U nderstanding Probability and Statistics Understanding Probability and Statistics A Book of Problems Ruma Falk with the cooperation of Raphael Falk The Hebrew University of Jerusalem Department of Psychology and School of Education Jerusalem, Israel LONDON AND In memory of Leslie V. Glickman (1945-1991). Friend, teacher, statistician. Contents Preface ix I PROBLEM S TO SOLVE 1 1 Descriptive Statistics I 3 1.1 Measures Characterizing Distributions I .......................... 3 1.2 Correlation and Linear Regression...........................................16 2 Probability I 21 2.1 Events, Operations, and Event Probabilities.........................21 2.2 Problems Involving Combinatorics..........................................27 2.3 Conditional Probabilities and Dependence/Independence between Events...........................................................................36 2.4 Bayes’ Theorem..........................................................................48 2.5 Probability Distributions and Expectations..........................59 3 Reasoning across Domains 69 3.1 Statistics, Probability, and Inference ....................................69 II M ULTIPLE-CHOICE PROBLEM S 85 4 Descriptive Statistics II 87 4.1 Scales of Measurement..............................................................87 4.2 Measures Characterizing Distributions I I .............................91 4.3 Transformed Scores, Association, and Linear Regression . 101 4.4 In Retrospect............................................................................110 vii viii Contents 5 Probability II 111 6 Normal Distribution, Sampling Distributions, and Inference 121 III ANSW ERS 131 Answers for Part I: Problems to Solve 133 Chapter 1. Descriptive Statistics I ..................................................133 1.1 Measures Characterizing Distributions I .....................133 1.2 Correlation and Linear Regression..................................142 Chapter 2. Probability I ..................................................................144 2.1 Events, Operations, and Event Probabilities...............144 2.2 Problems Involving Combinatorics..................................147 2.3 Conditional Probabilities and Dependence/Independence between Events . . . . 158 2.4 Bayes’ Theorem..................................................................170 2.5 Probability Distributions and Expectations..................189 Chapter 3. Reasoning across Domains ........................................203 3.1 Statistics, Probability, and Inference ...........................203 Answers for Part II: Multiple-Choice Problems 217 Chapter 4. Descriptive Statistics I I ...............................................217 4.1 Scales of Measurement.....................................................217 4.2 Measures Characterizing Distributions I I .....................217 4.3 Transformed Scores, Association, and Linear Regression.....................................................................218 4.4 In Retrospect.....................................................................218 Chapter 5. Probability II..................................................................218 Chapter 6. Normal Distribution, Sampling Distributions, and Inference.....................................................................................219 References 221 Index 232 Preface We certainly think we know whether we understand something or not; and most of us have a fairly deep-rooted belief that it matters. Skemp (1971, p. 14) Statistics and probability are often perceived as threats by students, particularly those in the humanities and the social sciences and not so rarely in the natural sciences as well. The very mention of statis tics evokes the notorious anxiety associated with mathematics. Fur thermore, dealing with uncertainty adds a disconcerting dimension to statistics. When you solve a mathematical problem, hard as it may be, you are rewarded by attaining certainty, but in working out a statistical problem, all your efforts may end up in an average or a probability value which is incapable of telling what will happen next. At best, you may learn what will happen in the long run, and that too, only in terms of proportions. Some students find this state of affairs quite frustrating. Without getting into the problem of the roots of math-anxiety and statistics-phobia and the detailed methods of overcoming them, we feel that the best way to allay these fears is by experiencing the revelation of understanding. One safe way toward this goal is to actively tackle problems concerning the concepts you learn. Nothing is more reassuring than coping with a problem by going back to the definitions of the concepts involved and manipulating them by the rules of the game. One solved problem, worked out on your own, may be worth tens of pages of passive reading. When some of our more diligent students ask for extra reading, we often discourage these good intentions and recommend instead that they use their energy for independent exploration of problems. Even reading a statistical text is best accomplished when conducted as a їх