 14 251y0731 4/30/07 distribution will be between 24.5 and 26.5 pounds. This means that there will be no area to shade between 20 and 24 pounds. P20 x 24 0 . If g 9, 2 g 2.9 and half the range is 1.45, the distribution goes from 25.5 – 1.45 = 24.05 to 25.5 + 1.45 = 26.95 pounds. Make a diagram! The box for this distribution will be between 24.05 and 26.95 pounds. This means that there will be no area to shade between 20 and 24 pounds. P20 x 24 0 . d) There are only 10 2g officers in your firm that are eligible to be on the Compensation Committee. Of these half make over $500,000.00 a year. (i) If you pick 4 officers at random, what is the probability that exactly three make over $500,000? (2) 5! 5 C 35 C15 3!2! Solution: If g 0, 10 2g 10 and half of this is 5. P 3 10! C 410 6! 4! 5 43 5 60 5 4 3 2 1 .2381 10 9 8 7 5040 4 3 2 1 If g 9, 10 2g 28 and half of this is 14. P3 C 314 C114 C 428 14! 14 11! 3! 28! 22!4! 14 13 12 14 2184 14 4 122304 3 2 1 .2489 28 27 26 25 491400 491400 4 3 2 1 (ii) The shareholders will be very unhappy if the see that more than half of the members of the Compensation Committee make over $500,000. If you pick 4 officers at random, what is the probability that more than half make over $500,000? (Note that the answer is not the same as the answer to (i).) (1) [11] 5! C 45 C 05 4 !1! Solution: If g 0, 10 2g 10 and half of this is 5. P4 10 10! C4 6! 4! 5 4 3 2 120 4 3 2 1 .0238 . P3 P4 .2381 .0238 .2689 10 9 8 7 5040 4 3 2 1 If g 9, 10 2g 28 and half of this is 14. P4 C 414 C 014 C 428 14! 1 10! 4! 28! 24!4! 14 13 12 11 4 3 2 1 24024 .0489 P3 P4 .2489 .0489 .2978 28 27 26 25 491400 4 3 2 1 Why does the probability rise as the group gets larger? If there are only 10 eligible officers, the probability that the first pick yields someone with earnings above $100,000 15 251y0731 4/30/07 5 1 . If we get someone in the high-earnings group, the chance that the second pick 10 2 4 yields one too is .4444 . ? If there are 28 eligible officers, the probability that the first 9 14 1 . If we get someone in the pick yields someone with earnings above $100,000 is 28 2 13 .4815 , so we high-earnings group, the chance that the second pick yields one too is 27 can see that the larger the pool of high earnings people, the more slowly it will get depleted. is e) Only 65% of the many workers in a high security scientific installation seem to be capable of remembering to carry their badges. If they are among the 35% that forget they will be sent home to get them! (i) If the installation opens at 6am, what is the probability that the second employee to arrive is the first one sent home? (1) Geometric: P2 .65.35 .2275 (ii) On the average, if we number the employees by their arrival order, what is the average number of the first employee to be sent home? (The third? The 4 th? The 3.74th?) 1 1 2.857 (1) Geometric: p .35 (iii) If a group of 6 g employees arrive at 9am, what is the probability that at least one will not be sent home? (1) You need 1 Pall 1 .65n . If the probability of forgetting is .65, the probability of all forgetting is Pall .65n . If n 6 , 1 .65 6 1 .075419 .9246 . If we want to use the binomial table with n 6 and p .35 , what we want is the probability that at least one remembers. This is 1 P0 1 .07542 .9246 If n 15 , 1 .6515 1 .001562 .9984 . You could also get this by taking 1 P0 using the Binomial table when n 15 and p .35 . (iv) In the same group, what is the probability that at least one will be sent home? (1) Binomial: n is between 6 and 15. You need 1 Pall 1 .35n . If the probability of remembering is .35, the probability of all remembering is Pall .35n . If n 6 , 1 .35 6 1 .001838 .9982 .Note that if there are 6 individuals, at least one will be sent home if 5 or fewer remember. According to the binomial table with n 6 and p .35 , Px 5 .99816 . 