Qualification

The data qualification is a functionality of Sphinx iQ which allows especially to identify the variables poorly-documented or badly-informed, to distinguish the singular observations to replace the non-responses. For example, some respondents always give the same answer to the scale questions, in order to respond to the survey faster. The dataset that we obtain may be biased, and necessary to qualify its data in order to make them more pertinent. This operating mode is accessible by two methods :

• In the tab Data of the Spreadsheet, click on the button Qualifiy.
• From the home panel, click on the button Qualifiy in the stage Data management.

You reach a dialog box in which you must select the type of changement you wish to perform on your data :

• Spot the singular observations or poorly-documented
• Replace the non responses (by the most cited modality for example)
• Spot the non-relevant variables
• Extract a cylindered sample

1 Select the operation you wish to perform on your dataset. In our example, we will replace the non-responses (empty cells in the table of the central area)

2 Click possibly on Dataset quality in order to have a graphical overview of the relevance of your data. The button Options allows to perform the analysis of dataset quality on a group of variables selected.

3 Select the variables you wish to replace by the non-responses

4 Choose the value to replace the non-responses : average, mode (for numerical variables), or previous value (previous value for all types of variables).

Click on Detailed results if you wish to visualize the set of observations concerned with replacing the non-responses, and Finish to validate the replacements in the dataset.