STATISTICS FOR MANAGEMENT-II LEARNING OBJECTIVES: To take sample from an entire population and use it to describe the population. Introduction to sampling and types of sampling. Introduction to sampling distribution and its concepts. To understand the trade-offs between the cost of taking larger samples and the additional accuracy this gives to decision made from them. INSTRUCTOR’S NAME: 1 CHAPTER – 6 Parameters: The observations taken from population used to calculate mean, median and standard deviation etc. are called parameters. For example, if we calculate mean from population, it is called population mean and is denoted by µ. Statistics: The observations taken from sample used to calculate mean, median and standard deviation etc .are called statistics. For example, if we calculate mean from sample, it is called sample mean and is denoted by x . There are two types of population, 1-Finite Population: Population having limited size is called finite population. 2- Infinite Population: Population, in which it is theoretically impossible to observe all items. There are two types of sampling, 1-Non-random or judgment sampling: It is process in which personal judgment determine which units of population are selected for sample. 2- Random or Probability Sampling: It is a process in which each unit of population has a known (non zero) probability of its being included in sample. There are four methods for random sampling. iiiiiiiv- Simple Random Sampling. Systematic Sampling. Stratified Sampling. Cluster Sampling. 2 i-Simple Random Sampling: It is method of selecting samples in which each possible sample has an equal probability of being picked and each unit of population has an equal chance of being included in the sample. ii- Systematic Sampling: It is method of selecting samples from population at uniform intervals that is measured in time, order or space. Iii-Stratified Sampling: In this method we divide the population in to groups, called STRATA. Then (a) We select sample from each stratum according to proportion of that stratum in population. (b) We draw an equal number of samples from each stratum and give weight to result according to stratum’s proportion of total population. We use stratified sampling when there is small variation in each group but wide variation between groups (i.e. group to group) iv- Cluster Sampling: In this method we divide the population into groups, called CLUSTERS, and then select a random sample of these cluster We use cluster sampling when there is wide variation within each group but small variation between groups (i.e. group to group). EXERCISE: (EX. Sc 6-1 pg 303) List the number of elements selected for the data mention in the question, based on the random digits table. Solution: 0892 1652 2963 2913 3181 9348 4959 7695 7712 8136 9659 2526 6988 1781 2204 4339 6299 3397 7652 8559 3 EXERCISE: (EX. 6-11 pg 303) A population is made up of groups that have wide variation within each group but little variation from group to group. The appropriate type of sampling for this population is, (a) Stratified. (b) Systematic. (c) Cluster. (d) Judgment. Solution: The cluster sampling is appropriate type of sampling for this population. H.W: Do EX.6-10 pg 303. Sampling with replacement: A method in which sampled items are returned to the population after being picked. In this method, each item of population can appear in the sample more than once. Sampling Without Replacement: A method in which sampled items are not returned to the population after being picked. In this method no item of population can appear in sample more than once. Sampling Distribution: A probability distribution of all values of a statistic (mean, median and standard deviation etc.) is called sampling distribution. There are many kinds of sampling distributions. Some are as normal distribution, t-distribution and F-distribution etc. 4 Sampling Distribution of mean: A probability distribution of all the means of the samples (taken from population) is called sampling distribution of mean. This distribution has its own mean µ and standard deviation σ . Standard Error: Standard deviation of sampling distribution of statistic (mean, median and standard deviation etc.) is known as standard error of that statistic (mean, median and standard deviation etc.). A sampling distribution of mean that has small standard error is better estimator of population than sampling distribution mean that has larger standard error. Relationship between Sample Size and Standard Error: As sample size increases the standard error decreases. As the standard error decreases the value of any sample mean will probably be closer to the value of population. Sampling from normal population: When population is normally distributed then, = σx̄ = µ 𝜎 Here σx̄ is standard error of √𝑛 mean from infinite population and we sample with replacement. σx̄ = 𝜎 √𝑛 𝑁−𝑛 × √𝑁−1 Here σx̄ is standard error of mean from finite population and we sample without replacement Here √ 𝑁−𝑛 𝑁−1 is finite population multiplier. If sampling fraction 𝑛 𝑁 is less than 0.05 then we will not use this multiplier. 