PROBABILITY DISTRIBUTIONS OBJECTIVES OF THIS UNIT Alt the end of this unit, you should be able to 4 Distinguish between frequency and probability distributions 4 Explain the Normal, Poisson and Binomial distributions & Compute the probabilities in Poisson and binomial probability distributions.Biostatistics Unit Three: Probability Distribution. For: L3 & MI classes Index of Unit Three 3.1. INTRODUCTION 3.2. MAIN CONTENTS 3.2.1. PROBABILITY DISTRIBUTION 3.2.2. THE NORMAL DISTRIBUTION 3.2.3. STANDARDIZING THE NORMAL CURVE 3.2.4. POISSON DISTRIBUTION 3.2.5. BINOMIAL DISTRIBUTION 3.3. TUTOR MARKED ASSIGNMENT 3.4. REFERENCES Prepared by: Dr. Zoubida Belli, Created b: Nigeria 1. Hiya S. Ndams Ahmadu Bello University, Zaria, EBBiostatistics Unit Three: Probability Distribution. For: L3 & M1 classes 3.1. INTRODUCTION Probability is a branch of mathematics which as a general concept can be defined as the chance of an event occurring. It is the basis of inferential statistics. This unit looks at three particular distributions, the Normal, the Poisson and the Binomial ~ all of which are important in sampling theory. 3.2. MAIN CONTENTS 3.2.1. PROBABILITY DISTRIBUTION A distribution is a scatter of related values, such as the assortment of weights in a group of cattle. A frequency distribution shows us how many ‘Probably dstibuton pill adel gl Lol aay JL wish Probability distribution is very similar because it Rea Rie el ome ga gpa times given values in a range of values occur. A shows us how probable given random variable values in a range of such values are. Example: if we toss two coins, we can obtain 0, 1 or 2 “heads”. If we prepare a table showing the probabilities of all the random variable values, we will have the probability distribution as shown below. ‘Number of heads? ‘Sequential event | Probal 0 a 05x05 025 1 ar 05x 05}.050 {TH 05x 03) 2 HH 05x05 =025 Total 00 Therefore, a probability distribution is simply a complete lis ing of all possible outcomes of an experiment, together with their probabi 3.2.2. THE NORMAL DISTRIBUTION This is the most important distribution in statistics. It is also known as the Gaussian distribution named after Gauss, a German astronomer who showed its use in statistics. The ¢ normal distribution is defined by just two statistics, the mean and the standard deviation (Gstat! 1,531). Normal distribution is concerned with results obtained by taking measurements Prepared by: Dr. Zoubida Belli, Created by: Dr. Iliya S. Ndams Ahmadu Bello University, NigeriaBiostatistics Unit Three: Probability Distribution. For: L3 & M1 classes on continuous random variable (ie the quantified value of a random event) like weight, yield etc, The normal distribution is a particular pattern of variation of numbers around the mean. It is symmetrical (hence we express the standard Normal dstbuton SLAM chal Sob ge Uple Upeasll ete gl LY te cae at cane In gy etal lol al numbers falls off equally away from the mean in aly Rye ange Ugly Regan deviation as #) and the frequency of individual both directions, It so happens that the curve given by this probability’s distribution approximates very closely to a Mathematical curve. This curve is called the Normal curve. Figure 3.1: Normal curve with standard deviations Yeaxis be toy He He X-axis Example: In terms of human height, progressively larger and smaller people than the average ‘occur symmetrically with decreasing frequency towards respectively giants or dwarfs. > Properties of a Normal Curve = It is a Unimodal symmetrical curve ~The mean, mode & median all coincide, thereby dividing the curve into two equal parts — Most items on the curve are clustered around the mean — No kurtosis (qh,4:) or skewness (1,3!) in the curve. Figure 3.2: Normal curve with Skewness Prepared by: Dr. Zoubida Belli, Created by: Dr. Iliya S. Ndams Ahmadu Bello University, NigeriaBiostatistics Unit Three: Probability Distribution. For: L3 & M1 classes Mean Median Median