0% found this document useful (0 votes)
16 views18 pages

Stat For Econ. CH 3

Chapter Three discusses special probability distributions, focusing on discrete types such as binomial, hypergeometric, and Poisson distributions. It explains the characteristics, formulas, and examples for calculating probabilities, means, variances, and standard deviations for these distributions. Additionally, it introduces multinomial distributions and continuous probability distributions, emphasizing the importance of area under the curve for continuous variables.

Uploaded by

Lami Daba
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
0% found this document useful (0 votes)
16 views18 pages

Stat For Econ. CH 3

Chapter Three discusses special probability distributions, focusing on discrete types such as binomial, hypergeometric, and Poisson distributions. It explains the characteristics, formulas, and examples for calculating probabilities, means, variances, and standard deviations for these distributions. Additionally, it introduces multinomial distributions and continuous probability distributions, emphasizing the importance of area under the curve for continuous variables.

Uploaded by

Lami Daba
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 18
CHAPTER THREE SPECIAL PROBABILITY DISTRIBUTIONS A) Types of discrete probability distributions 1. Binomial probability distribution This is one of the most widely used discrete probability distributions. A random variable with only ‘two outcomes ( 0 & 1, success & failure) is simply known as Bernoulli random variable after a Swiss mathematician Jacob Bernoulli during 17" century. ‘Trials having the following characteristics are called Bernoulli tria ‘We are interested in the number of successes in a fixed number of trials, b. Each trial has only two possible outcomes, success or failure, ¢. The probability of success is constant from trial to trial, 4. All trials are independent. Many processes produce outcomes that are classified into one of two distinct categories. For instance, -each item from a production process is frequently labelled as either good or defective, -class room attendance is recorded as either present or absent and the like. An experiment that results in outcomes of this type is called a Bernoulli process. A Bernoulli process is a random experiment that results in one of two mutually exclusive outcomes for each trials. Let the Greek letter ‘ze’ represents the probability of the outcome success and ‘1 — 7° represents the probability of the outcome failure for each trial of a Bernoulli proc s, then the probability of a specific sequence (arrangement) of x-successes in ‘n-x” failures out of trials is a*(1— 1)" where 1 = probability of success; 1 — 1 = probability of failure The binomial probability distribution expresses the probability of ‘x’ successes out of n-identical trials of a Bernoulli process for fixed values of n and 1. That is, P(X =x/n,7) lags] lrla- m= aw 47 Examplel; a national advertising agency estimates that only 50% of all new products introduced in Ethiopia succeed, i.e, become profitable enough to continue being produced. What is the probability that exactly 4 out of 6 new products succeed? Given: 1 = 0.50, n=6, x=4, then Solution: P(X = x/ n,m) =| lw" — my = leare=wl lo-s*a 0.5 rea =4/ 6.0.8) - [S| egy = 15 (3) (= = o.2see This is the probability that exactly 4 out of 6 new products succeed. Example2: An owner of a Mantain room has 15 rooms available for rent and they are rented independently. The probability that any one of them will be rented for a single night is 0.8. Compute the probability that 12 rooms will be rented in a single night Given; m= 0.80, n=15, x=12,1-2=0.2 Solution: from P(X = x/n,7) = | leyca = mr laa aIin=2)! |aayazml lo 8" ]0.2'" = BDO"@ Exercise: Consider the case of 4 trials and denote success by ‘S” and failure by ‘F’, then 1) A single success can occur in any one of the following 4 ways: (SFFF, FSFF,, EFSF, FFFS) tog gg P(I-P)® PU-P) ~—-P(-Py’—P(I-P)’, therefore, F(L4,P) = @) P(1— P)3, where P = probability of success and 1-P = probability of failure. 2) Two successes can occur in six distinct ways: 48, (SSEF, _ SFSF, SFES, FSFS, ESSE, PtHP) The probability of each success = (*) pre py? =) 1— mw? £(2,4,P) = (3) P (a-P)? or £24.) = (3) (-7) ‘The general form of probability distribution of a binomial random variable is given by: SGenw) =]CD = wy for x= 0,1, 25. 