First Steps In Research and Statistics

A Practical Workbook for Psychology Students

Dennis Howitt, Duncan Cramer

• Publisher: Routledge
• Year: 2003
• Print ISBN: 9781138464537
• eBook ISBN: 9781134635498
• Topic: Psychology
• Subtopic: History & Theory in Psychology
• Index: Psychology

First Steps in Research and Statistics is a new, very accessible approach to learning about quantitative methods. No previous knowledge or experience is assumed and every stage of the research process is covered.
Key topics include:
* Formulating your research questions
* How to choose the right statistical test for your research design
* Important research issues, such as questionnaire design, ethics, sampling, reliability and validity
* Conducting simple statistics to explore relationships and differences in your data
* Using statistics to explore relationships and differences in your data
* Writing up your research report and presenting statistics
Simple and helpful worksheets and flow diagrams guide you through the research stages. Each chapter contains exercises with answers to check whether you've understood.

Part I

The basics of research

1

Asking research questions

INTRODUCTION

We all have our own ideas of what determines people’s thoughts and actions. Some of these ideas are seen as common sense. For example, it may seem obvious that similar people are likely to be attracted to each other. After all, ‘birds of a feather flock together’. Confusingly, we believe that people who are very different from each other are likely to get involved in relationships—‘opposites attract’.

Testing the truth (validity) of such ‘theories’ is an important part of the work of many psychologists. Empirical research (collecting evidence by observation) is the main method by which this is achieved.

Try Exercise 1.1. This will get you thinking about possible psychological explanations of human behaviour.

Introductory textbooks in psychology contain numerous examples of the importance of research in psychological theory and knowledge. Research is an essential component of the discipline. Consequently students must know the broad principles of doing research. But, in addition, they must understand the more detailed practicalities of doing research. These are not the same activities. Reports of psychological research contain much detail but, nevertheless, lack many of the important details about the practicalities of doing research. This workbook is designed to help you to actually plan and do research.

In other words, we hope that you will learn how to do research to evaluate the explanations of human behaviour you listed in Exercise 1.1. Have you any ideas about how to do this now? As you work through this book you should begin to get more ideas.

A number of basic concepts such as variables and hypotheses need to be understood before you can get a clear idea about how psychologists answer questions through research.

EXERCISE 1.1
LIST EXPLANATIONS OF BEHAVIOUR

Write here three or more of your own or other people’s favourite or most frequent explanatory principles of human behaviour. We have given suggestions in grey— you probably have better ones. You are welcome to write over our humble efforts.

1 Blood is thicker than water.

2 Men with big powerful cars are disguising their sexual inadequacies.

3 Children of lone-parent families are at risk of delinquency

VARIABLES

The concept of variable is basic to psychological research. A variable is anything that varies and can be measured.

Put another way, a variable is any characteristic that varies in the sense of having more than one value. For example, the variable of attraction (as in opposites attract) consists of two values at the very least. An individual is either ‘attracted’ to the other person or ‘not attracted’. ‘Attracted’ and ‘not attracted’ are the two different values of the variable when we measure it in this way.

Of course, the variable of attraction may have a whole range of different values— from ‘strongly attracted’ through ‘neither attracted nor unattracted’ to ‘strongly unattracted’. The number of different values a variable has depends on just how the psychologist decides to measure that variable. A common way of measuring psychological variables is to use a simple five-point measuring scale such as:

1 Very attracted

2 Attracted

3 Neither attracted nor unattracted

4 Unattracted

5 Very unattracted.

As you can see, this scale has five different possible answers or values.

THE HYPOTHESIS

Very simple research may be concerned with counting how common a particular form of behaviour is among people. For example, we could research the number of people who have ever fallen in love at first sight. The answer to this question may be of great interest. For example, if research shows that 75 per cent of people had fallen in love at first sight then we would know that this is a very common human experience.

However, relatively little psychological research aims only to count the frequency of occurrence of things. More often, psychologists want to find the reasons why things occur, that is to explain their occurrence. Consequently most research tests a specific idea or hypothesis. The hypothesis is sometimes called the alternate hypothesis. An example of a hypothesis might be based on the suggestion that we are likely to fall in love at first sight with physically attractive people. Thus the hypothesis might be written: ‘The physical attractiveness of the other person affects the likelihood of falling in love with them at first sight.’

