Learning to Use Statistical Tests in Psychology | ||

Greene (Judith) & D'Oliveira (Manuela) | ||

This Page provides (where held) the Abstract of the above Book and those of all the Papers contained in it. | ||

Colour-Conventions | Disclaimer | Notes Citing this Book |

**BOOK ABSTRACT: **__Back Cover Blurb__

- Praise for the first edition: An excellent textbook which is well planned, well written, and pitched at the correct level for psychology students. I would not hesitate to recommend Greene and d’Oliveira to all psychology students looking for an introductory text on statistical methodology.

→*Bulletin of the British Psychological Society* - The second edition of this widely acclaimed text is an accessible and comprehensible introduction to the use of statistical tests in psychology experiments: statistics without panic. Presented in a new textbook format, its key objective is to enable students to select appropriate statistical tests to evaluate the significance of data obtained from psychological experiments. Improvements in the organization of chapters emphasize even more clearly the principle of introducing complex experimental designs on a ‘need to know’ basis, leaving more space for an extended interpretation of analysis of variance. In an important development for the second edition, students are introduced to modern statistical packages as a useful tool for calculations, the emphasis being on understanding and interpretation.
- This book shows psychology students:
- how psychologists plan experiments and statistical tests;
- why they must plan them within certain constraints;
- how they can analyse and make sense of their results.

- The approach is that:
- theory is always presented together with practical examples;
- theoretical points are summarized and understanding of them tested;
- statistical principles are introduced as part and parcel of the principles of experimental design.

**Judith Greene**was educated at Oxford and University College London. After ten years lecturing in psychology at Birkbeck College, she has been Professor of Psychology at The Open University since 1976.**Manuela d’Oliveira**did her first degree in experimental psychology at Birkbeck College. London and her PhD at The Open University where she worked from 1976 to 1985 in the Faculty of Social Sciences. After 14 years as Science Officer for The British Council in Portugal she is now attached to the Museum of Science of the University of Lisbon.

- There have been an enormous number of textbooks which have claimed to present statistics in a simple way. Despite this, many psychology students still find the whole statistical business something of a mystery.
- How does this book differ from these other attempts?
- We believe that virtually all books on statistics feel obliged to start with the mathematical principles underlying probability distributions, samples and populations, and statistical testing. But, however simply these are presented, in our view they obscure the basic reasons why psychologists use statistical tests.
- So we had better come clean straight away. This book sets out to achieve one single aim. This is to enable students to select appropriate statistical tests to evaluate the significance of data obtained from psychological experiments. In other words, this book is concerned with inferential statistics as used in psychological experimental studies.
- We have concentrated on this to the exclusion of much else. Topics like descriptive statistics and the basic principles of probability are well covered in other statistical texts. Moreover, we will be concentrating on psychological experiments rather than other types of psychological investigation. There is nothing here about the use of surveys, observational techniques or psychometric tests of intelligence and personality. All we have included is the battery of statistical tests which are usually introduced to psychology students as part of their undergraduate laboratory course. We hope that, by aiming at a single target, we will maximize our chances of scoring a bull’s-eye.
- While this is definitely a ‘beginners’ book, it takes students from the simplest non-parametric tests, like the Wilcoxon test, through to complex analysis of variance designs. The principle is the same throughout: always to give the rationale for using appropriate statistical analyses for particular experimental designs. It is certainly our expectation that anyone who has mastered the why and how of the statistical tests given in this book will be in a position to understand the basic principles of statistical tests as presented in advanced textbooks of psychological statistics.
- Our belief is that, with the aid of this book, students will feel comfortable about the basis for selecting and applying all kinds of statistical tests. We hope that teachers will wish to use a book which frees students from much of the panic usually associated with statistics. That way they should be in a far more receptive state to learn.
- The major change in the second edition is that the book is now organized into three parts.
- Part I contains a general introduction to the principles of research and design.
- Part II presents all the information required for non-parametric tests.
- It is not until Part III that multivariable designs arc introduced to prepare students for analysis of variance.