0 Otherwise, Where, 1-1 = probability of failure and 1 = probabilty of success Mean, Variance and Standard Deviation of a Binos Di Binomial probability distribution is a diserete probability distribution, And hence, the technique used to compute mean and standard deviation for a discrete randam variable is similar with the method used to compute pt and 6 for a binomial distribution. A binomial distribution has an expected value or long run average, which is denoted by where. ~ aP. Where n ~ sample size and P is probability of success. Like for other discrete variables, the variance of the binomial distribution is calculated as es Ya- w2. P(X) = nm(1 — 2) and the standard deviation of the binomial distribution is b= fe = ue w?.P(X) = van —m) Examplel: Suppose the production process of a certain company produces component parts for machineries. Information shows that when the company is operating at standard quality, 10% of the component parts will be defective. Suppose the quality controller plan to dismantle four component parts each week selected randomly to check for defect. Determine: 49 a) The expected number of defective components bb) The value of the standard deviation of the defective components Given: m = 0.10, thus, 1-7 =1—0.10 = 0.90, and n=4 Solution: a) x =n = 4X0.10= 0.40, b) 6? = na(1— 72) = 4X0.10X0,90 = 0.36 Then, ine — =) = V4X0.10X0.90 = 036 = 0.60 Example2: the probability is 0.2 that an item produced by a factory is defective. If 40000 items are made by the factory, compute the value of the standard deviation of the defective items. Solution: with n = 4000, = 0.20 and 1 m = 0.80, then 5? = nn (1 — x) = 40000(0.20)(0,80) = 6400 and 5 = Jnn(1—m) = ¥6400 = 80 Thus, the standard deviation of the defective items is 80. 2, HYPERGEOMETRIC PROBABILITY DISTRIBUTION ‘Some situations have outcomes that may be classified into one of two categories-such as success and failure, but, the probability of each outcome does not remain constant from trial to trial as in the binomial probability distribution. ‘Suppose that ‘n’ objects are to be chosen from a total of “N’ objects cons ting of success and failure and we are interested in the probability of getting *x’ successes and ‘n-x’ failures. It follows that the probability of ‘x’ successes in ‘n’ trials for fixed values of n, N and k is given by (Ky NK POC = x/nn,b) = BeDUtACID _ (ID <—eperGeometre. probability Distribution formula Examplel: A repair and service center for household appliances has a carton of 10 replacement tubes for use in certain models of microwave ovens. An oven, now being repaired, requires one of these replacement tubes and the repair person randomly chooses a tube from the carton. If 2 of 10 tubes are defective and 4 of 10 tubes are installed in microwave ovens, what is the probability that s defective? exactly oni Given: N = 10, k = 2 (defective/success), x = I(success from total success), n=4 50 -Oa 4 Solution: P(X = x/n,N,k) = Example2: Among an Ambulance service's 16 Ambulances, 5 emit excessive amounts of pollutants. If eight of the Ambulances are randomly chosen (picked) for inspection, what is the probability that this sample will include atleast three of the Ambulances that emit excessive amounts of pollutants? Solution: The probability we must find is P(3) + P(4) +P(5), thus, we have N = 16; n= 8; k = 5, then using the above formula: 10x462 i. P(X =3/8,16,5) WOxAe? — 9,359 ae sx230 ii, P(X = 4/8,16,5) = = 4339 0,128 P(X = 5/8,16,5 ai 44165 _ 0,013 (x = 5/8,16,5) = SX 8-8) — MAES _ 9 Thus, the probability that this sample will include atleast three of the Ambulances that emit ve amounts of pollutants is P(3) + P(4) +P(5) = 0.359 + 0.128 + 0.013 = 0.5 ex: er suppose an electronic component factory ships components in lots of 100 of which a few 10 of them are defective. A quality controller draws a sample of 5 to test. What is the probability that all of the five are defective? Given: N= 100; k= 10; n=5;x=2, then, Soh CO _ ELT _ AD _ sz06600 av TO8. 100, 75287520 @ ey ey P(X = x/n,N,k) = 0.0702 sa Exercise2: Among the 20 solar collectors on display at a trade show, 12 are flat-plate collectors and the others are concentrating collectors. If'a person visiting the show randomly selects six of the that three of them will be flat-plate collectors? solar collectors to check out, what is the probability Solution: from N = 20; n= IBUTION 3. POISSON PROBABIL A variable which can take only one discrete value in an interval of time, howsoever small, is known as Poisson variable. A Poisson process produces a discrete number of occurrences in a continuous interval. The interval may refer to any continuous measure of time (t), distance, area, et. Any experiment meeting the following conditions is the number of occurrences (x) within the interval. 