A research hypothesis has two essential features: 1) it contains a minimum of two variables; and 2) it suggests that there is a relationship between these two variables.

Another feature is not essential but nevertheless desirable. The hypothesis may describe what the researcher expects the relationship between the variables to be. We expect that people will be more likely to fall in love with someone generally thought to be physically attractive. This is the same as saying that they will be less likely to fall in love with someone generally considered physically unattractive. A directional hypothesis is one in which we specify what we expect the relationship between the variables to be.

The alternative to the directional hypothesis is, not surprisingly, called a non-directional hypothesis. A non-directional hypothesis is employed when we expect that there is a relationship between the variables but cannot specify its nature with any certainty. For example, we might think that there is a relationship between falling in love at first sight and physical attractiveness but feel uncertain quite what the relationship is. It is possible that we fall in love at first sight because the physical attractiveness of the other person makes them super-sexy to us. However, the fact that they are physically attractive may give us the impression that they are attractive to many people. This may warn us that they are likely to reject our advances and chat-up lines. Consequently we protect ourselves by not falling in love with them. Both of these alternatives may appear equally feasible to us. In these circumstances, because we are not sure which relationship is likely to be correct, we may prefer a non-directional hypothesis.

Directional hypotheses are more exacting than non-directional hypotheses. The convention is that in order to test a directional hypothesis in a statistical analysis, the researcher must be able to strongly justify its use on sound theoretical grounds or on the grounds of expectations generated by previous research. Obviously students must develop a degree of familiarity with the theoretical and empirical research literature before they can convincingly use directional hypotheses. Nevertheless, it is relatively easy to rephrase a non-directional hypothesis as a directional hypothesis. Exercise 1.2 gives you practice in using the appropriate phrases, though not the theoretical and research justification for your prediction.

EXERCISE 1.2
MAKING NON-DIRECTIONAL HYPOTHESES

Reword the following directional hypotheses to make them non-directional.

1 People with higher self-esteem will feel more accepted by their parents.

(e.g. Self-esteem is related to parental acceptance.)

2 Boys brought up without fathers will be less masculine than boys brought up with fathers.

3 Unemployment leads to increased stealing.

Note

Suggested answers are found at the end of the chapters for this and most of the remaining exercises in the book.

EXERCISE 1.3
CHANGING NON-DIRECTIONAL INTO DIRECTIONAL HYPOTHESES

Reword the following non-directional hypothesis so that they become directional hypotheses.

1 The number of children in a family will be related to children’s co-operativeness with others.

(e.g. Children from large families are more co-operative with others.)

2 Emotional stability will be related to satisfaction with close relationships.

3 Pet ownership is related to stress.

It is also useful to be able to turn non-directional hypotheses into directional ones. Exercise 1.3 gives you some practice in doing this.

Note that each of the hypotheses in Exercises 1.2 and 1.3 consist of two variables. So in Exercise 1.2 the pairs of variables are self-esteem and feeling accepted, father absence and masculinity, and employment status and theft. What are the pairs of variables in Exercise 1.3?

There is a big difference between psychological language and ordinary language. Psychologists when they put forward a research hypothesis, such as ‘Unemployment leads to increased stealing’, do not assume that all unemployed people steal and all employed people are honest. In psychology, the assumption is merely that the likelihood (or probability) of unemployed people stealing is higher than that for employed people. In other words, we anticipate that the percentage or proportion of unemployed people stealing will be greater than the percentage or proportion of employed people stealing.

THE NULL HYPOTHESIS

For every hypothesis in research, there is a null hypothesis. This is virtually the same as the hypothesis but states that there is no ...

Table of contents

• Cover Page
• Half Title page
• Title Page
• Copyright Page
• Contents
• List of boxes
• List of figures
• List of tables
• List of workshees
• Preface
• Part I The basics of research
• Part II The statistical analysis of single variables and simple surveys
• Part III Exploring correlational relationships in survey/non experimental studies
• Part IV Exploring differences between group means
• Part V The essentials of designing experiments
• Part VI The report
• References
• Index