SPSS for Researchers
This course by Associate Professor Brian Phillips covers many of the procedures SPSS researchers use to set up, modify and present data.
Monday 18 – Wednesday 20 February 2019
Associate Professor Brian Phillips
Swinburne University of Technology
Some basic knowledge of descriptive statistics, Introduction to Statistics, or equivalent experience presumed. The course is designed for people with little or no experience in the use of SPSS, though basic keyboard skills and/or experience with Microsoft Windows are expected. It is also useful for people with some SPSS experience to enhance their knowledge of the package. This course provides the beginner in quantitative data analysis with the basic requirements for analysis in an SPSS computing environment but is also useful for those with some experience with SPSS and statistical analysis.
This course will teach you many of the procedures that researchers who use SPSS regularly use to set up, modify and present their data. The focus is on setting up the data in a way suitable for the analysis, and will use descriptive statistical techniques to demonstrate many features of SPSS. Some use will be made of basic statistical concepts, though it is not a comprehensive course in statistics.
The course is suitable for those with little or no knowledge of SPSS, but will also cater for those who wish to improve their SPSS skills. You will learn how to establish an SPSS file and how to generate appropriate summaries, tables and graphs suitable for statistical reports. We will focus on the data management features in SPSS that researchers commonly use. Real life datasets will be used throughout the course to demonstrate features in SPSS.
The target audience for this course is anyone who has to analyse statistical data for their employment or research but has limited background in statistical analysis. After you have mastered the fundamentals of SPSS, you will have confidence in learning the more advanced statistical procedures that you can do in SPSS.
We will start by using data obtained from an annual survey which aims to assess people's opinions on a number of social issues and use the SPSS menus to learn about the key elements of a SPSS 'job' – defining and preparing data for analysis, data formats, variable names, labels, missing values and levels of measurement. We will see how to develop a codebook which will be used to establish a SPSS job and demonstrate SPSS with some procedures such as basic tables and simple graphs.
Graphs are an important feature of a statistical report to help get the researcher's message across in a clear and simple way. We will learn about the main features of graphs, what graphs to use for different types of data and how to use the extensive graphing facilities in SPSS. Throughout the course graphs will be used as appropriate, either from the SPSS Graphs menus or within specific procedures.
Modifying and selecting data
Often the original data has to be modified and subsets selected so we can carry out the tasks we want to do. We will initially use the menus to form new variables using COMPUTE, group the data in different ways using RECODE, determine the number of occurrences certain answers across a set of questions using COUNT, and apply logical conditions on the data using IF, SELECT IF and SPLIT FILE.
From menus to syntax
While the pull down menus are popular, any regular SPSS user should make some use of SPSS syntax. Here you will learn how to use SPSS syntax to make using SPSS more efficient. In particular you will learn how we can convert Excel files to SPSS files and then use syntax to add variable definition such as labels, missing values, formats and measurement levels. Descriptive SPSS procedures, selected from FREQUENCIES, DESCRIPTIVES, EXPLORE and MEANS will be used as appropriate to demonstrate some statistics obtained from SPSS. Then we will learn how syntax can be used to modify and select data more efficiently. Many researchers use a combination of drop down menus and syntax as best suits them.
Using different formats
Sometimes we have data that is in alphanumeric, date or money formats as well as numeric. We will learn how to use different formats and some special SPSS procedures for handling them. For example we learn about some of the date functions such as how to calculate the time between two dates.
Much of the data collected in the social sciences is categorical and cross tabulating variables is important when looking for relationships. We will learn how to do this when we have two or more variables, which statistics are applicable and how to proceed when we start with a table of counts rather than the original raw data. Further, sometimes respondents can give several responses to a question so we will learn how to analyse such questions using Multiple Response procedures.
Other topics can be included depending on student needs and time constraints. These include Merging files, Restructuring data, and Reading raw data into SPSS.
Working on your own data
It is not necessary for students to have their own data, but for those who do will have the opportunity to work on their own dataset and develop it into a form suitable to answer their research questions.
Note that what is covered and the order of the course may vary depending on the needs of the class.
This course will take place in a computer lab. All notes and data files will be provided. The instructor's course notes will serve as the course texts.
Students needing to convert from another package to SPSS will find that this course provides a bridge to understanding the SPSS language and the operation of the basic procedures, although advanced procedures of the package will not be covered.