STATISTICS A Complete Solution Guide CM UM ea sey # For Homework, Test Preparation, Exams @ For use with introductory and advanced texts ETC Mero ett 9 # Includes every type of problem that may be assigned by your instructor or given on a test @ Each problem worked out in step-by-step detail, Ulett Me elm CoM aCe (Lecd cute MC Meteo) (Tot am tT St LC ol Meee) mM aCe LTT) solutions to problems Premier12 Research & Education AssociationTHE STATISTICS PROBLEM SOLVER? Year 2004 Printing Copyright © 2002, 2000, 1998, 1996, 1994, 1978 by Research & Education Association. All rights reserved. No part of this book may be reproduced in any form without permission of the publisher. Printed in the United States of America Library of Congress Control Number 2001091423 International Standard Book Number 0-87891-515-X PROBLEM SOLVER® and REA® are registered trademarks of Research & Education Association, Piscataway, New Jersey 08854.CONTENTS Cc No. Page No. UNITS CONVERSION FACTORS —_ xiii 1 Graphs and Frequency Distributions 1 Measures of Central Tendency 9 Measures of Dispersion _16 Measures of Grouped Data 30 Moments, Skewness, Kurtosis 51 2 ELEMENTARY PROBABILITY AND STATISTICS 56 Fundamental Principle of Counting 56 Permutations _57 Combinations _ 60 Venn Diagrams _‘62 Classical Model of Probability 65 Conditional Probability 16 Bayes' Theorem _96 Random Sampling 99 3 PROBABILITY DISTRIBUTIONS __103 Discrete Distributions _103 Geometric Density Function _106 Poisson Density Function _108 Continuous Distributions _109 Uniform Density Function 114 Exponential Density Function _116 Change of Variable Technique 177 Moment Generating Function Technique _183 Distribution Functions _ 186 ix¢ Distribution 313 Additional Applications 364 Difference of Means of Normal Populations Urheberrechtlich geschiltztes MaterialLarge Samples With Mean of Unknown Variance 437 Small Samples With Mean of Unknown Variance 447 Proportions 480 Difference of tions 494 Bay Confidence Interval 502 15_POINT ESTIMATION _506 Estimating the Population Variance 506 The Method of Moments 509 Maximum Likelihood Estimators 512 Unbiased Estimators 530 Sufficient Statistics 539 Completeness 556 The Cramer-Rao Inequality and Efficiency 557 Bayesian Estimation 566 16 _ HYPOTHESIS TESTING 570 Level of Significance 570 Differences Between Proportions 613 The Norma! Approximation to the Binomial. 1 Variances _626 Difference Between Variances 630 Tests Involving the Poisson Distribution 637 Combinii ts of Different les 638 Type Errors 640 The Power of a Test _655 Likelihood Ratio Test for the Best Critical Region 659 17_REGRESSION AND CORRELATION ANALYSIS 665 Simple Regression 665 Linear Correlation 680 Spearman's Rank Correlation 699 Fisher's Z2-Transformation 708 Properties of the Regression Coefficients 720 Applications of Regression and Correlation Analysis 734 Factor Analysis xi Urheberrechtlich geschiltztes MaterialContingency Tables 809 Chi-square Tests and Contingency Tables 811 Fisher's Test and 2X2 Tables 823 ‘ov-Smirnov Statistic _ 876 xii Urheberrechtlich geschiitztes MaterialUNITS CONVERSION FACTORS This section includes a particularly useful and comprehensive table to aid students and teachers in converting between systems of units. The problems and their solutions in this book use SI (International System) as well as English units. Both of these units are in extensive use throughout the world, and therefore students should develop a good facility to work with both sets of units until a single standard of units has been found acceptable internationally. In working out or solving a problem in one system of units or the other, essentially only the numbers change. Also, the conversion from one unit system to another is easily achieved through the use of con- version factors that are given in the subsequent table. Accordingly, the units are one of the least important aspects of a problem. For these reasons, a student should not be concerned mainly with which units are used in any particular problem. Instead, a student should obtain from that problem and its solution an understanding of the underlying principles and solution techniques that are illustrated there.5.787 x104 1,639 x 105 4329x107 0.7646 0.91 x 10°? 1 5.73x 10% 57.3 4448 x 10° 0.624 x 10"8 1.356 x 107 3.766 x107 192 3,785 x 107 0.8327 gallons (liquid British Imperial)gallons (iquid US.) pints (liquid U.S.).. 