Median Mode i Meany | Mode ‘symmetrical Negative Distribution Skew Example: Let us take a very common example of house prices. Suppose we have house values ranging from $100k to $1,000,000 with the average being $500,000. If the peak of the distribution was left of the average value, portraying a positive skewness in the distribution. It would mean that many houses were being sold for less than the average value, Le. $500k. If the peak of the distributed data was right of the average value, that would mean a negative skew. This would mean that the houses were being sold for more than the average value. >What is Kurtosis? Kurtosi statistical measure used to describe the degree to which scores cluster in the tails or the peak of a frequency distribution. The peak is the tallest part of the distribution, and the tails are the ends of the distribution. There are three types of kurtosis: Mesokuni — Mesokurtic: Distributions that are moderate in Bap Inge pub pn ly PLAIN Jaa age breadth and curves with a medium peaked height. — Leptokurtic: More values in the 4) Leptokuitie General distribution tails and more values close * Forms of (0 Mesokurt Kurtosis to the mean (ie. sharply peaked with (Nonmal) heavy tails) 1 Patylurtte — Platykurtic: Fewer values in the tails and fewer values close to the mean (i.e. the curve has a flat peak and has more dispersed scores with lighter tails). Prepared by: Dr. Zoubida Belli, Created by: Dr. Iliya S. Ndams Ahmadu Bello University, Zaria, NigeriaBiostatistics Unit Three: Probability Distribution. For: L3 & M1 classes > Important Aspect or characteristics of the curve The important aspect of the curve is the area in relationship with probability, if perpendiculars are erected at a distance of 10 from the mean and in both directions, the area covered by these perpendiculars and the curve will be about 68.26% of the total area. (It means that 68.26% of all the frequencies are formed within [Lu cyt ave tua aul ipl Gae g eee (Gane Sie 01 hire Se cL GF gl Al 6625 Gy tone oS gl go Ly stag go 1 § 350 SLY Sa Ail, 68,26 curve is 1 (100%). (100%) 1 gaa Slee LaLa ay one standard deviation of the mean). The total probability encompassed by the area under the Figure 3.2: Normal curve (y) 040 035 > 5030 5 8025 2o20 B01 8 Eso 0.05 0.00 u-30 u-20 ko to ut20 ut30 Normal distribution is defined by: 1 oven Usually, we do not know the values of 1 (Population mean) and o (Population standard £-G-W7/20), where: — 00 200] = P [E =172 _ 200-172 — «L VI96 VI196 The proportion of French men are between 165 and 185 centimeters tall F(2) = 0.02275 65-172 X= 172 185-172 vise * Vise ~~ Vise 4. The tall would be the 9000-th is: PIL6S 0]=P > t 1 340 340 — F(-0.5423) = 0.7062 3.2.4. POISSON DISTRIBUTION A Poisson distribution is a discrete probability distribution that is useful when n is larger and pis small and when the independent variables occur over a period of time. It can be used when a density of items is distributed over a given area or [ qau> osSicuny Alsat Ghia ula las! pisisa os lae Sa pie S pte atoll volume, such as the number of plants growing per Maal lyse! ge Boar Bus acre(s\ss). It can also be used to discover whether organisms (a> <8) are randomly distributed. Example: in ecological studies, Poisson distribution is used to describe the spread of organisms like insects, trees, and snails’ etc. by the following: 1 Divide the large area into small squares of equal size. 2.Count the particular animal or plant species under study in each square Prepared by: Dr. Zoubida Belli, Created by: Dr. Iliya S. Ndams Ahmadu Bello University, NigeriaBiostatistics Unit Three: Probability Distribution. For: L3 & M1 classes 3.You can also randomly select a number of squares, if the area is two large. The probability of X occurrences in an interval of time, volume, area etc. for a variable where 4 (lambda) is the mean number of occurrences per unit (time, volume, area ...etc) is given by: ay PG, a) == where: x = 0,1,2,3 and e = constant, approximately equal to 2.7183 ... Example: In a study on the distribution of tree-roosting birds, if there are 200 birds randomly distributed on 500 trees, find the probability that a given tree contains exactly three birds. Solution: First, find the mean number A of birds on each tree 200 _ 500 That is 0.4 birds per tree. Therefore 2 = 0.4, while x = 3, We can then substitute in the formular above. 