What are Crosstabs?
Crosstabulated data is useful for showing a side by side comparison of two or more survey questions to determine how they are interrelated. In statistical terms, it is a joint distribution between two (or more) discrete variables such as product usage and demographics Within SurveyMonkey, crosstabs are created within the Analyze section of a paid account.
Take a look through our video to get started, or follow the written instructions below. (Note: This video is currently available in English only.)
What are Variables?
These are properties of some event, object, or person that can take on different values. Discrete variables are ones that contain possible values of "discrete" points on a scale. Example: The number of students attending college is a discrete variable.
When conducting survey analysis, a researcher often manipulates variables in order to test for an outcome. Survey data shows a synopsis of tallied numbers and information. Based on specific sets of that data, a researcher formulates a hypothesis and tests it to see if certain discrete variables prove what the overall data presents. Example: A researcher might compare the effectiveness of multiple cleaning products. In this case, the discrete variables are the cleaning products.
What is a Crosstabulation Table?
A crosstab gets its name from the layout of variable definitions into rows and columns thus making a table or matrix format.
- Rows and columns in the table correspond to the values of the first and second variables.
- The cells contain the frequencies or number of occurrences of the corresponding pairs of values of the selected variables.
Part 1: Building Crosstabs
A. Select the Question to Crosstabulate:
Select one question to represent the control or independent variable. A control variable is any factor that remains unchanged and strongly influences values.
- Questions like age, sex, education, etc. are commonly used for the crosstab/control question.
- This question is visible within the column (x-axis) of the table.
B. Select the Answer Choices to Include:
Within the control, choose the answer choices to include. You can include up to 5 options in SurveyMonkey. Example: If you want to filter by sex, pick "Male" and "Female" as the variables.
NOTE: You can pick up to 5 answer choices to include in the comparison.
C. Select Specific Questions to Compare:
By default, the Response Summary page presents the crosstabbed question compared with all questions in the design. If you do not create a Custom Report first, this can be done at any time. Decide which questions you want to see in relation to the crosstab question.
- They will be visible along the rows (y-axis) of the table.
- These questions are the ones you think are affected by the control variable.
Part 2: Crosstab Configuration
Females are more likely to use SurveyMonkey and provide more positive feature evaluations than males.
- Choose the "sex" question as the control/crosstab and select both the male and female answer choices as the independent variables.
- Next, view your product question(s) crosstabbed with the independent variables.
- Click the [Analyze] icon to the right of the survey on the My Surveys page. The Response Summary page opens.
- Click the [Crosstab Responses] button to open the Crosstab Editor page.
- Select the question to crosstabulate (the independent variable). Next, choose the answer choices to include in the tabulation.
Name it and click the [Save Crosstab] button when you are finished. The Response Summary page opens with the applied cross tab active. This page shows the Male/Female columns in teal in relation to all questions in your survey.
What questions can be used as the crosstab question?
The following can be used as the control/crosstab question:
- Multiple Choice (One or Multiple Answers)
- Matrix (One or Multiple Answers Allowed)
Matrix with Menu questions currently cannot be used as the control because they are three dimensional. The other Matrix and the Rating types can be used because they are two dimensional. This means they consist of columns and rows without menus.
- Because there is the potential for many answer choices to be included within each menu, the overall size of the question could become quite large in the Analyze page.
- The result would be a non useful view of the data.
Tips: Create Custom Reports and Add Filters
For this comparison of how Males and Females evaluated SurveyMonkey, you are also interested in the overall income of the respondents. You now want to see if a specific income results in a higher or lower evaluation of the crosstabbed data.
To learn more about creating this type of report, review the following: Creating a Crosstabbed Report