5 Sampling from non normal population: Relationship between shape of sampling distribution of mean and shape of population distribution is called the central limit theorem. It states as, The sampling distribution of mean approaches to normal distribution, as the sample size increases, regardless of the shape of population distribution . Standard Normal Random Variable: Z= x −𝜇 σx̄ Here Z is standard normal random variable. With this we can determine the probability that the sample mean will lie between the given limits. EXERCISE: (EX. Sc 6-3 pg 311) Is the conclusion by quality control manager, with the information given in question, correct? Solution: The conclusion is not correct as the mean of sample does not equal the population mean because of sampling error. EXERCISE: (EX. 6-20 pg 311) The term error, in standard error of the mean, refers to what type of error? Solution: The term error, in standard error of the mean refers to sampling error. EXERCISE: (EX. 6-22 pg 311) Solution: In general, over estimating the mean is neither better nor worse than under estimating. In this case underestimate ($ 3.0) is closer to true mean ($ 3.14) than overestimate ($ 3.5).So first sample is better one. 6 EXERCISE: (EX. Sc 6-5 pg 321) Solution: 𝜇 = 98.6 𝜎 = 17.2 n = 25 (a) σx̄ = = Z = For √𝑛 17.2 √25 = 17.2 5 = 3.44 x −𝜇 σx̄ x = 92 Z = For 𝜎 x 92−98.6 3.44 = −6.6 3.44 = -1.92 = 102 Z = P(92 ≤ x 102−98.6 3.44 = 3.4 3.44 = 0.99 ≤ 102) = P ( -1.92 ≤ Z ≤ 0.99) = P (-1.92 ≤ Z ≤ 0) + P (0 ≤ Z ≤ 0.99) = 0.4726 + O.3389 = 0.8115 7 EXERCISE: (EX. Sc 6-6 pg 321) Solution: 𝜇 = 112 𝜎 = 56 n = 50 (a) σx̄ = = Z = For 𝜎 √𝑛 56 √50 56 = 7.07 = 7.92 x −𝜇 σx̄ x = 100 Z = P ( x < 100) 100−112 7.92 = −12 7.92 = -1.52 = P ( Z ≤ -1.52) = P (-∞ ≤ Z ≤ 0) - P (-1.52 ≤ Z ≤ 0) = 0.5 – O.4357 = 0.0643 8 EXERCISE: (EX. 6-28 pg 321) Solution: 𝜇 = 18 𝜎 = 4.8 n = 19 (a) σx̄ = = Z = For x = 16 Z = For x 𝜎 √𝑛 4.8 = √19 4.8 = 1.10 4.36 x −𝜇 σx̄ 16−18 1.10 = −2 = -1.82 1.10 = 20 Z = P(16 ≤ x 20−18 1.10 = 2 1.10 = 1.82 ≤ 20) = P ( -1.82 ≤ Z ≤ 1.82) = P (-1.82 ≤ Z ≤ 0) +P (0 ≤ Z ≤ 1.82) = 0.4656 + O.4656 = 0.9312 H.W: Do EX. Sc 6-5 (b) Pg 321 EX. Sc 6-6 (b) Pg 321 EX. 6-28 (c) Pg 321 9 EXERCISE: (EX. Sc 6-7 pg 327) Solution: (𝑎) 𝜇 = 105 𝜎 = 17 n = 64 N = 125 σx̄ = = 𝜎 √𝑛 17 √64 = 17 8 × √ 𝑁−𝑛 𝑁−1 125−64 × √ 125−1 61 × √124 = 2.125 X √0.492 = 2.125 X 0.70 = 1.49 (b) 𝜇 = 105 𝜎 = 17 n = 64 N = 125 σx̄ = 1.49 ( already calculated) 10 Z = For For x −𝜇 σx̄ x = 107.5 x Z = 107.5−105 1.49 = Z = 109−105 1.49 = 2.5 1.49 = 1.68 = 109 P(107.5 ≤ x 4 1.49 = 2.68 ≤ 109) = P ( 1.68 ≤ Z ≤ 2.68) = P (0 ≤ Z ≤ 2.68) --P (0 ≤ Z ≤ 1.68) = 0.4963 - O.4535 = 0.0428 EXERCISE: (EX. 6-40 pg 327) Solution: (𝑎) 𝜇 = 364 𝜎 2 = 18 𝜎 = √18 n = 32 N = 75 σx̄ = 𝜎 √𝑛 = 4.24 𝑁−𝑛 × √ 𝑁−1 11 4.24 = 75−32 × √ 75−1 √32 4.24 43 5.66 × √74 = 0.75 X √0.58 = = 0.75 X 0.76 = 0.57 𝜎 σx̄ = ( c) √𝑛 4.24 = √32 = 4.24 5.66 = 0.75 H.W: Do EX. 6-40 Pg 327 EXERCISE: (EX. 6-30 pg 321) Solution: e = |x − 𝜇 = 375 𝜎 = 48 𝜇 |=5 P (Z) = 95% = 0.95 = 0.95 2 = 0.4750 = 1.96 n = [ 𝑍𝑋 𝜎 ]2 |x −𝜇| 12 1.96𝑋48 2 ] 5 [ = = (18.82)2 = = [ 94.08 2 ] 5 = 354 EXERCISE: (EX. Sc 6-8 pg 327) Solution: (𝑏) e =|x − 𝜇 = unknown 𝜎 = 1.25 𝜇 | = 0.5 P (Z) = 98% = 0.98 = 0.98 2 = 0.4900 = 2.33 𝑍𝑋 𝜎 n = [ = [ 2.33𝑋1.25 2 ] 0.5 = [ 2.91 2 ] 0.5 ]2 |x −𝜇| = (5.83)2 = 34 H.W: If population has standard deviation $500.How many observation would be needed in order to be 95% certain that sample mean is within $100 of population mean? 13 OBJECTIVE SECTION Q-1 WRITE SHORT ANSWERS FOR THE FOLLOWING. 1-Write the name of the theorem which describes the relationship between the shape of sampling distribution of mean and shape of population distribution. Answer: Central Limit Theorem. 2- Define standard error. Answer: Standard deviation of sampling distribution of statistic (mean, median etc) is known as standard error of that statistic (mean, median etc) 3- Define parameters. Answer: The observations taken from population are used to calculate mean, median and standard deviation etc are called parameters. 4- Write the other name of random sampling. Answer: Probability sampling. Q-2 TICK THE CORRECT ONE. 1- Standard deviation of sampling distribution of statistic is called a) Cluster b) Standard error c) Population 2- As the sample size increases the standard error , a) Increases b) Decreases c) Remains same 14 3- The other name of non-random sampling is, a) Systematic sampling b) Judgment Sampling c) Non of these 4- Standard error of sampling distribution of mean is denoted by, Q-3 1- a) b) N c) σx̄ FILL IN THE BLANKS: In cluster sampling we divide the population in to groups called, . 2- A sampling distribution of that has estimator. 3- standard error is the best The observations taken from sample are called . 15 16