2.7183)-°4(0.4)? P0304) = £ yO)" = 0.0072 Thus, there i less than 21% probability that any given tree will contain exactly three birds. 3.2.5. BINOMIAL DISTRIBUTION a 04 ‘A binomial experiment is a probability experiment that satisfies the following four requirements: 1. Each trial can have only two outcomes or [yun pt Jus! hips win Qld wd outcomes that can be reduced to two | JS Mii cite cba a os baie SHhett Ryn Sg OAS ane me sll go Spat LAS Sigua! Jara! Gy oly cinaall Genes Ge ite success or failure. No two events can occur a se outcomes. i.e these outcomes can either be simultaneously. Prepared by: Dr. Zoubida Belli, Created by: Dr. Iliya S. Ndams Ahmadu Bello University, NigeriaBiostatistics Unit Three: Probability Distribution. For: L3 & M1 classes 2. There must be a fixed number of trails. 3. The outcomes of each trial must be independent of each other. 4, The probability of a success must remain the same for each trial. ‘A binomial distribution is a special probability distribution that describes the distribution of probabilities when there are only two possible outcomes for each trial of an experiment. Examples: 1. The answer to a multiple-choice question (even though there are four or five answer choices) can be classified as correct or incorrect. 2. When tossing a coin, you get either a head or a tail 3. In selecting individuals from human population, you select either a male or a female, a boy ora girl etc. The binomial probability formula is given by: n o PO) = Gry"? Where’ "= Numerical probability of a success = P(s) = Numerical probability of a failure = P(F) Number of trials The number of successes in n trials ‘Mathematical symbol called ‘factorial’. So, n! means multiple all the numbers in a count down from the total number in the sample. Example: 7! =7x6x5x4x3x2xland4!=4x3x2x1 Example: A survey on birds showed that one out of five fire finch was trapped, using mist net, ina given season. If 10 birds are selected at random, find the probability that 3 of the birds were trapped in the previous season. Solution: n=10, x-3, P= 1/5 and q-1/5 Substituting these values in the formula above, we have tora (0 —3)13! OQ P@)= Prepared by: Dr. Zoubida Belli, Created by: Dr. Iliya S. Ndams Ahmadu Bello University, NigeriaBiostatistics Unit Three: Probability Distribution. For: L3 & M1 classe The mean (p), variance (02) and standard deviation (a) of a variable that has the binomial distribution can be found by using the formular: pen.p;o=n.p.q; o=Vnp.q From our example above: The mean: p= 10* '/5 =2 The variance: «? = 10x? x4= 1.6 The Standard deviation, 0? = V1.6 = 1.265 Example: 40% of the snails breed by a certain farm are of the type reticulata, determine the probability that, out of 12 snails chosen at random, (a) 2 (b) at most 3 will be of the type reticulate Solution: 0.4 is the probability of chosen reticulata breed, and 0.6 is the probability of chosen a non-reticulata breed. P (2 reticulata breed of 12 snails) =(12)(0.4)?(0.6)° b. (b) P (at most 3 reticulata snails breed) = P(0 reticulata) + P(1 reticulata) + P(2 reticulata)+ P(3 reticulata) = (77) (0.4) (0.6)? + (77)(0.4)1(0.6)!*+ (17) (0.4)(0.6)1°+ (47) (0.4)3(0.6) Prepared by: Dr. Zoubida Belli, Created b: Nigeria 1. Hiya S. Ndams Ahmadu Bello University, Zaria, as |Biostatistics Unit Three: Probability Distribution. For: L3 & M1 classe 3.4. TUTOR MARKED ASSIGNMENT 1) What is a Normal Curve? 