To learn more about the how projects & questions are organized in general, visit this article.
Number of rows / responses
The number of rows is displayed on the Manage Projects page on the very right.
In this example, the project has 2 questions with 11 responses each. Duplicates as well as empty rows are counted as well.
What are duplicates?
In every feedback context it can happen, that different respondents give the exact same response(s). These additional responses are called duplicates. As identical responses are being treated consistently by our AI, duplicates are displayed in a summarized way. Duplicates can be spotted in the Coding View as a response with a small number attached to it, which depicts the amount of times the response occurs.
In this example "Great service" was mentioned 10 times.
As duplicates are grouped together in the coding view, the number of rows in the coding view (screenshot below) might be lower than the count on the Manage Projects view. The number of rows matches the number of unique responses (see paragraph below).
What is the difference between “number of responses” and “unique responses”?
In the "Statistics" box (visible in the Cockpit and Coding View), the most important numbers characterizing the question are displayed.
Deducting all duplicates from the total number of responses results in the number of unique responses.
How are empty responses handled?
Empty responses are responses that have been omitted by the respondent. Caplena automatically identifies these and groups them together (i.e. you will never see more than 1 row being empty in the coding view). Empty responses do not require any credits and will never be billed.
How are duplicates and empty responses being handled within the visualizations?
In the visualizations, duplicates are treated like every other response. For example: Two people mentioning "Great service" will count as 2 in the charts.
Empty responses are being counted the same way as any other response. If you would like to omit empty responses from being counted in your charts, you can exclude them by selecting the "Empty responses" entry in the filter menu of your dataset and then ticking "Skip empty verbatims" checkbox.