Table Of Contents:
- The retest that would not move
- Two different things wear the same word
- Why the distinction decides everything for development
- The uncomfortable bit: stable scores can mean the scale is working
- Which of the common scales measure trait, and which measure state
- How this plays out with a state-based instrument
- The question to ask before you choose
- Where to take this next
- FAQ
- References
The retest that would not move
Sofia had done everything right, or so it looked from the outside. She had won a competitive corporate contract to run a six-month resilience programme for a company’s top 100 leaders, and to sharpen her pitch she had added a measurement component at no extra cost. She chose a well-known, published, freely available resilience scale, exactly the kind of decision that looks prudent: a peer-reviewed instrument, no licence fee, a number to show the client. She measured all 100 leaders at the start.
The programme itself went well. The feedback was strong. Participants described real shifts, harder conversations handled better, less reactivity under pressure, a clearer sense of what they were working toward. Sofia could feel the work landing.
Then, at the six-month mark, the company asked for the obvious thing. Retest everyone, and show us the improvement. So she did. And the scores had barely moved.
She was left in a genuinely difficult position, holding two things that would not reconcile: a room full of leaders reporting meaningful change, and a spreadsheet saying almost nothing had changed at all. To the client, the flat scores were the real result and the glowing feedback was just sentiment. Sofia knew the work had been real. She could not prove it with the instrument she had chosen. And she could not work out why.
The reason was not the programme. It was the scale. She had, without realising it, measured the one thing she most needed to move with a tool designed never to move.
Two different things wear the same word
Resilience is used to mean two quite different things, and the difference is the whole of Sofia’s problem.
Trait resilience is stable. It sits close to personality, it is largely settled by adulthood, and it changes slowly if at all. When researchers talk about resilience as a durable characteristic that distinguishes people over the long run, this is what they mean. A trait is something you have.
State resilience is dynamic. It reflects your current capacity to cope, it responds to circumstances, and, crucially, it responds to deliberate practice. When a coach helps someone build a specific capability and that person copes better as a result, the thing that moved was state. A state is something you can change.
Both are real. Both are legitimately called resilience. But they answer different questions and they behave in opposite ways, and an instrument is almost always built to capture one or the other. That design choice, buried in the scale and rarely advertised, decides what the instrument can do for you.
Why the distinction decides everything for development
Here is the rule that would have saved Sofia, stated plainly: you can describe a trait, but you cannot coach one.
If your job is to understand or predict, a trait measure is exactly right. It is stable, so it gives you a reliable read on where someone sits relative to others, and that stability is a virtue. But if your job is development, the same stability becomes the problem, because a good trait measure is built specifically not to move. Running a development programme and then measuring it with a trait scale is like putting someone through months of strength training and assessing the result with a tape measure around their height. The intervention can be excellent and the instrument will still, correctly, report almost no change, because it was never measuring the thing you were changing.
This is why the state-trait question is not academic. It is the first question a practitioner should ask of any instrument, before price, before item count, before brand: does this measure the kind of resilience that moves, or the kind that does not? If you coach for change, only one of those answers is any use to you.
The uncomfortable bit: stable scores can mean the scale is working
There is a twist here that catches good practitioners out, and it caught Sofia. When a retest comes back flat, the instinct is to assume the instrument failed. Often the opposite is true. If the scale is a trait measure, flat scores after six months are the instrument doing its job perfectly. Trait measures are validated on their stability, their ability to return the same result over time, and a large body of research treats high test-retest stability as a mark of quality.
So “the scores did not change” can mean two completely different things, and they look identical on the spreadsheet. It can mean the programme did not work. Or it can mean the programme worked and the instrument was built to ignore exactly that. Without knowing which kind of scale you are holding, you cannot tell these apart, which is precisely the bind Sofia found herself in: unable to prove real change, and unable to say why the number would not show it.
Which of the common scales measure trait, and which measure state
You do not have to guess at this, because the field’s own quality data gives a usable signal. A 2025 systematic review examined 24 resilience scales and scored each on a set of psychometric properties, including stability, which is test-retest reliability, how consistently a scale returns the same result over time [1]. That stability score is the cleanest available proxy for where a scale sits on the state-trait question, because high stability is the signature of a trait measure and the opposite of what a change-sensitive instrument would show. The same review anchors our companion piece on whether any resilience measure yet counts as a gold standard: is there a gold standard for measuring resilience?
Read that way, the review is revealing.
A small number of scales score highly on stability, and they behave like trait measures by design. The Hardiness Resilience Gauge and the Tachikawa Resilience Scale both earned full marks on stability in the review [1]. These are well-built instruments, and if your aim is a stable, comparative read on a durable characteristic, that is a strength. For measuring the effect of a six-month programme, that same stability is the trap Sofia fell into.