- We hope this reorganization will have several advantages. Both students and teachers will gain from the clear division between non-parametric and parametric tests. The new approach embraces the principle of introducing more complex designs on a ‘need to know’ basis, providing room for a more extended treatment of concepts students find difficult, like degrees of freedom and interactions.
- The second innovation is to introduce students to the use of computerized statistics packages. Although many readers will have access to computer programs, we still provide step-by-step instructions for those who do not. One danger is that when students can simply press buttons on a computer they can lose sight of the purpose of statistical analysis. We have concentrated on ensuring that students understand the inputs and can interpret the outputs of programs.
- This book does not attempt to provide complete instructions for logging into particular packages, all of which differ slightly. A brief account is given in Appendix 1 about one of the most commonly used computer packages and the appropriate terminology. The intention is that students will have no problem in adapting easily to whichever programs are available in a particular institution.
- Note about subjects and participants: Traditionally psychologists used the word ‘subjects’ to describe the people taking part in experiments in order to distinguish them from ‘objects’. More recently, it has been agreed by the British Psychological Society that a better term would be 'participants’ in order to remind researchers that people are participating in their experiments. However, it is not common to use this term when describing experimental designs and statistics. For instance, ‘within-participants’ and ‘between-participants’ would sound quite odd, rather than the usual ‘between-subjects’ and ‘within-subjects’. The usage in this book refers to the people taking part in experiments and the differences between people as a source of variability. But, when the discussion from Chapter 2 onwards moves to types of experimental designs and the consequences of using same and different subjects, ‘subjects’ is used in its technical sense (as in the Decision Charts).

- The aim of this book is to explain the rationale for using statistical tests to evaluate the results of psychological experiments. The problem is that in psychology you have to carry out these experiments on human beings, often other students. Unlike most physical objects, human beings are unique, each interpreting and performing whatever task you set them in a slightly different way.
- You will find that the data and observations obtained from the people doing a psychological experiment are often extremely varied and that many of the things which influence their behaviour may have nothing to do with the experiment. It is for this reason that you have to sort out whether experimental results are really significant. And, as you will see, this is just what statistical tests enable you to do.
- You will probably be relieved to hear that the chapters which introduce the basic rationale for statistics and summarize all you need to know in order to select an appropriate statistical test are the shortest chapters in the book. These are aspects of using statistical tests that students often find rather puzzling, but we hope that these chapters will clear up all your worries.
- Other chapters in the book present statistical tests, explaining the rationale for each one, taking you ‘step by step’ through any necessary calculations and giving precise instructions about how to use the statistical tables in Appendix 2. For more complex types of statistical analysis you will be introduced to the latest types of computer programs for carrying out numerical calculations.
- One essential feature of this book is the questions which occur throughout the text. It is not enough to read the summaries presented in the progress boxes. The only way you can make sure that you understand the context of each section is to attempt the questions before looking up the answers at the back! It is equally important to work your way through the step-by-step instructions given for each statistical test. Otherwise, you will never gain the confidence which comes from fully understanding the rationale for a statistical test as you successfully complete the necessary arithmetical calculations.
- Let us end by making some important, and, we hope, encouraging points. The first thing to grasp is that statistical tests are not magic formulae to be turned to, desperately wondering which on earth to choose. They simply follow as a natural result of the kind of experiment you have chosen to do. What makes most people give up all hope of mastering statistics is the thought that they will find themselves presented with a huge amount of numerical data without the foggiest idea of how to deal with them. But this is quite the wrong way to go about things. The important thing is to decide what experiment you want to carry out. You will find that such a decision immediately narrows your possible choice of statistical tests to only one or two, and that there are good reasons for selecting one or the other.
- With statistical tests selection is all; the actual calculations are quite easy once you have understood the reasons for doing them. The aim has been to introduce the principles of using statistical tests without referring to any mathematical concepts. And, in order to do the calculations for the tests themselves, you will only need to know how to add, subtract, multiply, divide and square numbers. With modern pocket calculators and computer programs this really should be child’s play.
- Good luck - and if, in spite of everything, you do find yourself getting disheartened by statistics, turn back and reread this study guide.

The Open University, Open University Set Book, Second Edition, 1999 (2001 Reprint), Paperback. Naomi's book.

- Blue: Text by me; © Theo Todman, 2020
- Mauve: Text by correspondent(s) or other author(s); © the author(s)

© Theo Todman, June 2007 - July 2020. | Please address any comments on this page to theo@theotodman.com. | File output: Website Maintenance Dashboard |

Return to Top of this Page | Return to Theo Todman's Philosophy Page | Return to Theo Todman's Home Page |