2) Outline any five characteristics of a normal curve, 3) Construct a probability distribution for three patients given a headache relief tablet. The probabilities for 0, 1, 2 or 3 success are 0.18, 0.52, 0.21 and 0.09, respectively. 4) Inacertain rabbit farm in the southern part of Nigeria 25% of the rabbit breed are of the type bauscat, determine the probability that, out of 3 rabbits chosen at random, (a) 1 (b) at most 4 will be of the type bauscat Prepared by: Dr. Zoubida Belli, Created b: Nigeria 1. Hiya S. Ndams Ahmadu Bello University, Zaria, as |Biostatistics Unit Three: Probability Distribution, For: L3 & M1 classes 3.6. REFERENCES 1) Bailey, N.TJ. (1994). Statistical Methods in Biology. Third Edition, Cambridge University Press. United Kingdom. 2) Bluman, A.G. (2004). Elementary Statistics. A Step by Step Approach. Fifth Edition. McGraw-Hill Companies Incorporated. London. 3) Daniel, W.W. (1995). Biostatistics: a foundation for Analysis in Health sciences. Sixth Edition. John Wiley and sons Incorporated. USA 4) Fowler, J.A. and Cohen, L. Statistics for Ornithologist. British Trust for Ornithology Guide 22. 5) Harper, W.M. (1991). Statistics. Sixth Edition. Pitman Publishing, Longman Group, United Kingdom 6) Hoel, P.G. (1976). Elementary Statistics. Fourth Edition. John Wiley and Sons Incorporated, NewYork. Pp 151-204. 7) Mukhtar, F.B. (2003). An Introduction to Biostati Nigeria. Pp 1-112 8) Sanders, D.H., Murph, AF. and Eng, RJ. (1980). Statistics: A Fresh Approach. ics. Samarib Publishers, Kano McGraw-Hill Kogakusha, Limited. Kosaido Printing Company Limited,Tokyo, Japan. Prepared by: Dr. Zoubida Belli, Created b: Nigeria 1. Hiya S. Ndams Ahmadu Bello University, Zaria, ElBiostatistics Unit Three: Probability Distribution. For: L3 & M1 classe NOTES: EXEREISE 01 (CONTINUOUS RANDOM VARIABLES): In a population with 40000 persons, 2276 have between 0.8 and 0.84 milligrams of bilirubin per deciliter of blood, and 11508 have more than 0.84. Assuming that the level of bilirubin in blood follows a normal distribution model. 1.Calculate the mean and the standard deviation. 2.How many persons have more than 1 mg of bilirubin per dl of blood? IERERSISETORVCONTINUSUSRANBEMIVARIABLED: It is known thatthe blood pressure of people in a population with 20000 persons follows a normal distribution model with mean 13 mm Hg and interquartile range 4 mm Hg. 1.How many persons have a blood pressure above 16 mm Hg? 2.How much have to decrease the blood pressure of a person with 16 mm Hg in order to be below the 40% of people with lowest blood pressure? iSRERGIsE/GS(GONTINUSUE!RANBOMIVARIABLED):-Let Z be a random variable following a standard normal distribution model. Calculate the following probabilities using the table of the distribution function: 1. P(Z < 1.24) 2. P(Z > -0.68) 3. P(—1.35 < Z < 0.44) : Let X be a random variable following a normal distribution model N(10,2). 1. Calculate P(X < 10). 2. Calculate P(8 < X < 14), 3. Calculate the interquartile range. 4, Calculate the third decile. EREHEISEIOS(DISERETE|RANSEWIVARIABUER): Let X be a discrete random variable with the following probability distribution: Prepared by: Dr. Zoubida Belli, Created b: Nigeria 1. Hiya S. Ndams Ahmadu Bello University, Zaria, ElBiostatistics Unit Three: Probability Distribution. For: L3 & M1 classes xl[4[s]eé[7|[8 Fe) | 0.18 | 0.35 | 0.10 | 0.25 | 0.15 1. Calculate and represent graphically the distribution function, 2. Calculate the following probabilities a. P(X < 7.5).b. P(X > 8).c. PAS X <65).4.P(S 1) = 0.1149 = 11.5 persons . ne EXERCISE 02; 2 PU < 75) = 088, P(X > 8) =0,PS 2565) = 08m 1. P(X > 16) = 0.1587 = 3174 persons. PG -0.68) = 0.7517. 2. P(X > 2) = 0.2639. 3. P(-1.35 < Z < 0.44) = 0.5815. apel 1 P(X < 10) = 05 Let. X bethe numberoffemsesna group of students, 2. P(@< X< 14) = 0.8186 P(X <2) =0.17% 3. IQR = 2.608 Let be the number of groups of students without males ina sample of 4. Ds = 8.9512. groups, P(Y > 0) = 0.2125. Prepared by: Dr. Zoubida Belli, Created by: Dr. Iliya S. Ndams Ahmadu Bello University, Nigeria