1 .3515 1 .00000 1.0000 . Note that you could also get this by taking Px 14 using the Binomial table when n 15 and p .35 . (v) In the same group, what is the probability that more than half will not be sent home? (1) For a group of 6, the probability that any given person is sent home is .35 and a majority is 4. Px 4 1 Px 3 1 .80015 .19985 according to the binomial table with n 6 and p .35 . For a group of 15, the probability that any given person is sent home is .35 and a majority is 8. Px 8 1 Px 7 1 .88677 .11323 using the Binomial table when n 15 and p .35 16 251y0731 4/30/07 (vi) (Extra Credit) By late morning, the employees are trickling is at only 12 per hour. The guard decides that if the probability of someone arriving in the next 5 minutes after you come in is less than 50%, he will hide behind the warning sign and take a few nips from his bottle. He asks you whether the probability is less than 50%. Do not use the Poisson distribution for this problem. (2) Solution: For the exponential distribution, the probability that the waiting time is less 1 than or equal to x is F x 1 ecx when x 0 and the mean time to a success is . If c 1 12 people arrive in an hour, the average time to a success is 5 minutes. If 5, c .2 c .25 1 and F 5 1 e 1 e 1 .3679 .6321. No drink! 2. As everyone knows, a jorcillator has two components, a phillinx and a flubberall. It seems that the jorcillator only works as long as both components work (so that it fails in the first month if either component fails). Note change. The probability of the phillinx failing is given by a Normal distribution with 2 .1h and 0.5 .01h , where h is the second-to-last digit of your student number. For example, if the life of the phillinx is represented by x1 , the chance of the phillinx failing in the 3rd month is P2 x1 3 and the probability of it failing after the third month is Px1 3 . For example: Ima Badrisk has the number 375292, so her distribution has a mean of 2 .1h 2.9 and a standard deviation of 0.5 .019 0.59 . Your mean will be between 2.0 and 2.9 and your standard deviation will be between 0.50 and 0.59. The probability of the flubberall failing is: in the first month 40%; in the second month 30%; in the third month 20% and in the last month 10%. We will divide time into four periods with the last period being ‘beyond the third month.’ The jorcillator is guaranteed to fail in one of the four periods. Failure of components is assumed to be independent, so if the probability of the phillinx failing in the first month is .1, and the probability of the flubberall failing in the first month is .4, the probability of both components failing in the first month is (.1) (.4) =.04 (This is not the necessarily the probability that the jorcillator will fail in the first month!) In order to maintain my sanity, use the following events. Failure of the phillinx in period 1, 2, 3, 4 are events A1, A2 , A3 , and A4 . Failure of the flubberall in period 1, 2, 3, 4 are events B1, B2 , B3 , and B4 . Failure of the jorcillator in period 1, 2, 3, 4 are events C1, C2 , C3 , and C 4 . If you haven’t figured it out already, one of the easiest ways to do this is to make a joint probability table. Put the A events across the top. Put the B events down the side. Figure out what the probability of the joint events must be if they are independent. Now make a similar table. This time, instead of probabilities, fill in the period in which the jorcillator fails. 