What Is an Analysis Level?
The analysis level setting is a feature that lets you choose the minimum unit of analysis when analyzing data.
This feature is only available for 'Diary survey' types, where a single respondent leaves multiple responses over a set period of time.
💡 Usage tip | The diary survey type is available for panel-based research conducted through the Opensurvey expert research service.
Why Should You Set an Analysis Level?
1️⃣ Interpret data using the right standard for your analysis goals.
Even with the same data, the appropriate analysis standard may differ depending on whether you want to know "how often a behavior occurred" or "how many people experienced it." Setting the analysis level lets you view data using the approach that fits your goals — either response count (frequency of behavior) or respondent count (proportion of people who experienced it).
2️⃣ Accurately interpret results even in surveys with repeated responses.
In surveys like diary surveys where a single respondent answers multiple times, the same data can have different meanings depending on the analysis standard used. Setting the analysis level lets you distinguish between the total frequency of all responses and the number of people who experienced the behavior, enabling an interpretation suited to the situation.
Difference Between the Response Count and Respondent Count Options
1️⃣ Response count (Response level)
This method analyzes data based on the total number of responses collected in the survey. It is useful for checking how frequently a specific behavior or activity occurred (total frequency).
🧚 Understanding response count through an example
If 100 panelists recorded their meals over 7 days and a total of 500 responses were collected, the analysis basis would be 500. In other words, the data is reviewed from the perspective of "how many times meal records were made."
2️⃣ Respondent count (Respondent level)
This method analyzes data based on the number of unique respondents who participated in the survey. It is useful for checking how many people experienced a specific behavior or event (proportion of people).
🧚 Understanding respondent count through an example
Suppose 100 panelists recorded their meals over 7 days and a total of 500 responses were collected. If the analysis is based on "people who made at least one meal record during the week," the basis becomes 100 people.
💡 Usage tip | For standard surveys where each respondent answers only once, the response count and respondent count are always identical, so the analysis level setting option is not provided for these surveys.
How to Set the Analysis Level
step 1. In the right sidebar of the Analytics screen (Results / Cross-tab / Report tab), locate the [Analysis level] section and select either [Response count] or [Respondent count] based on your analysis goals.
Step 2. Once you select an option, the figures in all charts and tables are immediately recalculated to match the selected standard. This setting is also applied in the same way when downloading a PPT report.
Have you got a good understanding of how to set the analysis level?
If you encounter any difficulties during the setup process, please contact us anytime via the customer support icon in the bottom right corner of your screen. We will do our best to help you resolve any issues you're experiencing.