Introduction to Statistics with SPSS for Social Science

Gareth Norris, Faiza Qureshi, Dennis Howitt, Duncan Cramer

• Publisher: Routledge
• Year: 2014
• Print ISBN: 9781408237595
• eBook ISBN: 9781317861829
• Edition: 1
• Topic: Education
• Subtopic: Education Theory & Practice

This is a complete guide to statistics and SPSS for social science students. Statistics with SPSS for Social Science provides a step-by-step explanation of all the important statistical concepts, tests and procedures. It is also a guide to getting started with SPSS, and includes screenshots to illustrate explanations. With examples specific to social sciences, this text is essential for any student in this area.

PART 1

Descriptive statistics

Chapter 1

Why you need statistics

Types of data

Overview

• Statistics is a word used to describe data but also to assess what reliance we can place on information based on samples.

• A variable is any concept that we can measure and that varies between individuals or cases.

• Variables should be identified as nominal (also known as category, categorical or qualitative) variables, or score (also known as numerical or quantitative) variables.

• Nominal variables are simply named categories.

• Score variables are measured in the form of a numerical scale which indicates the quantity of the variable.

1.1	Introduction

Every day we are bombarded with statistical information, most times without realising. Simply by reading a newspaper, watching the television or even going out shopping, the chances are you will come across some form of statistical information, such as stories on crime rates, figures on how well or badly your football team has played in recent matches, discounts on clothes during the sales (10% off), television adverts stating that nine out of ten cats prefer a certain brand of cat food, opinion polls on how the three main political parties in the UK are viewed by the public, and so forth. The list is endless.

Statistics is a tool that can be used to answer a researcher’s questions, as research (such as dissertation projects) is very varied and can be quite complex. Only the researcher can fully know what they want their research to achieve – what issues they want resolving through collecting research data and analysing it. Unless the researcher clearly understands what they want the research to achieve, statistics can be of little help. Very often, when approached for statistical advice, we find that we have to clarify the objectives of the research first of all – and then try to unravel how the researcher thought that the data collected would help them. These are not statistical matters but issues to do with developing research ideas and planning appropriate data collection. So the first thing is to list the questions that the data were intended to answer. Too often, sight of the purpose of the research is lost in the forest of the research practicalities.

There are three types of statistical techniques:

1. Descriptive statistics provide ways of summarising and describing information that we collect from different sources. This is achieved using tables and diagrams to summarise data, and simple formulae which turn fairly complex data into simple indexes that describe numerically the main features of the data (Part 1).

2. Inferential statistics is about generalising from a sample to the entire population, such as from the British Crime Survey (BCS). The current BCS sample stands at 51,000 people aged 16 or over (The National Archives, 2011). The data can be weighted to make the analysis representative of the entire population of England and Wales (Part 2).

3. Data exploration techniques simplify large amounts of data that otherwise would be much too confusing. They allow researchers to clarify trends in their data (Parts 3 and 4).

1.2	Variables and measurement

A variable is anything that varies and can be measured. So, the age, height and hair or eye colour of a person are variables. There are only two different types of measurement in statistics – category and numerical measurements.

1. Category (nominal/qualitative) measurement is deciding to which category of a variable a particular case belongs. If we were carrying out research looking into the legal marital status (variable) of people, we would have to decide whether or not they were never married, married, divorced or widowed (categories). There are...

Table of contents

• Cover Page
• Half Title page
• Title Page
• Copyright Page
• Contents 1
• Contents
• Guided tour Page
• Introduction
• List of figures
• List of tables
• List of boxes
• List of calculation boxes
• Acknowledgements
• 1 Descriptive statistics
• 2 Inferential statistics
• 3 Introduction to analysis of variance
• 4 More advanced statistics and techniques
• Appendices
• Appendix A Testing for excessively skewed distributions
• Appendix B Extended table of significance for the Pearson correlation coefficient
• Appendix C Table of significance for the Spearman correlation coefficient
• Appendix D Extended table of significance for the t-test
• Appendix E Table of significance for chi-square
• Appendix F Extended table of significance for the sign test
• Appendix G Table of significance for the Wilcoxon matched pairs test
• Appendix H Tables of significance for the Mann–Whitney U-test
• Appendix I Tables of significant values for the F-distribution
• Appendix J Table of significant values of t when making multiple t-tests
• Appendix K Some other statistics in SPSS Statistics
• Glossary
• References
• Index