In this chapter:
- Go to Analyze Results
- Test Data and Auto Reports
- Statistical Analysis: Correlation
- Statistical Analysis: Regression
- Residuals Table
- Top-bottom Data Organization
Data Analysis Mode
If you want to analyze the collected data, click Analyze Data on the top bar to the left from the survey name in the editor mode:
First, you need to create a new report for your survey data. QUESTIONSTAR offers different types of reports where data is presented in different formats.
Test Data and Auto Reports
If you want to get a report while the data collection is still in progress and you do not have enough data for analysis, you can choose to fill the missing data with test input. All fields, including text fields, will be randomly populated with test data. Once you have collected all survey data and set up the report, you can delete test data by clicking one button. This will not affect the real survey data.
Visualizations will be automatically generated for each question, depending on the type of used data.
We will use test data and automatic visualizations to show you what QUESTIONSTAR can do.
An example of automatic visualizations:
On the same report page, you can add another graph type by clicking Add Visualization or use the right pane to change the layout of your current visualization.
Statistical Analysis: Correlation
To go to statistical data analysis, select the required type of analysis on the left pane:
If you want to conduct correlation analysis, first, you need to select the variables. You can do so by clicking on them (you don’t need to press any keys); the selected items will be highlighted in dark color:
By clicking the Run button, you can generate a report that will contain a correlation analysis table where you can choose the type of analysis and its graphical representation:
Statistical Analysis: Regression
To launch regression analysis, you also need to select variables on your left, but the first selected item will be regarded as a dependent variable and marked with a white key. The remaining items will be treated as independent variables (grey key). By clicking on the key sign, you can change the status of any selected variable to dependent or independent.
After clicking the Run button, you will see a table with regression coefficients on your screen. On your right, you will see multiple settings for setting up data representation; you can turn on and off data categories and graphical analysis for your data in line with your needs:
After setting up relevant statistics on your left, selecting two variables from the list, and clicking the Run button, you will see a page containing t-test coefficients. You can run this test for both dependent and independent samples:
If you click on Cross Table after selecting relevant variables, you will get a table that can be customized to view residuals, expected frequencies, etc. (use the settings on the right pane).
Top-bottom Data Organization
Sometimes, you may need not only to assess data for each individual answer, but also group several lower and upper values. You can organize your data in the Data Source menu on your right.
Here is what a diagram based on all answers without grouping data may look like:
To group your data, open the Combine Alternatives tab in the Data Source and check the relevant boxes for one pair of variables, click Combine, and then do the same for the second pair of variables.
If you want to create different visualizations with different data groups for one question, use the left pane to create a new stats page for this question. Otherwise, every time you group your data differently, this will affect all visualizations for a given question.
A diagram with grouped data with disabled No Response data:
You can also use the Add Visualization option to insert a pivot table showing the correlations between the response groups in text format.