Chapter06
- 2. A chance experiment is any activity or situation in which there is uncertainty about which of two or more possible outcomes will result. The collection of all possible outcomes of a chance experiment is the sample space for the experiment. Chance Experiment & Sample Space
- 3. Example An experiment is to be performed to study student preferences in the food line in the cafeteria. Specifically, the staff wants to analyze the effect of the student’s gender on the preferred food line (burger, salad or main entrée).
- 4. Example - continued The sample space consists of the following six possible outcomes. A male choosing the burger line. A female choosing the burger line. A male choosing the salad line. A female choosing the salad line. A male choosing the main entrée line. A female choosing the main entrée line.
- 5. Example - continued The sample space could be represented by using set notation and ordered pairs. sample space = {(male, burger), (female, burger), (male, salad), (female, salad), (male, main entree), (female, main entree)} If we use M to stand for male, F for female, B for burger, S for salad and E for main entrée the notation could be simplified to sample space = {MB, FB, MS, FS, ME, FE}
- 6. Example - continued Yet another way of illustrating the sample space would be using a picture called a “tree” diagram. This “tree” has two sets of “branches” corresponding to the two bits of information gathered. To identify any particular outcome of the sample space, you traverse the tree by first selecting a branch corresponding to gender and then a branch corresponding to the choice of food line. Male Female Burger Salad Main Entree Outcome (Male, Salad) Outcome (Female, Burger) Burger Salad Main Entree
- 7. Events An event is any collection of outcomes from the sample space of a chance experiment. A simple event is an event consisting of exactly one outcome. If we look at the lunch line example and use the following sample space description {MB, FB, MS, FS, ME, FE} The event that the student selected is male is given by male = {MB, MS, ME} The event that the preferred food line is the burger line is given by burger = {MB, FB} The event that the person selected is a female that prefers the salad line is {FS}. This is an example of a simple event.
- 8. Venn Diagrams A Venn Diagram is an informal picture that is used to identify relationships. The collection of all possible outcomes of a chance experiment are represented as the interior of a rectangle. The rectangle represents the sample space and shaded area represents the event A.
- 9. Forming New Events Let A and B denote two events. The shaded area represents the event not A.
- 10. Forming New Events Let A and B denote two events. The shaded area represents the event A B.
- 11. Forming New Events Let A and B denote two events. The shaded area represents the event A B.
- 12. More on intersections Two events that have no common outcomes are said to be disjoint or mutually exclusive . A and B are disjoint events
- 13. More than 2 events Let A 1 , A 2 , …, A k denote k events These k events are disjoint if no two of them have any common outcomes.
- 14. Some illustrations A B C A, B & C are Disjoint A B C A B C A B C
- 15. Probability – Classical Approach If a chance experiment has k outcomes, all equally likely, then each individual outcome has the probability 1/k and the probability of an event E is
- 16. Probability - Example Consider the experiment consisting of rolling two fair dice and observing the sum of the up faces. A sample space description is given by {(1, 1), (1, 2), … , (6, 6)} where the pair (1, 2) means 1 is the up face of the 1 st die and 2 is the up face of the 2 nd die. This sample space consists of 36 equally likely outcomes. Let E stand for the event that the sum is 6. Event E is given by E={(1, 5), (2, 4), (3, 3), (4, 2), (5, 1)}. The event consists of 5 outcomes, so
- 17. Probability - Empirical Approach Consider the chance experiment of rolling a “fair” die. We would like to investigate the probability of getting a “1” for the up face of the die. The die was rolled and after each roll the up face was recorded and then the proportion of times that a 1 turned up was calculated and plotted.
- 18. Probability - Empirical Approach The process was simulated again and this time the result were similar. Notice that the proportion of 1’s seems to stabilize and in the long run gets closer to the “theoretical” value of 1/6.
- 19. Probability - Empirical Approach In many “real-life” processes and chance experiments, the probability of a certain outcome or event is unknown, but never the less this probability can be estimated reasonably well from observation. The justification if the Law of Large Numbers. Law of Large Numbers : As the number of repetitions of a chance experiment increases, the chance that the relative frequency of occurrence for an event will differ from the true probability of the event by more than any very small number approaches zero.