The Ego Resilience Scale, one of the oldest and most widely used measures, is explicit about its nature: it was built to assess resilience as a personality trait (Ego Resilience Scale vs PRI). It remains genuinely useful for research. For development work, its premise is its limitation, because a trait, by definition, is not what you train.
Most of the reviewed scales scored zero on stability, meaning the property was either not established or not a design priority [1]. That does not automatically make them state measures in the developmental sense, some were simply never tested that way, but it does mean their makers were not chasing the unmoving-over-time quality that defines a trait instrument.
And then there is the cautionary case. One scale in the review is actually named the State Trait Assessment of Resilience Scale, which sounds like exactly the instrument a developmental coach would want. Yet it scored zero on stability in the review’s assessment [1]. The lesson is worth holding onto: naming the state-trait distinction is not the same as handling it well. The label on the box does not tell you what the instrument actually does. Only the design does.
How this plays out with a state-based instrument
Now run Sofia’s situation again, with a different tool in her hands.
The Personal Resilience Indicator was built from the start to measure resilience as something that develops. It is delivered by a certified practitioner and debriefed in person, and it reports not a single figure but a profile across six domains and twelve underlying drivers, each of them a specific, trainable capacity rather than a fixed trait. That design choice is the entire point: an instrument built to track change has to be able to show change, which means measuring at the level where change actually happens.
In Sofia’s scenario, that changes both the story and the retest. At the start, instead of one number for each leader, she would have had a profile showing where each person’s resilience was thin and where it was strong, which drivers to work on and which to leave alone. The programme would have had somewhere specific to aim. And at six months, the retest would not have been a single score stubbornly refusing to move. It would have been a driver-level comparison, showing which specific capacities had shifted, by how much, and in which people. The leaders who reported handling pressure better would have a change in the relevant drivers to point to. The strong feedback and the data would tell the same story instead of contradicting each other.
This is not a claim that PRI scores always rise, or that every programme works. It is a simpler point about fit. An instrument built to detect trainable change can show you a flat result and a real result and tell them apart, because it is measuring at the level where development lives. A trait instrument cannot, because it was designed not to. Sofia’s problem was never that she measured. It was that she measured the right work with the wrong kind of instrument.
The question to ask before you choose
If you take one thing from Sofia’s six months, let it be a single question to ask of any resilience instrument before you commit to it, especially one you are choosing because it is free.
Does this measure the resilience that moves, or the resilience that does not? A trait measure and a state measure can look almost identical on the page, similar items, similar scoring, both peer-reviewed, both respectable. They diverge only when you try to use them for development, and by then, as Sofia found, it is too late to change your mind. The free instrument that made her offer look more competitive cost her the one thing she most needed to show. The right question, asked at the start, is a great deal cheaper than the wrong answer discovered at the retest.
Where to take this next
If you coach or train for change, learning to read resilience as a set of trainable drivers, and to measure it with an instrument built to show movement, is part of what certification in the Personal Resilience Indicator develops.
If you want to see a state-based resilience profile for yourself, including how it is designed to change with deliberate work, a certified PRI practitioner can take you through one in a structured debrief.
FAQ
What is the difference between state and trait resilience?
Trait resilience is stable and close to personality; it is largely settled by adulthood and changes slowly if at all. State resilience is dynamic; it reflects your current capacity to cope and responds to circumstances and to deliberate practice. Both are real and both are called resilience, but a trait is something you have, while a state is something you can change. Most instruments are built to measure one or the other.
Which kind matters for coaching and development?
State resilience, because it is the kind that moves. You can describe a trait, but you cannot coach one. If your aim is development, you need an instrument that measures the capacity that responds to your work, otherwise your programme can succeed and the numbers will still barely change, because a trait measure is built specifically to stay stable over time.
If a retest shows no change, does that mean the programme failed?
Not necessarily, and this is the trap. If you used a trait measure, flat scores after months of work can mean the instrument is doing its job perfectly, staying stable is what trait measures are validated on, rather than that nothing changed. Without knowing whether your scale measures trait or state, you cannot tell a failed programme from a real result measured with the wrong tool. They look identical on the spreadsheet.
How can I tell whether a resilience scale measures trait or state?
A useful proxy is the scale’s test-retest stability, how consistently it returns the same result over time. High stability is the signature of a trait measure; instruments built to detect development are designed to show change instead. A 2025 systematic review scored 24 resilience scales on exactly this property, which lets you see where common instruments sit. And be wary of names: one reviewed scale is called the State Trait Assessment of Resilience Scale yet scored zero on stability, so the label does not tell you what the instrument actually does.
References
[1] Huerzeler, H. E., Boss, P., & Thoma, M. V. (2025). Addressing the heterogeneity of resilience scales: A systematic review and development of a unified resilience construct within a standardized resilience framework. The Journal of Positive Psychology, 1–20. https://doi.org/10.1080/17439760.2025.2574049
