Methodology
Taps are not people.
HABLA™ counts interactions. It does not count the people who had them, and it cannot — there is nothing in the anonymous experience that distinguishes one person from another, which is the point. Everything below follows from that one fact.
What every report carries
This panel ships with every report surface, every export, and every board brief. It is not collapsible, and it is not a footnote.
How to read this report
HABLA counts interactions, not unique people. Counts do not measure prevalence, diagnosis, or severity. A site value reflects the organizational placement experience used and does not confirm the person's physical location. An audience-pack value reflects the experience accessed and does not authenticate the visitor as a member of that group. Differences may reflect population size, awareness, access, placement, repeat use, or other factors. Use these signals to identify questions worth examining — not to identify individuals or estimate how many people have a condition.
The rules that produce those numbers
- Activity is not unique people, prevalence, diagnosis, or severity.
- Raw counts are not rates. A rate needs a denominator HABLA does not have.
- Reporting periods are fixed and non-overlapping. No custom or sliding windows.
- Below ten activity events, a value is suppressed. The floor is applied to each topic in each hour of each day, before the month is added up. So a topic that never reaches ten activity events inside a single hour reports a zero for the month — and that zero means no activity there reached the floor, not that nothing happened. The floor is never described as ten people.
- A placement or audience pack covering fewer than fifty people is not reportable on its own.
- No unrestricted raw export of anonymous activity, and no unrestricted query builder.
- No overlapping filters that could reconstruct a suppressed value.
What a count is a count of
Every figure HABLA reports is a count of activity events. It is not a count of whoever was behind them, and there is no operation anywhere in the product that turns the first kind of number into the second.
So a share of people is not a more careful figure — it is an uncomputable one. Nothing in the product counts people, so there is no quantity anywhere in it that “how many people” names. A percentage of a denominator nobody has is not a cautious number; it is a number about nothing. Where a topic is broken down by how much a statement fit, the three parts are shares of the reportable activity for that topic, and never shares of people, in any wording, in either language.
And some interactions produce no figure at all. Reaching a crisis resource is counted nowhere: no total, no trend, no alert, and no row in any report. It is not that the number is suppressed. It is that it is never made.
Raw counts are not rates either. A rate needs a denominator, and the denominator a rate would need is the one this page has just said does not exist.
Three values a reader could over-read
Misreading a report is not a presentation problem. A leader who believes a number means something it does not may act on a false belief about real people, so each of these says what it reflects and what it is not.
- A site value
- It reflects the organizational placement experience that was used — a particular poster, QR code or link. It does not confirm where the person was. HABLA never asks a device for its location and never stores one, so a value attributed to a campus is a fact about the material rather than about presence. HABLA draws no maps, for the same reason.
- An audience-pack value
- It reflects the experience that was opened. It does not authenticate whoever opened it as a member of that group: a staff member can open the student pack, a parent can open the staff pack, and nothing anywhere asks either of them to prove otherwise. Packs are open experiences, not gated identities.
- An hour
- It is access timing — when the HABLA experience was used. It is not need timing. It does not say when a difficulty arose, when somebody was struggling, or when anything happened to them, and every hour visualization is labelled as access timing so that the label travels with the chart into whatever deck it is pasted into.
When one period can be compared to another
Not automatically. Every comparison carries one of four states, and HABLA renders no trend arrow, no percentage change and no up-or-down language for a pair that is not comparable — a missing arrow is information, and a manufactured one is not.
Comparable
The same deployment configuration, the same period length, and both periods above the floor.
Not comparable — configuration changed
Sites, audience packs, placements, the topic library or a declared population band changed between the two periods. The signal may have moved because the deployment did.
Not comparable — below the floor
One period or both are suppressed. A suppressed value has no size to compare, and inferring one from the pair is the arithmetic suppression exists to defeat.
Not comparable — partial period
The period is not closed. HABLA uses fixed, non-overlapping periods precisely so that nobody can narrow a window until a small value falls out of it.
Two floors, not one
The activity floor answers how few taps may be shown. The population floor answers how few people a number may describe. They protect different things, so both apply.
Below the floor, nothing is shown
Two floors, not one. A count is suppressed below ten activity events, and the floor is applied to each topic in each hour of each day, before the month is added up; a placement or audience pack is only reportable on its own if it also covers at least fifty people.
-
Suppressed. Not rounded, not estimated, and never shown next to another number that could reveal it by subtraction.
-
A fixed threshold. It is not lowered for a larger contract — that refusal is the product.
-
Clears the activity floor. Reported as forty-two activity events — never as forty-two people.
-
The second floor. A placement or audience pack covering fewer than fifty people is not reportable on its own however high its count climbs — its activity rolls into the parent site.
The same two rules, in plain words
Imagine your school puts up a poster with a button on it. Any time something is bothering you, you tap it. You are not asked for your name, and there is nowhere to type one. What the school gets back later is a figure: how often that button was tapped.
Rule one: fewer than ten taps, and the figure is not shown
Not a smaller figure, and not a rounded one. Below ten taps the value is suppressed, and the report says that it is suppressed rather than quietly printing something else in its place.
A small figure invites a guess. Somebody sees a low number and starts wondering who it was — and they would be guessing twice over, because ten taps are not ten people. The same person can tap more than once, and nothing in HABLA can tell anyone whether they did. Guessing about one particular child is the exact thing this exists to prevent, so a small figure is not shown discreetly. It is not shown.
Rule two: fewer than fifty people, and that spot does not get a row of its own
Picture a poster in a corner of the gym that covers only twelve people. Ten taps there clear rule one easily, and they still say far more about that small group than the same figure would say about a whole school.
So a placement or an audience pack earns a row of its own only when it covers at least fifty people. Under that, its taps are still counted: they are added into the total above it instead of sitting on a line of their own. And when the whole organization is below fifty, there is one total for everybody and no breakdown underneath it.
And a breakdown that names topics is narrower still: it releases about the whole organization and about nothing smaller, at every tier. Which topics cleared the floor is never reported per site, per audience pack or per placement — because the list of topics that did clear it is also a list of the ones that did not, and at a unit that small the absence is a statement about whoever runs it.
Where does the fifty come from, if nobody is counted? The organization declares it when a placement is set up — a band saying roughly how many people it covers. That is a fact about the organization’s own map, not about anybody who tapped anything.
Rule one is about how few taps may be shown. Rule two is a floor on the population behind them: how few people a figure may describe. A large figure from a small group gives away as much as a small figure does, so bolting one door and leaving the other one open protects nobody.
- And neither floor is for sale. Not to the largest customer, not for any amount, not on any plan.
Every figure carries the limits of the figure.
The panel above is not a disclaimer added at the end. It is printed on every report HABLA produces, in the same words, because a number a reader can over-read is a number that gets over-read.