- 20. Relative Frequency Approach The probability of an event E , denoted by P(E) , is defined to be the value approached by the relative frequency of occurrence of E in a very long series of trials of a chance experiment. Thus, if the number of trials is quite large,
- 21. Methods for Determining Probability The classical approach: Appropriate for experiments that can be described with equally likely outcomes. The subjective approach: Probabilities represent an individual’s judgment based on facts combined with personal evaluation of other information. The relative frequency approach: An estimate is based on an accumulation of experimental results. This estimate, usually derived empirically, presumes a replicable chance experiment.
- 22. Basic Properties of Probability For any event E, 0 P(E) 1. If S is the sample space for an experiment, P(S)=1. If two events E and F are disjoint, then P(E or F) = P(E) + P(F). For any event E, P(E) + P(not E) = 1 so, P(not E) = 1 – P(E) and P(E) = 1 – P(not E).
- 23. Equally Likely Outcomes Consider an experiment that can result in any one of N possible outcomes. Denote the corresponding simple events by O 1 , O 2 ,… O n . If these simple events are equally likely to occur, then
- 24. Example Consider the experiment consisting of randomly picking a card from an ordinary deck of playing cards (52 card deck). Let A stand for the event that the card chosen is a King.
- 25. Example Consider the experiment consisting of rolling two fair dice and observing the sum of the up faces. Let E stand for the event that the sum is 7. The sample space is given by S ={(1 ,1), (1, 2), … , (6, 6)} and consists of 36 equally likely outcomes. The event E is given by E ={(1, 6), (2, 5), (3, 4), (4, 3), (5, 2), (6, 1)} and consists of 6 outcomes, so
- 26. Example Consider the experiment consisting of rolling two fair dice and observing the sum of the up faces. Let F stand for the event that the sum is 11. The sample space is given by S ={(1 ,1), (1, 2), … , (6, 6)} and consists of 36 equally likely outcomes. The event F is given by F ={(5, 6), (6, 5)} and consists of 2 outcomes, so
- 27. Addition Rule for Disjoint Events Let E and F be two disjoint events.
- 28. Example Consider the experiment consisting of rolling two fair dice and observing the sum of the up faces. Let E stand for the event that the sum is 7 and F stand for the event that the sum is 11.
- 29. A Leading Example A study 1 was performed to look at the relationship between motion sickness and seat position in a bus. The following table summarizes the data. 1 “Motion Sickness in Public Road Transport: The Effect of Driver, Route and Vehicle” ( Ergonomics (1999): 1646 – 1664).
- 30. A Leading Example Let’s use the symbols N, N C , F, M, B to stand for the events Nausea, No Nausea, Front, Middle and Back respectively.
- 31. A Leading Example Computing the probability that an individual in the study gets nausea we have
- 32. A Leading Example Other probabilities are easily calculated by dividing the numbers in the cells by 3256 to get P(F) P(N and F) P(M and N C ) P(N)
- 33. A Leading Example The event “a person got nausea given he/she sat in the front seat” is an example of what is called a conditional probability. Of the 928 people who sat in the front, 58 got nausea so the probability that “a person got nausea given he/she sat in the front seat is
- 34. Conditional Probability If we want to see if nausea is related to seat position we might want to calculate the probability that “a person got nausea given he/she sat in the front seat.” We usually indicate such a conditional probability with the notation P(N | F). P(N | F) stands for the “probability of N given F.
- 35. Conditional Probability Let E and F be two events with P(F) > 0. The conditional probability of the event E given that the event F has occurred , denoted by P(E|F) , is
- 36. Example A survey of job satisfaction 2 of teachers was taken, giving the following results 2 “Psychology of the Scientist: Work Related Attitudes of U.S. Scientists” ( Psychological Reports (1991): 443 – 450).
- 37. Example If all the cells are divided by the total number surveyed, 778, the resulting table is a table of empirically derived probabilities.
- 38. Example For convenience, let C stand for the event that the teacher teaches college, S stand for the teacher being satisfied and so on. Let’s look at some probabilities and what they mean. is the proportion of teachers who are college teachers. is the proportion of teachers who are satisfied with their job.
