You are always anonymous when using &frankly! We believe anonymity is essential to get the frankest answers and insights. Your employer defines the number of answers required to get an aggregated result (however it can never be set to lower than 3).
Anonymity and integrity framework
To ensure that all answers are anonymous, we use a number of methods. The most basic method that we employ is requiring a minimum number of answers before we show a given result. For any result that we produce, the system counts the amount of collected answer, and if the number of answers is below the required minimum, the result won't be shown.
We provide two such minimum thresholds in &frankly: one for average answers (e.g. where an average has been produced), and one for plots (e.g. where we plot/show many individual answers in a graph to illustrate how scattered results are). The default of these values is 3 but is customizable by the organization – however, a strict system limit is set that never allows the values to be below 3 answers.
Only administrators in &frankly can see and edit this threshold. In order to change the limit, please follow the instructions found in this article: Anonymity threshold
The second method employed to protect individual answers from being divulged is to restrict slicing of results on attribute level to fixed ranges, which cannot be changed during an ongoing measurement.
The third method that is used to ensure anonymity is an integrity framework, which looks at other general possible situations of isolation of individual answers in the organizational structure and/or by the attributes defined on a company in &frankly. The results are restricted to be shown when there are identified isolation cases. Our integrity framework makes sure that no groups can be compared in cases where an individual answer would be possible to expose.
In order to address this possible case of individual answer isolation, &frankly looks at all possible differences in the current organizational structure, as well as all combinations of attributes and discrete attribute ranges to identify if there are potential isolation cases where an individual may be possible to isolate out.
When such a case is identified (where the number of uniquely answering individuals is less than or equal to 2), the case is resolved by granting visibility only to the largest group result and restricting visibility to results of the smaller group.
If you need more information or have any further questions regarding this, please contact us at firstname.lastname@example.org