17 251y0731 4/30/07 a) What is the probability that the phillinx will fail in month 1? Month 2? Month 3? After Month 3? (2) We have 2 .1h , .5 .01h 1 2 02 z If h 0 , P0 x 1 P P 4.00 z 2.00 . 5 .5 P4.00 z 0 P2.00 z 0 = .5000 - .4772 .0228 P A1 22 1 2 P1 x 2 P z P 2.00 z 0 = .4772 P A2 .5 .5 3 2 22 P2 x 3 P z P0 z 2.00 = .4772 P A3 .5 .5 3 2 P x 3 P z Pz 2 Pz 0 P0 z 2.00 =.5 - .4772 = .0228 P A4 .5 Check: .0228 + .4772 + .4772 + .0228 = 1.0000 1 2.5 0 2.5 z If h 5 , P0 x 1 P P 4.54 z 2.72 .55 .55 P4.54 z 0 P2.72 z 0 = .5000 - .4967 = .0033 P A1 2 2.5 1 2.5 P1 x 2 P z P 2.73 z 0.91 . 55 .55 P2.72 z 0 P0.91 z 0 = .4967 - .3186 = .1781 P A2 3 2.5 2 2.5 P2 x 3 P z P 0.91 z 0.91 .55 .55 P0.91 z 0 P0 z 0.91 2.3186 = .6372 P A3 3 2.5 Px 3 P z Pz .91 Pz 0 P0 z 0.91 = .5 - .3186 = .1814 P A4 .55 Check: .0033 + .1781 + .6372 + .1814 = 1.0000 1 2.9 0 2.9 x If h 9 , P0 x 1 P P 4.91 z 3.22 .59 .59 P4.91 z 0 P3.22 z 0 = .5000 - .4994 = .0006 P A1 2 2.9 1 2.9 P1 x 2 P x P 3.22 z 1.53 .59 .59 P3.22 z 0 P1.53 z 0 = .4994 - .4370 = .0624 P A2 3 2.9 2 2.9 P2 x 3 P x P 1.53 z 0.17 .59 .59 P1.53 z 0 P0 z 0.17 = .4370 + .0675 = .5045 P A3 3 2.9 Px 3 P z Pz 0.17 Pz 0 P0 z 0.17 .5 .0625 P A4 .59 Check: .0006 + .0624 + .5045 + .4325 = 1.0000 b) What is the probability that the jorcillator will fail in the first month? (2) c) What is the probability that the jorcillator will fail in the second month? (1) d) What is the probability that the jorcillator will fail in the third month? (1) 18 251y0731 4/30/07 e) What is the probability that the jorcillator will last beyond 3 months? (1) Assume that h 0 . The solution for parts b) – e) follows. For the Flubberall PB1 .4 , PB2 .3 , PB3 .2 and PB3 .1 For the Phillinx, if h 0 we have found P A1 0228, P A2 .4772, P A3 .4772 and P A4 .0228 The way this was done in the last exam was as below. Remember that our new failure rule says whichever component fails first downs the machine. Joint Event Probability When Fails Period 1 .0228 .4 = .00912 A1 B1 A1 A1 A1 A2 A2 A2 A2 A3 A3 A3 A3 A4 A4 A4 A4 .0228 .3 = .00684 .0228 .2 = .00456 .0228 .1 = .00228 .4772 .4 = .19088 .4772 .3 = .14316 .4772 .2 = .09544 .4772 .1 = .04772 .4772 .4 = .19088 .4772 .3 = .14316 .4772 .2 = .09544 .4772 .1 = .04772 .0228 .4 = .00912 .0228 .3 = .00684 .0228 .2 = .00456 .0228 .1 = .00228 B2 B3 B4 B1 B2 B3 B4 B1 B2 B3 B4 B1 B2 B3 B4 Period 1 Period 1 Period 1 Period 1 Period 2 Period 2 Period 2 Period 1 Period 2 Period 3 Period 3 Period 1 Period 2 Period 3 Period 4 I had suggested using a joint probability table for the joint probabilities and a similar table for failure times. This sure saves space. You can also see from the failure time table that the probability of failure in the first period is P A1 B1 . However, if you are going to try something similar for the next period, note that you have to exclude failure in the first period. The result is a nightmare and some difficult logic. P A2 B2 A1 B1 P A2 B2 A1 B1 P A2 A1 B1 B2 A1 B1 P A2 B1 A1 B2 P A2 B1 P A1 B2 P A2 B2 .30000 .00684 .4772 .19088 .14316 .43632 . Here is the joint probability table and the table relating joint events to failure times. A1 A2 A3 A4 B1 B2 B3 B4 .00912 .19088 .19088 .00912 .40000 .00684 .14316 .14316 .00684 .30000 .00456 .09544 .09544 .00456 .20000 .00228 .04772 .04772 .00228 .10000 B1 .0228 .4772 .4772 .0228 1.0000 A1 A2 A3 A4 1 1 1 1 B2 B3 B4 1 2 2 2 1 2 3 3 1 2 3 4 b) What is the probability that the jorcillator will fail in the first month? (2) PC1 P A1 B1 A1 B2 A1 B3 A1 B4 A2 B1 A3 B1 A4 B1 19 251y0731 4/30/07 P A1 B1 = .02280 + .40000 - .00912 = .41368 c) What is the probability that the jorcillator will fail in the second month? (1) PC 2 P A2 B2 A2 B3 A2 B4 A3 B2 A4 B2 = .14316 + .09544 + .04772 + .14316 + .00684 = .43632 d) What is the probability that the jorcillator will fail in the third month? (1) PC 3 P A3 B3 A3 B4 A4 B3 .09544 .04772 .00456 .14772 e) What is the probability that the jorcillator will last beyond 3 months? (1) PC 4 P A4 B4 .00228 Note that these probabilities add to 1. Assume that h 0 . The solution for parts f) – h) follows. f) (Extra Credit) Find the probability that the jorcillator and the Phillinx both fail in the third month (1) The problem asks for PC 3 A3 . C 3 is the union of three mutually exclusive events, two of which involve A3 . So PC 3 A3 P A3 B3 A3 B4 .09544 .04772 .14316 . Note that because C 3 depends on A3 , PC 3 A3 PC3 P A g) (Extra Credit) Find the probability that the phillinx fails in the third month, given that the jorcillator fails in the 3rd month i.e. P A3 C3 (1) PA3 C3 P A3 C 3 .14316 .96913 PC 3 .14772 h) Demonstrate Bayes’ rule by finding showing how to get the probability that the jorcillator fails in the third month, given that the phillinx fails in the third month, two ways. The first would be directly by working with joint probabilities that both the jorcillator and the phillinx fail. The second would be by using Bayes’ rule and your result in g). (1) P A3 C 3 .14316 .3 P C3 A3 P A3 ..4772 PC3 A3 PA3 C3 PC3 P A3 .96913 .14772 .3 ..4772 Assume that h 5 . The solution for parts b) – e) follows. For the Flubberall PB1 .4 , PB2 .3 , PB3 .2 and PB3 .1 For the Phillinx, if h 5 we have found that P A1 .0033, P A2 .1781, P A3 .6372 and P A4 .1814. Here is the joint probability table and the table relating joint events to failure times. B1 B2 B3 B4 B1 B 2 B3 B 4 A1 A2 A3 A4 .00132 .07124 .25488 .07256 .40000 .00099 .05343 .19116 .05442 .30000 .00666 .03562 .12744 .03628 .20000 .00033 .01781 .06372 .01814 .10000 .0033 .1781 .6372 .1814 1.0000 A1 A2 A3 A4 1 1 1 1 1 2 2 2 1 2 3 3 1 2 3 4 b) What is the probability that the jorcillator will fail in the first month? (2) 20 251y0731 4/30/07 PC1 P A1 B1 A1 B2 A1 B3 A1 B4 A2 B1 A3 B1 A4 B1 P A1 B1 = .00330 + .40000 - .00132 = .40198. c) What is the probability that the jorcillator will fail in the second month? (1) PC 2 P A2 B2 A2 B3 A2 B4 A3 B2 A4 B2 = .05343 + .03562 + .01781 + .19116 + .05442 = .35244. d) What is the probability that the jorcillator will fail in the third month? (1) PC 3 P A3 B3 A3 B4 A4 B3 .12744 .06372 .03628 .22744 . e) What is the probability that the jorcillator will last beyond 3 months? (1) PC 4 P A4 B4 .01814 Note that these probabilities add to 1. Assume that h 9 . The solution for parts b) – e) follows. For the Flubberall PB1 .4 , PB2 .3 , PB3 .2 and PB3 .1 For the Phillinx, if h 9 we have found that For the Phillinx P A1 .0006, P A2 .0624, P A3 .5045 and P A4 .4325. Here is the joint probability table and the table relating joint events to failure times. B1 B2 B3 B4 B1 B 2 B3 B 4 A1 A2 A3 A4 A1 .00024 .00018 .00012 .00006 .0006 1 1 1 1 A2 .02496 .01872 .01248 .00624 .0624 1 2 2 2 A3 .20180 .15135 .10090 .05045 .5045 1 2 3 3 A4 .17300 .12975 .08650 .04325 .4325 1 2 3 4 .40000 .30000 .20000 .10000 1.0000 b) What is the probability that the jorcillator will fail in the first month? (2) PC1 P A1 B1 A1 B2 A1 B3 A1 B4 A2 B1 A3 B1 A4 B1 P A1 B1 = .0006 + .40000 - .00024 = .40036 c) What is the probability that the jorcillator will fail in the second month? (1) PC 2 P A2 B2 A2 B3 A2 B4 A3 B2 A4 B2 = .01872 + .01248 + .00624 + .15135 + .12975 = .31854 d) What is the probability that the jorcillator will fail in the third month? (1) PC 3 P A3 B3 A3 B4 A4 B3 .10090 .05045 .08650 .23785 e) What is the probability that the jorcillator will last beyond 3 months? (1) PC 4 P A4 B4 .04325 Note that these probabilities add to 1. Assume that h 9 . The solution for parts f) – h) follows. f) (Extra Credit) Find the probability that the jorcillator and the Phillinx both fail in the third month (1) The problem asks for PC 3 A3 . C 3 is the union of three mutually exclusive events, two of which involve A3 . So PC 3 A3 P A3 B3 A3 B4 .10090 .05045 .15135 g) (Extra Credit) Find the probability that the phillinx fails in the third month, given that the jorcillator fails in the 3rd month i.e. P A3 C3 (1) 21 251y0731 4/30/07 PA3 C3 P A3 C 3 .15135 .63633 PC 3 .23785 h) Demonstrate Bayes’ rule by finding showing how to get the probability that the jorcillator fails in the third month, given that the phillinx fails in the third month, two ways. The first would be directly by working with joint probabilities that both the jorcillator and the phillinx fail. The second would be by using Bayes’ rule and your result in g). (1) P A3 C 3 .15135 .3 P C3 A3 P A3 .5045 PC3 A3 PA3 C3 PC3 P A3 .63633 .23785 .3 .5045 3. (Extra Credit) Do the following Binomial problems. If you substitute another distribution for the Binomial, justify the substitution by showing that a substitution condition is satisfied. (6) All problems below are done first using a continuity correction. It is obviously necessary in a). a) 45% of our many management employees have MBAs. If we select 150 at random, what is the probability that exactly 72 have college degrees? We need the binomial probability of 72 successes when the expected number of successes is np 150 .45 67.5 , the expected number of failures is 150 – 67.5 = 82.5 and npq 67 .5.55 37 .125 6.09303 . We cannot use the Poisson distribution here because the probability is large. Since the expected numbers of successes and failures are both above 5, we can use the Normal distribution. 72.5 67.5 71.5 67.5 PB 72 PN 71 .5 x 72 .5 P z P0.66 z 0.82 6 . 09303 6.09303 150! 150 72 78 .2939 .2454 .0485 . This can also be done as C72 .45 72 .55 78 , but this p q 78!72! looks like much too much work. b) 45% of our many management employees have MBAs. If we select 150 at random, what is the probability that more than half have college degrees? 75 .5 67 .5 PB x 76 PN x 75 .5 P z Pz 1.31 .5 .4049 .0951 6.09303 Without the continuity correction 76 67 .5 PB x 76 PN x 76 P z Pz 1.40 .5 .4192 .0808 6.09303 Or if p x , E p p .45 and p n pq .45.55 .00165 .04062 n 150 .5 .45 PN p .5 P z Pz 1.23 .5 .3907 .1093 0.04062 c) 5% of our many management employees have MBAs. If we select 90 at random, what is the probability that more than 6 have MBAs? Note that the expected number of successes, np 90.05 4.5 , is below 5 so that we should not use the Normal distribution. However, the criterion for use of the Poisson distribution n 90 1800 . If we use the Poisson table with a parameter of 4.5 500 is met since .05 p Px 6 1 Px 6 1 .83105 .16895 . 22