- 39. Example Restated: This is the proportion of satisfied that are college teachers. The proportion of teachers who are college teachers given they are satisfied is
- 40. Example Restated: This is the proportion of college teachers that are satisfied. The proportion of teachers who are satisfied given they are college teachers is
- 41. Independence Two events E and F are said to be independent if P(E|F)=P(E). If E and F are not independent, they are said to be dependent events. If P(E|F) = P(E), it is also true that P(F|E) = P(F) and vice versa.
- 42. Example
- 43. Multiplication Rule for Independent Events
- 44. Example Consider the person who purchases from two different manufacturers a TV and a VCR. Suppose we define the events A and B by A = event the TV doesn’t work properly B = event the VCR doesn’t work properly 3 This assumption seems to be a reasonable assumption since the manufacturers are different. Suppose P(A) = 0.01 and P(B) = 0.02.
- 45. Example Consider the teacher satisfaction survey
- 46. Sampling Schemes Sampling is with replacement if, once selected, an individual or object is put back into the population before the next selection. Sampling is without replacement if, once selected, an individual or object is not returned to the population prior to subsequent selections.
- 47. Example Suppose we are going to select three cards from an ordinary deck of cards. Consider the events: E 1 = event that the first card is a king E 2 = event that the second card is a king E 3 = event that the third card is a king.
- 48. Example – With Replacement If we select the first card and then place it back in the deck before we select the second, and so on, the sampling will be with replacement.
- 49. Example – Without Replacement If we select the cards in the usual manner without replacing them in the deck, the sampling will be without replacement.
- 50. A Practical Example Suppose a jury pool in a city contains 12000 potential jurors and 3000 of them are black women. Consider the events E 1 = event that the first juror selected is a black woman E 2 = event that the second juror selected is a black woman E 3 = event that the third juror selected is a black woman E 4 = event that the forth juror selected is a black woman
- 51. A Practical Example Clearly the sampling will be without replacement so
- 52. A Practical Example - continued If we “treat” the Events E 1 , E 2 , E 3 and E 4 as being with replacement (independent) we would get
- 53. Notice the result calculate by sampling without replacement is 0.003900 and the result calculated by sampling with replacement is 0.003906. These results are substantially the same. Clearly when the number of items is large and the number in the sample is small, the two methods give essentially the same result. A Practical Example - continued
- 54. An Observation If a random sample of size n is taken from a population of size N, the theoretical probabilities of successive selections calculated on the basis of sampling with replacement and on the basis of sample without replacement differ by insignificant amounts under suitable circumstances. Typically independence is assumed for the purposes of calculating probabilities when the sample size n is less than 5% of the population size N.
- 55. General Addition Rule for Two Events
- 56. Example Consider the teacher satisfaction survey
- 58. Example 18% of all employees in a large company are secretaries and furthermore, 35% of the secretaries are male . If an employee from this company is randomly selected, what is the probability the employee will be a secretary and also male. Let E stand for the event the employee is male. Let F stand for the event the employee is a secretary. The question can be answered by noting that P(F) = 0.18 and P(E|F) = 0.35 so
- 59. Bayes Rule
- 60. Example A company that makes radios, uses three different subcontractors (A, B and C) to supply on switches used in assembling a radio. 50% of the switches come from company A, 35% of the switches come from company B and 15% of the switches come from company C. Furthermore, it is known that 1% of the switches that company A supplies are defective, 2% of the switches that company B supplies are defective and 5% of the switches that company C supplies are defective. If a radio from this company was inspected and failed the inspection because of a defective on switch, what are the probabilities that that switch came from each of the suppliers.
- 61. Example - continued Define the events S 1 = event that the on switch came from subcontractor A S 2 = event that the on switch came from subcontractor B S 3 = event that the on switch came from subcontractor C D = event the on switch was defective From the problem statement we have P(S 1 ) = 0.5, P(S 2 ) = 0.35 , P(S 3 ) = 0.15 P(D|S 1 ) =0.01, P(D|S 2 ) =0.02, P(D|S 3 ) =0.05
- 63. Example - continued These calculations show that 25.6% of the on defective switches are supplied by subcontractor A, 35.9% of the defective on switches are supplied by subcontractor B and 38.5% of the defective on switches are supplied by subcontractor C. Even though subcontractor C supplies only a small proportion (15%) of the switches, it supplies a reasonably large proportion of the defective switches (38.9%).