61.02 0.2642 2.413 1 0.0929 0.0929 ila i 3.28 39.37 6214x104 28.35 14.1 0.4047 2.357 x 10% 3.03 x 105 1.182% 109 0.9863 1341 0.39377.9577 x 10 3.125107CHAPTER1 DESCRIPTIVE STATISTICS GRAPHS AND FREQUENCY DISTRIBUTIONS © PROBLEM 1-1 Discuss and distinguish between discrete and continuous values. Solution: The kinds of numbers that can take on any fractional or integer value between specified limits are categorized as continuous, whereas values that are usually restricted to whole-number values are called discrete. Thus, if we identify the number of people who use each of several brarids of toothpaste, the data generated must be discrete. If we determine the heights and weights of a group of college men, the data generated is continuous. However, in certain situations, fractional values are also integers. For example, stock prices are generally quoted to the one-eighth of a dollar. Since other fractional values between, say, 24.5 and 24.37 cannot occur, these values can be considered discrete. However, the discrete values that we consider are usually integers. @ PROBLEM 1-2 (a) Suppose a manufacturer conducts a study to determine the average retail price being charged for his product in a particular market area. Is such a variable discrete or continuous? (b) In conjunction with the previous study the manufacturer also wants to determine the number of units sold in the area during the week in which an ad- vertising campaign was conducted. Is this variable dis- crete or continuous? Solution: {a) Since an average may take any fractional Value, the average retail price is continuous. (b) We have a count. This variable must be discrete. The number of units sold is a discrete variable. © PROBLEM I-3 Twenty students are enrolled in the foreign language department, and their major fields are as follows: Spanish, Spanish, French, Italian, Prench, Spanish, German, German, Russian, Russian, French, German, German, German, Spanish, Russian, German, Italian,German, Spanish. (a) Make a frequency distribution table. (b) Make a frequency histogram. Solution: The frequency distribution table is con- structed by writing down the major field and next to it the number of students. Number of Students A histogram follows: Fon om wo Garaan Russian Spanish french italian In the histogram, the fields are listed and spaced evenly along the horizontal axis. Each specific field is represented by a rectangle, and all have the same width. The height of each, identified by a number on the vertical axis, corresponds to the frequency of that field. * PROBLEM 1-4 The IQ scores for a sample of 24 students who are entering their first year of high school are: 115 119 119 134 121 128 128 152 97 108 98 130 108 110 111 122 106 142 143 140 141 151 125 126 (a) Make a cumulative percentage graph using classes of seven points starting with 96 - 102. (b) What scores are below the 25th percentile? (Cc) What scores are above the 75th? (d) What is the median score?Interval Cumulative Cumulative Interval Midpoint Frequenc Frequen, Percentage 96-102 103-109 110-116 117-123 124-130 131-137 138-144 145-151 152-158 The frequency is the number of students in that interval. The cumulative frequency is the number of students in intervals up to and including that interval. The cumula- tive percentage is the percentage of students whose IQ's are at that level or below. Cumulative Percentage = Cumulative Frequency , 100 & We, will plot our graph using the interval midpoint as the x coordinate and the cumulative percentage as the y coordinate. 100 90 60 70 60 Cumulative 5 Percentages > 4a 99 106 113 120 127 134 141 148 155 Student 1.9. Scores (b) The 25th percentile is defined to be a number that is exactly greater than the lowest 25 % of the scores. We want to know the score that is at least greater than (.25)24=6 other students. The 6 lowest scores are 97, 98, 106, 108, 108 and 110. We cannot use 111 as a 25th percentile since another student has that score so we use 110.5. (c) The 75th percentile is the score which exceeds the lowest 75% of the population but is less than the top 25 % of scores. We want the score below 6 students and above 18. The 6 highest scores are 152, 151, 143, 142, 141, 140. The next highest is 134. As our 75 % percentile we can take any value between 134 and 140. We will take 137, the average of 134 and 140. (d) The median is the value which half of the values of the Population exced and half do not. There are 12 values & 123