Methodology
How the Divorce Recovery Index is built, what it can support, and what it cannot.
This page describes the method behind every monthly reading we publish. It is linked from every issue, and it is the page to check before citing anything we report. Where the method has limits, they are stated here rather than discovered later.
What is measured
The Index measures divorce adjustment: the extent to which a person has recovered from the ending of a love relationship. It is not a measure of happiness, mental health, or life satisfaction, and it should not be read as one.
Adjustment is reported across the four phases of the RIFT framework.
| Phase | What it captures |
|---|---|
| Thinking | Mental disentanglement from the former partner |
| Feelings | Processing of grief and anger |
| Identity | Sense of self independent of the relationship |
| Relationships | Capacity to trust other people |
Full definitions are on the definitions page.
Who is measured
Everyone in the Index sought out a divorce recovery assessment. They are people who went looking for help, found our assessment, and completed it.
This is not a random sample of divorcing adults, and no figure we publish should be read as describing the general population.
It describes the population that clinicians, coaches, courts and support services actually encounter — people in enough difficulty to go looking. That is what makes the Index useful as a professional reference, and it is also its principal limitation.
Wherever we describe the population, we use language that states it: "among people seeking divorce recovery support," not "among divorcing adults."
Composition
Because the population is self-selected, its composition matters and we publish it. Drawn from assessments completed between 2023 and 2026:
| Where they are in the process | Share |
|---|---|
| Considering divorce | 23% |
| In the divorce process | 46% |
| Divorced | 31% |
n = 5,309 people, counted once each at first assessment.
| Who ended the relationship | Share |
|---|---|
| Their partner did | 51% |
| They did | 37% |
| Mutual decision | 12% |
n = 5,319 people, self-reported.
Instruments
Two assessments feed the Index.
| Fisher Divorce Adjustment Scale | Divorce Recovery Scorecard | |
|---|---|---|
| Items | 100 | 30 |
| Reports | Six subscales, total out of 500 | Four phase scores |
| Origin | Fisher, 1976; revised 1978 | Rebuilders short form |
| Independently validated | Yes | No |
The FDAS has been validated independently of its author. A study of the Persian version across 486 divorced adults reported internal consistency of .93 for the full scale and split-half reliability of .89.
The DRS has not been independently validated. It is a short form built for accessibility at first contact, and we do not claim more for it than that.
How subscales map to phases
| RIFT phase | FDAS subscales | Points available |
|---|---|---|
| Thinking | Disentanglement from the love relationship | 110 |
| Feelings | Symptoms of grief; feelings of anger | 180 |
| Identity | Feelings of self-worth; social self-worth | 170 |
| Relationships | Rebuilding social trust | 40 |
Combining the two instruments
Readings pool responses from both assessments. Pooling raises the number of assessments behind each monthly figure, which matters for precision, but it introduces a risk that has to be designed around.
The risk: if the two instruments produce even slightly different average readings, then a change in which assessment people happen to take would move the Index even though nobody's adjustment had changed. A month in which a local program promoted the short form heavily would look like a real shift. It would not be one.
The fix: each instrument is standardised separately against a fixed baseline period, then combined using fixed weights held constant over time. Weights are not sample-proportional. The Index therefore reflects adjustment, not instrument mix.
Weighting is by reliability rather than by headcount, so the longer instrument is not diluted by the shorter one simply because more people take the short form.
PENDING PUBLICATION
The baseline period and the specific fixed weights are being finalised against a concurrent validity analysis of people who completed both assessments. They will be published here in full, including the correlation between the two instruments at phase level, before the first monthly reading is released.
If that analysis shows the two instruments do not agree closely enough to pool, we will publish them as separate series instead and say why.
The scale
The Index runs from 0 to 100, where higher indicates better adjustment. It is derived from assessment point scores expressed as a share of the maximum available.
The Index is not a percentile. A percentile places one person against others. The Index measures the population as a whole. Both produce numbers between 0 and 100 and they answer different questions.
An Index reading of 65 does not mean the average person is at the 65th percentile. It means the population is scoring, on average, 65% of the maximum on the assessment.
Why not percentiles
Percentiles are a ranking rather than a measurement. Because scores cluster in the middle of the range, the points required to climb ten percentile places vary enormously depending on the starting position, so percentiles cannot meaningfully be averaged.
There is also a circularity problem. Percentiles are computed against the same population the Index measures, so a percentile-based index would sit near 50 permanently by construction, no matter what happened to anyone.
Where we report percentiles elsewhere — in individual reports, or in outcome research — every statistic is computed in points first and converted at the final step.
Reporting rules
Comparisons
Each issue reports the current reading against the previous month and against the same month a year earlier. Divorce has seasonal patterns, so year-over-year is generally the more meaningful of the two.
The noise floor
At current assessment volumes, a single-month change smaller than about 2.5 points cannot be distinguished from ordinary sampling variation. Movements below that threshold are reported as flat.
This threshold is built into the publication process rather than applied by judgment. We would rather publish a dull reading than a false trend.
Rolling three-month figures are more precise than single months and are used where a trend claim is being made.
Denominators
Every figure is published with the number of assessments behind it. Where a subgroup is reported, its own denominator is stated rather than the total.
Rounding
Index and phase readings are reported to one decimal place. Composition shares are rounded to whole percentages and may therefore not total exactly 100.
Safety responses
Both assessments include items that can indicate acute distress. Responses meeting that threshold trigger a direct human response rather than an automated result. Those items are never included in any published figure, in any form, including as an aggregate rate.
Limitations
Stated plainly, because anyone relying on this data should know them before they rely on it.
- Self-selected population. Everyone measured sought out an assessment. The Index describes people seeking divorce recovery support, not divorcing adults generally.
- Self-report. All measurement is self-reported on a single instrument type. No independent clinician rating is collected.
- Composition can shift. Changes in where our assessments come from — a new region, a marketing campaign reaching a different audience — can move the population behind a reading. We publish composition each month so that shifts are visible rather than hidden.
- Relationships rests on a narrow base. One subscale, 40 of 500 points.
- The DRS is unvalidated. It contributes to the Index; its psychometric properties have not been independently established.
- Not a clinical instrument in this form. The Index is a population measure. It should not be used to assess an individual, and no individual conclusion should be drawn from it.
Corrections
Published readings are permanent and remain at their original addresses. When we find an error we correct it in place and note the correction on the page, including what changed and when. We do not silently edit published figures.
If you believe a figure is wrong, tell us at [email protected]. We would rather hear it from you than not.
Full series available as data at /divorce-recovery-index/data.csv
About Rebuilders International
Rebuilders International trains mental health professionals and divorce coaches in the Rebuilding method, developed by Dr. Bruce Fisher over more than forty years. Individuals going through divorce can find programs, classes and a free self-assessment at rebuilders.net.
Rebuilders International
Empowering professionals to guide individuals through transformative recovery and rebuilding processes.
For Professionals
Looking for the consumer program? Visit Rebuilders.net.
© 2026 Rebuilders International. All rights reserved. | Privacy Policy | Terms | [email protected]