Table Of Contents:
- The question a careful coach asks
- Convergent validity, defined
- How to read the numbers
- A worked example
- The point most explanations miss
- What this lets Priya do
- Where to take this next
- FAQ
- References
The question a careful coach asks
Priya was deciding whether to bring a resilience instrument into her practice, and she did the responsible thing. She asked how she could know it actually measures resilience, rather than measuring something adjacent and calling it resilience. Then she said the honest part: “I can see the word validity all over these instruments’ websites, but I couldn’t tell you what it means, and I don’t want to pretend.”
That is exactly the right place to start. Coaches are asked to trust instruments whose evidence is written in a language most were never taught. One of the most important terms in that language is convergent validity, and it is far simpler than it sounds. Once you understand it, you can evaluate an instrument’s evidence yourself, rather than taking a claim on faith.
Convergent validity, defined
Convergent validity is the degree to which an instrument correlates with other established measures that it should, in theory, be related to. It is one of the core ways researchers demonstrate that an assessment measures what it claims to measure.
The logic is straightforward. If you build a new measure of resilience, it ought to relate in predictable ways to things resilience is already known to travel with. If it does, that is evidence you have built a resilience measure. If it does not, you have built something, but you cannot yet claim it is resilience.
It helps to separate two words that often get blurred. Reliability is about consistency: does the instrument give a stable result. Validity is about accuracy: does it measure the right thing. An instrument can be perfectly consistent and still consistently measure the wrong thing. Convergent validity is one of the checks that guards against that.
How to read the numbers
Convergent validity is reported as correlations, so you need to be able to read a correlation. It is less technical than it looks.
A correlation coefficient, written as r, runs from minus one to plus one. The sign tells you direction. A positive number means the two measures rise together. A negative number means one rises as the other falls. The size tells you strength. Values near zero mean little relationship; values approaching one, in either direction, mean a strong one. As a rough field guide, correlations around 0.1 are weak, around 0.3 are moderate, and 0.5 and above are strong.
That is the whole toolkit. Direction, then strength. With it, you can read an instrument’s convergent validity evidence and judge whether it holds up.
A worked example
Let me show you what good convergent validity looks like using real figures, so the concept stops being abstract.
When the Personal Resilience Indicator (PRI) was validated, its results were compared against two well-established external instruments. The first was the 13-Item Sense of Coherence Scale, the SOC-13, developed by Aaron Antonovsky in 1993 [1], which measures how far a person experiences life as comprehensible, manageable, and meaningful. The second was the Positive and Negative Affect Schedule, the PANAS, developed by Watson, Clark and Tellegen in 1988 [2], which measures the degree to which a person is experiencing positive and negative emotional states.
Now the predictions. A genuine resilience measure should correlate positively with sense of coherence, because resilient people tend to find life more manageable and meaningful. It should correlate positively with positive affect, and negatively with negative affect, because resilience travels with more of the former and less of the latter. Those are the directions convergent validity predicts before anyone looks at the data.
Here is what the data showed. The overall PRI score correlated with the SOC-13 at r = 0.72, positive and strong, as predicted. It correlated with PANAS positive affect at r = 0.67, again positive and strong. And it correlated with PANAS negative affect at r = -0.55, negative and strong, exactly the opposite direction, exactly as it should. Three predictions, three confirmations, all in the right direction and all substantial in size.
The point most explanations miss
Here is what nobody tells you about validity numbers: bigger is not always better, and a correlation that is too high is a red flag, not a triumph.
Think about what it would mean if the PRI had correlated with positive affect at 0.97. That would not show the PRI is an excellent resilience measure. It would show the PRI is measuring positive affect under a different name. If a new instrument correlates almost perfectly with an existing one, you have not built something new. You have built a more expensive way to measure the old thing.
Good convergent validity therefore sits in a band. The correlations need to be strong enough and in the right direction to prove the instrument relates to the construct, but not so high that the instrument is simply a clone of an existing measure. Correlations in the range of roughly 0.5 to 0.7, in the predicted directions, are close to ideal. They say the instrument is clearly related to known constructs, which validates it, while still measuring something of its own, which justifies its existence. Related, but not redundant. That is the target, and it is the standard a careful evaluator should hold an instrument to.
What this lets Priya do
Priya no longer has to take the word validity on trust. She can now ask three concrete questions of any instrument she is considering.
Does it report convergent validity at all, or does it just use the word. Against which established measures, and are those measures themselves credible. And are the correlations in the directions you would predict and sensible in strength, strong enough to prove the relationship, not so high that the instrument is a clone. An instrument that answers all three well has shown its work. One that cannot is asking for faith, and a careful coach is right to withhold it.
The PRI reports its convergent validity figures openly for exactly this reason, in its Technical Report [3], summarised on our Evidence page. You should expect the same transparency from any instrument you put in front of a paying client. The number is not the point. The willingness to be checked is.
Where to take this next
Knowing how to read an instrument’s evidence, and how to use a validated profile well with a client, is part of what certification in the Personal Resilience Indicator develops. It is the difference between trusting the numbers and being trained to work with them.
FAQ
What is convergent validity in simple terms?
It is a test of whether an instrument measures what it claims to. The idea is simple: a real resilience measure should relate, in predictable ways, to other things resilience is known to travel with. If a new resilience assessment correlates positively with a sense of meaning and with positive emotions, and negatively with distress, that is evidence it is genuinely measuring resilience. If it does not, the claim has not been earned yet.
How do I read a convergent validity correlation?
Two steps. Look at the sign, then the size. A positive number means the two measures rise together; a negative number means one rises as the other falls. Then the size: near zero is a weak relationship, and the closer to one in either direction, the stronger it is. Roughly, 0.3 is moderate and 0.5 or above is strong. So a resilience measure correlating at 0.67 with positive emotion and at minus 0.55 with negative emotion is showing strong relationships in exactly the directions you would want.
Is a higher validity correlation always better?
No, and this trips people up. A correlation that is too high is a warning, not a win. If a new instrument correlates at, say, 0.97 with an existing measure, it is not a great new tool. It is the old tool wearing a new label. Good convergent validity is strong and in the right direction, but not near-perfect, because you want an instrument that relates to known constructs while still measuring something of its own. Related, but not redundant.
What should I check before trusting a resilience assessment?
Ask three things. Does it actually report convergent validity, rather than just using the word. What credible established measures was it validated against. And are the correlations in the predicted directions and sensible in strength. An instrument that reports this openly has shown its work. One that will not is asking you to trust it, and with a paying client in the room, that is not a good enough basis.
References
[1] Antonovsky, A. (1993). The structure and properties of the sense of coherence scale. Social Science & Medicine, 36(6), 725-733.
[2] Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: the PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063-1070.
[3] Sinclair, N., Hafner, G., & Sinclair, P. D. (2022). Personal Resilience Indicator: Validation summary and psychometric properties (PRI Technical Report). Mind Matters. Available on the Evidence page.
