Somewhere in the middle of a survey about their own organization, most people pause and wonder who is going to read what they are about to say. That pause is the moment an assessment either earns the truth or settles for the polite version, and it is why I care about confidentiality more than almost anything else in this work.
It seems to me that people deserve to know three things before they answer: who can see their answers, what will appear in the report, and where the protection stops. KindTru limits what a report can reveal about one person or a small group. It cannot make every answer impossible to recognize, and I would rather say so at the start. A reassuring sentence in an invitation helps a little. What helps more, I think, is a system that keeps its rules on its own, on a busy Tuesday when nobody is thinking about them.
What KindTru does
A few rules are written into the code rather than left to good intentions. (The full list, with the reasoning behind each number, is on our confidentiality page.)
- No screen shows one person's answers, to anyone, including the pages only we can reach.
- No screen shows one group's average. If fewer than three people from a group (the board, the senior staff or the program staff) answer, their answers still count toward the organization's scores, but that group never gets an average of its own.
- If fewer than three people answer in total, there is no report at all.
- With three or four people in total, the report gives organization-wide scores and says nothing about views differing.
- With five or more people in total, the report may say that people in different roles see a topic differently, but only where at least two of the groups each had three or more people answer in that topic. It never says which group rated it higher.
Whoever runs the survey can see who has finished (so that reminders stop), but never what anyone said.
Where the protection stops
I want to be as plain about the limits as about the rules, because a confidentiality promise that reaches too far is its own kind of breach.
Your answers are saved as you go, tied to your private link. That is how you can stop and come back, and how your answers can be removed if you ask. It also means they sit in our database beside your invitation, and as the person who runs that database, I could look them up. I do not, and the only thing I will do with one person's answers is delete them when that person asks. (It is the same commitment the confidentiality page makes, word for word.)
In a very small organization, arithmetic is arithmetic. If four people answer and everyone knows which four were asked, a determined reader can narrow things down. The minimums make that harder; they cannot make it impossible.
And no system reaches the conversation afterward. If someone goes around the office asking what people said, the remedy is an old-fashioned one: agree out loud, before the survey goes out, that the answers stay in the report.
Two questions worth asking
If you are comparing assessment tools, one question tells you a good deal: if only two board members answer, what happens to their responses? I would look for an answer that says they count toward the whole but are never shown as a group of their own, and that the report says less as the number of people falls. Then I would ask the harder follow-up, the one I have tried to answer above: who could look at one person's answers, and what do they do with them?
My hope is that the people who answer your survey feel safe enough to say what they actually think, and that the conversation afterward is better for it. That, in the end, is what the whole exercise is for.