Career Path Appreciation (CPA) is one of the most rigorously validated instruments a Mexican board is likely to encounter, and one of the least well known outside the tradition. Here is what the primary research actually shows — and what a governance decision should do with it.
Every time we present the BIOSS tradition to a Mexican board or a family council, one question comes back within the first fifteen minutes: how do we know this actually works? The question is legitimate and the market has trained boards to ask it. The good news is that Career Path Appreciation — CPA, the appreciation instrument developed by Gillian Stamp inside the Brunel Institute of Organisation and Social Studies — sits on a longer, deeper, and more transparent evidence base than most instruments a board will be shown, including several household names. The less-good news is that its evidence looks unfamiliar, because it was assembled through a different discipline than the one boards are trained on. What follows is a plain-language reading of what the primary research shows, and what a governance decision should — and should not — infer from it.
The lineage, briefly, so the evidence sits in context
CPA descends from the work of Elliott Jaques at the Glacier Metal Company in the 1950s and Jaques's Stratified Systems Theory of the 1960s and 1970s. Gillian Stamp began work on the instrument at Brunel between 1976 and 1979, and the refined form standardized in 1979 has remained substantially stable since — with cards and structured conversation replacing the earlier pencil-and-paper protocols. Between 1979 and 1988 the instrument was applied in more than thirty field organizations across the United Kingdom, Norway, and southern Africa. From 1997 onward it entered a long South African re-validation programme under Employment Equity Act §6 fairness standards. What that means for a Mexican reader is simple: this is not a recent instrument in search of a market. It is an instrument with forty years of continuous field use and multiple independent research programmes behind it.
What the primary research actually shows
Four bodies of evidence carry most of the weight, and each is publicly citable. Boards do not have to take the tradition's word for it.
Stamp 1988 — predictive validity. The foundational validity study, published as US Army Research Institute Technical Report 819 and available in the public archive at the Defense Technical Information Center, reported four longitudinal correlations between initial CPA reading and later organizational level. Over a nine-year interval (n = 84), the correlation was r = .70. Over a four-to-thirteen-year interval (n = 182), r = .79. In the more tightly matched four-to-eight-year window using the refined form (n = 76), r = .89; among the refined-form participants who stayed in the same organization long enough for the reading to be tested against sustained progression (n = 59), r = .92 (Stamp, 1988). These are exceptional correlations by any standard used in industrial-organizational psychology, and they are longitudinal — the reading predicted what the person actually did in an organization over years, not what they said about themselves in an inventory.
Lewis 1993 — construct validity and interrater reliability. A second US Army Research Institute report, Technical Report 983, tested CPA against independent instructor ratings of strategic thinking at the US Army War College during academic year 1991-92 (sample n = 52) and reported interrater reliability of r = .81 (p < .0001) between trained CPA practitioners reading the same participants, and construct-validity correlations of r = .57 and r = .51 with independent instructor ratings of strategic thinking capability (Lewis, 1993). The interrater number matters especially: it says two trained practitioners reading the same person produce substantially the same reading, which is exactly the property a governance decision needs and exactly the property a lightly trained instrument does not have.
Mauer 2003 — test-retest stability, factor-analytic orthogonality. Published inside the Bioss South Africa network's research pipeline, the 2003 test-retest study (n = 75) reported nine-week stability of CLC (Current Level of Contribution) at r = .87 and Mode of Thinking at r = .84. Companion factor-analytic work in the same year showed that CPA CLC loads .95 on a common factor with the Occupational Personality Questionnaire's Analytical/Numerical scale, while the OPQ's personality dimensions load near zero on the same factor. Read plainly: CPA is measuring a cognitive-complexity construct, not a personality trait, and the two families of instruments read different things. This matters for the board because it settles the most common objection ahead of time — no, CPA is not a personality test in different packaging.
Retief 2003 and Kitching 2005 — large-sample fairness re-validation. The South African re-validation programme conducted under Employment Equity Act §6 fairness standards accumulated a sample larger than eight thousand readings, applying the instrument across race, language, gender, and educational cohorts. The programme produced two things a governance reader should note: an evidentiary base at scale, and an instrument that passed a formal legal fairness standard designed to detect adverse impact. Not all instruments a board is likely to encounter have been through anything comparable.