1, The number of occurrences in one interval is independent of the number of occurrences in another interval. 2. The expected number of occurrences in an interval is proportional to the size of the interval. 3. A very small interval may be identified so that no more than one occurrence is possible in any interval of this size. 4, The probability of two or more occurrences in a subinterval is small enough to be ignored. 5. It must be possible to divide the time interval of interest into many subintervals. ‘The Poi ‘on probability distribution gives, for any Poisson process, with parameter * At’, the probability of *x’ occurrences in an interval of size ‘t’. This is, aor(e™ P(x = x/At) , where 2 — represents the intensity of Poisson process or the expected number of occurrences in a specified interval (expected number of successes or average number of jon of interest and e = 2.71828. successes); tis the proportion of the specified interval for the que: 52 the Examplel; patients arrive randomly for treatment at the health clinic of a University Hospital. expected number of patients arriving each hour is 4, What is the probability that exactly two patients arrive in 1 hour? Given: 2=4; t=1; x=2; then, M=4, Solution; P(x-x/at) = CVE - we, Ss age = 0.1465, => Gnas This is the probability that exactly two patients arrive in I hour. The Poisson probability distribution has a single parameter, the expected value At. NB: for Poisson probability distribution the expected value and variance are equal, i.e. E(x) = 6,7 =At The difficulty of evaluating the binomial probability distribution increases with sample size. To be more specific when ‘n’ becomes larger and 7 becomes proportionally smaller and where ‘nm (sample size times success) remains unchanged, then, the binomial distribution approaches the Poisson distribution, Generally, Poisson distribution is a good approximation of the binomial distribution provided that n > 20 and £0.50. To use the approximation, set the expected number of successes “nn” equal to the expected number of occurrences “At” and evaluate the probability P(x =x/at) = gore”) = 4. MUL! NOMIAL PROBAB! (TY DISTRIBU, An important generalization of the binomial distribution arises when there are more than two possible outcomes for each trial, the probabilities of the various outcomes remain the same for each trial and the trials are all independent. This is the case, for example, when we repeatedly roll a die where each trial has six possible outcomes. If there are ‘k’ possible outcomes for each trial and their probabilities 53 are Pi, Po, Ps, « , and Pj, it can be shown that the probability of X; outcomes of the first kind, X> ‘outcomes of the second kind, ..., and X, outcomes of the k" kind in ‘n’ trials is given by: tw) (PE. P2...P&*), this is multinomial distribution formula, Examplel: For a car being tested at a state inspection station, the probability that it will pass on the first try is 0.70, the probability that it will pass on the second try is 0.20 and the probability that it will pass on the third try is 0.10. what is the probability that among 10 cars being tested, six will pass on the first try, three will pass on the second try and one will pass on the third try? Given: n= 10; x1 =6; x2=3; xs=1 and P)=0.70; P»= 0.20; Ps=0.10 Solution: (<=) (Pj. Pi#P3*) = (FE) (0.7°.0.280.1%), 20° (5) = #40Gsie0 Go) Go) = 0.079060 ~ 0.079 (sz) GO". B) CONTINUOUS PROBABILITY DISTRIBUTIONS, ‘There are many CPD, such as, Uniform distribution, normal distribution, the t-distribution, the chi- square (7) distribution, exponential distribution, and F-distribution. The probability distribution for continuous random variable is defined by mathematical function or equations. The probability for such variables is defined in terms of the area under the curve. The probability distribution for continuous random variable ‘X” is a function f(x) for which: i. f()20, 0 Wx ii, the (otal area under the curve f(x) equals 1 Examplel: the postal service of country ‘A’ provides air mail service to foreign countries for packages weighing no more than 4 pounds. Heavier items may be mailed by parcel post. Personnel at one post office constructed a probability distribution for a continuous random variable *X’, the weight (in pounds) of packages airmailed to foreign countries. This is given by: 54 , for 0 0, Vx and