Basic Statistics
Chapter 5 Discrete Probability Distributions
5.1 Exercises
-
5.1.1 Consider litters of \(n=3\) rabbit kittens. Construct the sample space considering the births of females and males, using a tree diagram, and consider the events being born male and being born female as equally likely.
-
a) Letting \(X\) be the occurrence of females, construct the probability distribution of \(X\);
-
b) Calculate the probabilities of the following events using the constructed probability distribution:
-
i) birth of exactly two females.
-
ii) birth of at least one male.
-
iii) birth of at least two females.
-
iv) birth of at most one female.
-
-
c) Suppose you sample \(500\) litters of \(3\) kittens. In how many, on average, do you expect to find exactly \(1\) female?
-
-
5.1.2 Consider births of \(n = 4\) rabbit kittens of a given breed. In this breed there is a genetic disorder and the probability of being born female is \(5/8\). Letting \(X\) be the occurrence of females and using the binomial distribution, obtain:
-
5.1.3 On a slide it was verified that there were on average \(4\) bacteria/cm\(^2\). The slide was subdivided into \(600\) squares of \(1\) cm\(^2\). What is the adequate probabilistic model to model the occurrence of bacteria per cm\(^2\), assuming that the spatial distribution follows a random pattern? In how many of the \(600\) squares, on average, do you expect to find at most \(1\) bacterium? What is the probability of finding more than \(2\) bacteria per square centimeter? What is the probability of finding no bacteria in a square taken at random from these \(600\) squares?
-
5.1.4 A researcher in the area of animal science obtained a series of data from the last \(120\) years with the record of the number of a rare disease in horses of the locality where he worked. The data obtained were:
Number of diseases \((x)\) 0 1 2 3 4 5 Number of years \((F_i)\) 55 40 17 5 2 1 -
a) Estimate the average number of diseases/year;
-
b) Calculate for each value of the random variable \(X\) the associated probabilities. Assume that \(X\) has a Poisson distribution and that the sample mean is the estimator of the parameter \(\lambda \) of the Poisson distribution;
-
c) Calculate the expected frequency (in years) for each value of \(X\);
-
d) Compare the expected results with the observed ones. Based on this comparison, can you state that the Poisson distribution is adequate to explain the occurrence of this disease in the study region? Justify.
-
-
5.1.5 A researcher in the area of entomology decided that he would evaluate his experiment weekly and that the sample size in number of sampled plants would be a value between \(21\) and \(30\). The choice of a sample size \(x\) would be made according to the discrete uniform distribution. Knowing that the cost of each sampled plant is R$ 1.50:
-
a) Specify the probability function of \(X\), the random variable that represents the number of plants sampled in a week.
-
b) What is the expected weekly average cost for the sampling?
-
c) What is the variance of this cost?
-
d) Use Chebyshev’s theorem and find the limits that guarantee a minimum probability of \(95\%\) that the cost of any given week does not exceed these limits.
-
e) Construct \(F(x)\) for all values of the support of the random variable \(X\).
-
-
5.1.6 If \(X\) is a random variable with binomial distribution, determine the probability \(P(X < \mu _X - k\sigma _X)\) for the following situations:
-
a) \(n = 30\); \(p =\)0.1; and \(k = 1\) or \(k = 2\)
-
b) \(n = 30\); \(p =\)0.5; and \(k = 1\), \(k = 2\), or \(k = 3\)
-
c) \(n = 30\); \(p=\)0.9; and \(k = 1\) or \(k = 2\)
-
-
5.1.7 A seed producer claims that the germination power of his seeds is 0.95. A buyer will test a sample of \(100\) seeds and will only buy the lot if the number of germinated seeds in the sample is greater than \(90\). What is the probability that the lot is bought? Obtain the mean and variance and compare the probability value with that obtained by Chebyshev’s theorem.
-
5.1.8 To carry out a survey of the size of the fish population in a tank, a fish farmer decided to use the “capture” and “recapture” method. For this, a sample of size \(n_1 = 30\) was taken and all fish were marked, then returned to the tank. In a new sample of \(n_2 = 32\) fish, it was verified that \(x\) \(=\) \(2\) of them bore the mark. Estimate the population size using the hypergeometric distribution, seeking to maximize the probability of finding \(2\) marked fish in the second sampling. Suggestion: use different values of \(N\) and a spreadsheet to facilitate the calculations and verify the behavior of the probabilities. In this way it is possible to track the maximum probability value, which will correspond to an estimate of \(N\). Another approach can be made by maximizing the probability function, which is a discrete function and must be maximized in a special way. This maximum, if \(n_1n_2/x\) is not an integer, is \(\lceil n_1 n_2/x\rceil - 1\). If \(n_1n_2/x\) is an integer, the maximum can be \(n_1 n_2/x - 1\) or \(n_1 n_2/x\).
-
5.1.9 The average number of spores of a given fungus per cm\(^3\) of soil is equal to 0.05. What is the probability that a sample of \(100\) cm\(^3\) presents no spores of the fungus? What is the probability that at least \(3\) spores are found in the sample?
-
5.1.10 A milk producer performs artificial insemination in his herd. Males are discarded, since the interest is in milk production. How many inseminations are necessary to guarantee at least \(95\%\) probability that the first success occurs, that is, that the first female is born? What is the probability that the producer has to wait until the third insemination to have the first success (first female)?
-
5.1.11 From the statement of exercise number 5.1.10, answer:
-
a) What is the probability that the producer obtains the second success (female) in the third insemination performed?
-
b) How many inseminations will the producer have to wait to guarantee at least \(95\%\) chance of obtaining the third success?
-
c) What is the expected average value of inseminations to obtain \(3\) successes (females)?
-
-
5.1.12 Construct the probability distribution of the variable \(X\), number of females in litters of size \(8\), knowing that all litters have at least \(1\) female. What would be the probability function of the variable \(X\), considering that the litters have at least \(2\) females among the \(n\) \(=\) \(8\) kittens?
-
5.1.13 Set \(\alpha \) \(=\) \(1\) and \(\beta \) \(=\) \(1\) in the probability function of the beta-binomial distribution and demonstrate that this probability function specializes to the discrete uniform. Calculate the mean and variance of the obtained distribution, using for this the expressions presented for the mean and variance of the beta-binomial.
-
5.1.14 Show that the values of the expectation \(E(X)\) and of the variance \(\sigma ^2_X\) of the discrete uniform distribution, for \(x_i = 1\), \(2\), \(\cdots \), \(k\), are equal to \((k+1)/2\) and \((k^2-1)/12\). Obtain the mean and variance for the discrete uniform if the support is changed to \(x_i = 0\), \(1\), \(2\), \(\cdots \), \((k-1)\).
Additional data:
\(\seteqnumber{0}{5.}{0}\)\begin{align*} \sum _{i=1}^k x_i^2 =&(1^2+2^2+\cdots +k^2)=\dfrac {k^3}{3}+\dfrac {k^2}{2}+\dfrac {k}{6}, && x_i = 1, 2, \cdots , k,\\ \sum _{i=1}^k x_i^2 =&[0^2+1^2+\cdots +(k-1)^2]=\dfrac {k^3}{3}-\dfrac {k^2}{2}+\dfrac {k}{6}, && x_i = 0, 1, 2, \cdots , (k-1). \end{align*}