Protecting the Respondent
To ensure that every individual participating in certain types of surveys can feel secure in providing accurate and confidential responses, there are several ways to protect the individual.
Disconnect Contact Information
One way is to completely disconnect the respondent's information from the response. This is typical in employee surveys where contact information is no longer needed once the individual has responded to the survey.
Grouping Background Information
Surveys where the respondent's identity needs to be protected should use minimal background information, and the information used should not be traceable to the individual. Therefore, it is common practice not to use details that are more specific than necessary, such as not using the exact age but age ranges, not using specific job titles (e.g., "CEO") but groupings (e.g., "management"), and not using exact length of employment but ranges for employment duration.
Limited Filtering
By limiting the number of ways we allow filtering to be tied to the role/purpose of each reader, we can also minimize the risk of misuse. For example, filtering by gender may be relevant for an HR department that wants to see how different questions perform based on gender, but for an individual manager with 8 employees, gender distribution may not be an appropriate tool. It can also be the case that as a manager, you can only see responses from your direct reports or a few levels below, while you cannot see detailed information for a specific department several levels below your own.
Automatic Protection with Few Responses
The system automatically provides protection if there are fewer than a certain number of responses. Most commonly, Quicksearch uses a threshold of 8 responses for dispersion (to view how responses are distributed for a question) and 5 responses to display calculated values such as NPS, eNPS, mean or index values, and employee indexes.
The thresholds of 8 and 5 responses work well in many cases, but depending on the organization, purpose, or context, it is possible for your organization to adjust these thresholds.
Time as a Tool to Protect the Respondent
If a person with access to the results can see exactly when another person responds to the survey, it is also possible to see the impact of the response on the overall index when the response comes in by examining before and after. A simple but effective way to protect the individual is to update the results at regular intervals rather than in real-time. This can reduce the risk of individual responses being interpreted.