Time commitment
Less than 2 minutes
Description
The purpose of this video is to explain how to use the "Crosstabs" feature in SPSS to analyze the relationship between two categorical variables. It covers how to select variables for rows and columns, generate a crosstab table, and visualize the results with clustered bar charts in the output window.
Video
Transcript
[Slide shows SPSS screenshot in Data View with the Analyze menu highlighted.]
There's one more descriptive statistic I want to show you within SPSS: this is the Crosstabs feature. So this is generally used if we have two categorical variables and as a side note, we can also use this for certain normality checking, which we won't go into today, but you can use this for other things and that's in the SPSS LibGuide if you need some help with that.
If you're trying to generate a Crosstabs, you're like, okay I know I've got 15 men and 15 women, but I want to know what group they're in: how many men are in group one, how many women are in group two, for example. You can do that generating crosstab. So it's not just one variable, you're taking two variables and crossing them to get crosstabs.
In SPSS, you click Analyze > Descriptive Statistics > Crosstabs. So again, if you're trying this on your own computer, Analyze > Descriptive Statistics > Crosstabs.
[Screenshot shows the Crosstabs dialog box. In the left panel, you see a list of available variables. On the right, there are boxes labeled Rows, Columns, and Layer. The buttons on the right let you specify additional analyses and formatting for the crosstab.]
This will open the Crosstabs dialog box which looks something like this, and again I've linked the SPSS LibGuide here at the top, so if you wanted to go and get some additional help on how to run Crosstabs, you can click that link. But you want two categorical variables, so in this case I have grabbed Gender and put that on the “Row(s):” and I've grabbed Group and I've put that on the “Column(s):”. You can switch these, they can go back and forth, it's just which way you want the output to look. So if you do it one way and you don't like it, you can always come back and run it a different way.
We make sure we select the box that says, “Display clustered bar charts”. It's kind of weird phrasing, but essentially what this is, is we want a graph to go with our data.
We want to know what's happening, what does it look like. And then once we do that, we click “OK” and in our output window, we'll get something that looks like this.
[Slide shows screenshots of the Output window beside a bar chart. The Output window shows a tree-style navigation pane on the left, with the right (main) pane showing Crosstabs tables.]
So again, we've got our data window, we've got our output window. In SPSS, it will spit out all of your analyses in the output window.
So we've generated a Crosstab. It gives you your case processing summary: So how many valid cases did you have? How many observations were in this analysis? How many missing data pieces were in this analysis? And then it gives you your crosstab table: So it gives you the breakdown of, in our case, Gender based on Group. And all of this is fake data, so I've said we've got 15 male participants, 15 female participants.
We've got some folks who like small drinks, some who like medium drinks, and some who like large drinks. This is all fake data. If you had real data, this would be a little bit more meaningful, but in this case, we've got equal groups, and we can see that in our graph that spit out as well. It's essentially saying, well, we've got five in each group. The count of how many participants we have in our two categorical variables.
We've done a 2 × 3 Crosstabs, there's five participants in each group. Pretty easy.
Again, Crosstabs is generally not recommended for continuous variables, because if you've got 30 different decimal points, for example 30 different continuous [values], it will do male by female times every single one of those options, which is a lot and probably not what you're looking for. Crosstabs, we generally use this for continuous variables.
[Questions? Contact us. UG Library. Website: lib.uoguelph.ca. Email: library@uoguelph.ca.]
License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
- Ask Chat is a collaborative service
- Ask Us Online Chat hours
- Contact Us