the total area under the curve f(x) equals 1 Also it is possible to find probabilities between a) | and 2 inclusive, and b) 2 and 4 inclusive. That is, a) P(1 Hy wa Figure 8.9 Normals with different means, same variance: Figure 8.10 Normals with different variane ‘The normal probability distribution for a continuous distributions normal random variable ‘x’ is a given by: f(x/8) = gee sfor 05x50, 5>0,-0 Sys, mand e are mathematical constants where m=3.14159... or = 22/7 and e= 2.71828. The parameters which define a normal distribution are referred to as the location and shape parameters, ie., the parameter mean (j1) determines the location of a normal distribution along a number line; the parameter standard deviation (8) determines the relative width and height of the distribution. Note: the width of the distribution decreases and the height increases as the standard deviation decreases. That is, width and height are having indirect relationship while width and standard 58 deviation have direct relationship. The probability that a normally distributed random variable ‘x’ assumes values between limits a & b; P(a > P(asx0.5) = 0.5 — 0.3085 = 01915 Orsimply P(OSZ<0.5) = Zs =0.1915 and P(OSZE1) =0.5 - P(Z>1,00) = 0.5 = 0.158" Therefore, P(-0.51.5)] = 0.5 +{0.5 - 0.0668] = = 0.5 + 0.4332 = 0.9332 Therefore, the probability that the volume is less than 345ml is 0.9332. ©) Given p= 330ml, 8 = 10ml, Required: P(x > 338) 330-330 70 xu Soluti 0.80 Then, P(x > 338) = P(Z > 0.80) =0.5- P(O0.8)] 0.5 -[0.5 - 0.2119] = 0.5000 - 0.2881 = 0.2119 ‘Therefore, the probability that the volume is greater than 338ml is 0.2119. Exercise: Annual salary for the professional employ: in the planning department of a company is on average (1) $26000 with standard deviation (3;) $2000. Annual salaries for cleric: employees in the same department average (2) $15000 with standard deviation (62) of $1000, 62 a) One of the professional employees has an annual salary of $29000 (x:) b) A clerical employee eams $17500 (x2). The manager of the planning department sets the salaries within each group of employees according to the individual's relative performance in that group. On this basis both A & B are above average performance. Is one a more outstanding performer than the other? Reading assignment: Normal Approximation to Binomial Distribution 63 CHAPTER FOUR 4.0 JOINT AND CONDITIONAL PROBABILITY DISTRIBUTIONS Introduction We have until now been concemed with probability distributions involving one variable only. ‘There are many types of experiments, however, in which we shall need to deal with more than one variable at a time. For instance, we might wish to select a sample of N individuals with a view to ascertaining the income, age, and educational background of each person in the sample. In this case each of the three characteristics measured is a random variable. In our introductory discussion of probability, we already referred to probabilities assigned to the joint occurrence of ‘vo or more events as joint probabilities. When joint probabilities are assigned to two or more random variables, the resulting relationship is known as a joint probability distribution, Joint probability distribution of two random variables is known as bivariate distributions, When more than two random variables are being considered, the distribution is said to be multivariate distribution. In this chapter, we extend some of the ideas presented under chapter on and two to the concept of a joint probability distribution two random variables. It is subdivided in to four sections. The first two sections focus on the joint and marginal distributions. Define the conditions to be considered as a bivariate probability distribution of two random variables and also shows how to derive the ‘marginal distribution from the given bivariate probability distributions. The third section discusses the conditional probability and the statistical independence. This section shows how to compute conditional probabilities and how to check statistical independence between two random variables using the joint, marginal and conditional probabilities. The last section focuses on the measures that used to summarize the joint probability distribution. Under this section the ‘most important measures of expectation, covariance, and correlation will be discussed, After reading this section you will be able to: Y Define and interpreted the joint probability distributions for both discrete and continuous random variables 64

You might also like