The SPARKS Strain Index
One measure of how a group is doing, reported at cohort level, never at the level of a person. Defined here precisely enough that a security reviewer, a benefits consultant, and a research methodologist can each check our work.
Strain Index v1.0 · Method questions: privacy@memores.me
We help a benefits consultant defend the number by publishing the method, the comparators and the suppression floor that produced it.
What it measures, and what it deliberately does not
The Strain Index is a single number between 0 and 100 describing the current stress load carried by a group of people. It is derived from one facet of the SPARKS Prism, How you’re doing now, which is the only facet designed to change. The other six facets describe stable traits and contribute nothing to this index.
That separation is the point. We do not claim that a quarter of coaching alters anyone’s attachment style or personality. We claim that strain moved, that the movement is measurable, and that the profile explaining why it moved is available to the individual and to no one else.
Direction and range
The example marks a cohort at 54 against an industry benchmark of 47, the figures published on the organisational reporting page. A score is always read against the benchmark and against the cohort’s own history, never on its own.
Six facets describe the instrument. One reports the reading.
What each band means and what it asks of you
A number without a recommended response is decoration. Every band carries an action, because the index exists to prompt one.
| Band | Range | Reading | Recommended response |
|---|---|---|---|
| Low | 0 to 34 | Strain is below what this population would normally carry. | Observe. Look at what this cohort is doing that others are not. |
| Typical | 35 to 49 | Within the normal band for the benchmark population. | No action. Watch the direction of travel rather than the level. |
| Elevated | 50 to 64 | Above benchmark. Sustained readings here precede measurable turnover. | Investigate at cohort level against workload, change events and manager span. Targeted skill-building. |
| High | 65 to 100 | Well above benchmark, or rising quickly out of Elevated. | Structural response, not an individual one. Escalate to the sponsor. Review conditions, not people. |
A band is a property of a cohort, never of a person.
There is no individual band, no individual score, and no view in the product that produces one.
What goes into the number, and what it is measured against
Six properties define the index. Where something is not settled yet, this page says so rather than leaving you to find out later.
One instrument, measured repeatedly
The Stress and Anxiety Measurement, taken at baseline and on a repeating cadence. Everything that contributes to the index is named on this page. Nothing contributes that is not named here.
Checked against measures you already know
PSS-10 and WHO-5 are reported alongside the Strain Index for the same cohort and the same period. Both are public domain and already familiar to benefits consultants and academic reviewers, so you can check a proprietary measure against a recognised one rather than taking it on trust.
Zero to one hundred, higher meaning more strain
Raw instrument scores are transformed onto a fixed 0 to 100 scale so that cohorts, quarters and organisations can be compared directly.
Compared against a stated population
A score means little alone. Every figure is shown against an industry benchmark and against the cohort’s own history, so you are reading a direction of travel rather than an isolated number.
Nothing below thirty people is ever shown
No cohort smaller than thirty individuals renders, in any view, at any time. Repeated or recombined queries cannot be used to narrow below that floor. This is the same threshold stated on the trust, privacy and organisational reporting pages.
Every figure carries the method that produced it
When inputs, scale or bands change, the version increments and a changelog records what moved and why. Historical figures keep the version label they were published under; we do not quietly restate old numbers under a new method.
Two views of one measure
There is one index. It is presented in two views for two audiences, computed identically and suppressed identically.
For the sponsoring organisation
Department, function or site, minimum thirty people. Live adoption, monthly movement, quarterly full breakdown, and before-and-after analysis on request.
For insurers and reinsurers
Whole book or segment, aggregated across organisations, reported quarterly. Same measure, same math, same suppression floor, one level further up.
Cohort view, illustrative
Industry benchmark
Illustrative figures. Two cohorts in this organisation are not shown at all, because each has fewer than thirty people and the report suppresses them rather than rounding them.
What the Strain Index may not be used for
Every limit below is written into the organisational agreement in the same words it appears in here. A limit that lives only in marketing copy is not a limit.
- It is not a diagnosis. It measures self-reported stress load in a group, not a clinical finding about anyone in it.
- It is not an individual score. No individual Strain Index exists, is computed, is stored, or can be produced on request, including by us.
- It may not inform any decision about a person. Hiring, promotion, performance management, discipline, redundancy selection and benefits eligibility are all excluded uses.
- It may not be used to underwrite an individual. The portfolio view prices books and segments. It cannot price a person, and any attempt to narrow it toward one meets the suppression floor first.
- It does not predict who will leave. It reports a cohort’s current strain against a benchmark and against its own history. Any forward-looking language must name what is projected, over what horizon, with what error.