What the evidence lets a board conclude
The practical implications for a Mexican board are three, and they matter more than any individual number.
The reading is stable. A trained practitioner reading a person today and again in nine weeks produces substantially the same reading. Two trained practitioners reading the same person independently produce substantially the same reading. That is the property that lets a board rely on the instrument to inform a real decision, rather than treating it as one snapshot to be argued with.
The reading is longitudinal. Unlike a point-in-time assessment, CPA is designed to trace a person's Mode of Thinking as it has developed and to project a realistic trajectory. The Stamp 1988 correlations are longitudinal correlations against what people actually did in organizations over years, not internal consistencies of an inventory. A board reading a CPA output is reading a projection built from evidence about growth curves, not a personality label.
The reading is orthogonal to personality. Because the CPA construct is cognitive-complexity, not disposition, a CPA reading answers a different question than a personality profile does — and can sit alongside one without either substituting for the other. Boards that already commission personality profiles for cultural-fit reads gain something additional from CPA, not the same thing dressed differently.
What the evidence does not let a board conclude
Equally important is what the numbers do not say. A correlation of r = .89 over four-to-eight years is not a prediction that person X will reach role Y. It is a statement about how well the reading corresponds to long-run organizational level across a research population. A board that reads a single CPA output should still read it inside the triad of work, person, and context — the number is context-free until the reading is placed against a specific role at a specific horizon.
Equally, high validity in a research population does not translate automatically into low error in an individual case. This is why the tradition insists on trained practitioners for administration and interpretation, and why the output is a qualitative narrative built from the conversation rather than a score to be plugged into a decision matrix. Rigour inside, prose outside. A board that treats the reading as a certainty is misusing an instrument the research does not license anyone to use that way.
Why the tradition treats appreciation as a distinct stance from assessment
One consequence of the evidence base is that CPA reads growth curves rather than fixed points. The tradition therefore names appreciation as a distinct stance from assessment and from evaluation: assessment produces the readings, evaluation is what the board or the CEO does with them, and appreciation is what leadership owes the person afterward — recognizing what their judgment carries and where it is likely to grow. The instrument is called Career Path Appreciation because it is oriented toward stewardship of the person read, not just toward a decision about them. A board that reads a CPA output correctly is receiving both a rigorous instrument and a stance discipline in the same package.
How CPA sits inside the wider evidence conversation
One caution boards frequently need is this: the volume of published, peer-reviewed papers is not, on its own, a proxy for the rigour of an instrument. Some heavily marketed personality instruments carry hundreds of papers whose replication behaviour is much weaker than the marketing suggests. CPA sits on a smaller number of studies, but the studies are longitudinal, independent (US Army Research Institute reports are external to the developer), and cross-cultural (UK, Norway, South Africa, and the US Army War College). What matters for a governance decision is not the total number of papers but the design of the studies and what the numbers actually claim to show. On both counts, CPA sits well.
What to ask when a firm claims to use CPA
Four questions distinguish a rigorous practitioner from a firm that has read the marketing.
First, ask which form of CPA — the standardized 1979 refined form, in the tradition's card-based protocol — and how the practitioner was trained and by whom. The instrument is licensed inside the network; a firm using it without formal training in the tradition is using something else.
Second, ask what the output will contain and, more importantly, what it will not contain. A rigorous CPA output is a qualitative narrative about Mode of Thinking, Current Level of Contribution, and projected trajectory, plus a stance recommendation for how the person should be stewarded going forward. It is not a scored composite, and it is not a ranking against other candidates.
Third, ask how the reading will be used inside the decision the board is making. A well-run engagement places the CPA reading inside a broader capability assessment that also names ability and capacity, and reads the composite against the role's actual timespan and complexity. A CPA reading pulled out of that context is at risk of being over-read, no matter how good the instrument is.
Fourth, ask what the practitioner will not do. Trained practitioners refuse to compress the reading into a number, refuse to rank candidates against each other outside role context, and refuse to trespass into appreciation-territory during an evaluation conversation. That refusal is a sign of the tradition, not a weakness.
Anker Bioss uses CPA inside a wider capability architecture — evaluation, capability, succession, and organizational design as one system. If the board is preparing to decide, or preparing not to have to decide under pressure, we can help. That is capability that complexity can't break.