Being able to present your data graphically is very important. SPSS allows you to create and edit a
range
of different charts and graphs in order to get an understanding of your data and the relationships between variables.
Though we can’t run through all of the different options it is worth showing you how to access some of the basics. The image below shows the options that can be accessed. To access this menu click on You will probably recognise some of these types of graph. Many of them are in everyday use and appear on everything from national news stories through to cereal boxes. We thought it would be fun (in a loose sense of the word) to take you through some of the LSYPE 15,000 variables to demonstrate a few of them.
Bar charts will probably be familiar to you – a series of bars of differing heights which allow you to visually compare specific categories. A nominal or ordinal variable is placed on the horizontal x-axis such that each bar represents one category of that variable. The height of each bar is usually dictated by the number of cases in that category but it can be dictated by many different things such as the percentage of cases in the category or the average (mean) score that the category has on a second variable (which goes on the horizontal y-axis). Let’s say that we want to find out how the participants in our sample are distributed across ethnic groups - we can use bar charts to visualise the percentage of students in each category of ethnicity. Take the following route through SPSS:
In this case we want the We could also alter the ‘ The ‘ The next thing we need to do is tell SPSS which variable we want to take as our categories. The list on the left contains all of the variables in our dataset. The one labelled
As you can see all categories were represented but the most frequent category was clearly White British, accounting for more than 60% of the total sample. Note how our chart looks somewhat different to the one in your output. We’re not
cheating...
we simply unleashed our artistic side using the
The line chart is useful for exploring how different groups fluctuate across the range of scores (or categories) of a given variable within your dataset. It is hard to explain in words (which are why graphs are so useful!) so let’s launch straight in to an example. Let’s look at socio-economic status ( This time take the route As before we want You will notice that the ‘
The line chart shows how average scores at age 14 for both males and females are associated with SEC (the category number decreases as the background becomes less affluent). Students from more affluent backgrounds tend to perform better in their age 14 exams. There is also a gender difference, with females getting better exam scores than males in all categories of SEC. What a useful graph!
Histograms are a specific type of bar chart but they are used for several purposes in regression
analysis
(which we will come to in due course) and so are worth considering separately. The histogram creates a frequency distribution of the data for a given variable so you can look at the pattern of scores along the scale. Histograms are only appropriate when your variable is continuous as the process breaks the scale into intervals and counts how many cases fall
into
each interval to create a bar chart. Let’s show you by creating a histogram for the age 14 exam scores. Taking the route
We are only interested in graphing one variable,
The frequency distribution seems to create a bell shaped curve with the majority of scores falling at and around ‘0’ (which is the average score, the mean). There are relatively few scores at the extremes of the scale (-40 and 40). We will stop there. We could go through each of the graphs but it would probably become tedious for you as the process is always similar! We have encouraged you to use the |