THE GOLD STANDARD OF JUDGEMENT · SPECIALIST STUDY
Let both transformation and input-dependent effort shape the mean only when the scale can support them
Bajraktarević means generalise transformed averages by combining a strictly monotone ratio of generator functions with positive weight functions that may depend on each input. This creates expressive, state-dependent aggregation, but also more ways to hide compensation or influence. The gold standard states the generators, weights and domain, tests internality, equality and comparison, and interprets every mathematical property in decision language.
This guide belongs to the Judgement subject guide, the Gold Standard of Education, and the How X Works learning library. Choose a reading route, work the examples, then use the transfer checks to test whether the method survives a changed context.
Choose a reading route
Build the foundations · chapters 1–2
Apply the core method · chapters 3–5
Test evidence and alternatives · chapters 6–7
Handle limits and disagreement · chapters 8–9
Transfer and govern · chapters 10
The whole problem at a glance
| Component | Formal role | Decision meaning |
|---|---|---|
| Ratio f/g | Defines transformed coordinate | How positions are compared |
| Positive weight p_i(x_i) | Changes contribution by input | State-dependent influence |
| Inverse ratio | Returns to original domain | Interpretable output |
| Domain conditions | Keep denominator and inverse valid | Boundary of use |
A familiar output on one dataset does not make the operator equivalent to a quasi-arithmetic mean.
CHAPTER 1 OF 10 · Build the foundations
1. Construct the mean from a monotone generator ratio
Core idea. A weighted Bajraktarević mean can be written using functions f and g with g nonzero and f/g strictly monotone, then inverting a ratio of weighted sums.
Method. Declare interval I, regularity, sign and monotonicity conditions, weight functions and inverse branch before calculating.
Worked example. With g=1 and constant positive weights, the form reduces to a weighted quasi-arithmetic mean generated by f.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. If the denominator vanishes or f/g is not one-to-one, the output can be undefined or ambiguous.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 1 is complete only when the claimed result survives that review.
Misconception to avoid. Any pair f,g produces a mean. The ratio and denominator conditions are essential.
Practical evidence check
- Verify equal inputs return the same value and test both endpoints.
- Return to chapter contents
CHAPTER 2 OF 10 · Build the foundations
2. Read quasi-arithmetic means as a special case
Core idea. Setting the denominator generator to a constant recovers transformed averaging, which provides a useful baseline for understanding what input-dependent weights add.
Method. Calculate the quasi-arithmetic baseline first, then introduce g or variable weights one component at a time.
Worked example. A logarithmic f with g=1 gives a geometric-type mean; variable positive weights can make large or small inputs exert different influence.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Equivalent current outputs can diverge sharply under new vectors because the weighting mechanism is state-dependent.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 2 is complete only when the claimed result survives that review.
Misconception to avoid. Bajraktarević means are merely renamed quasi-arithmetic means. They form a broader family.
Practical evidence check
- Find a vector where the broader rule departs from its baseline.
- Return to chapter contents
CHAPTER 3 OF 10 · Apply the core method
3. Interpret input-dependent weight functions
Core idea. When p_i depends on x_i, contribution changes with the value itself rather than only with a fixed importance coefficient.
Method. Plot every weight function across the domain, normalise contribution, and explain whether it represents effort, reliability, salience or another defensible construct.
Worked example. A reliability weight decreases near a sensor saturation boundary, reducing the influence of extreme readings rather than rewarding magnitude.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. If weights encode the same signal already present in f, the rule can double-count it.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 3 is complete only when the claimed result survives that review.
Misconception to avoid. Weights are neutral technical coefficients. Their state dependence can encode strong policy.
Practical evidence check
- Compare fixed and variable weights on ordinary and extreme inputs.
- Return to chapter contents
CHAPTER 4 OF 10 · Apply the core method
4. Test generator and weight equivalence
Core idea. Different function pairs can define the same mean under transformation conditions, so formula appearance alone cannot establish a distinct decision rule.
Method. Use the relevant equality theorem and regularity assumptions; supplement proof with diverse numerical fixtures and boundary tests.
Worked example. Two specifications differ by a permitted linear-fractional transformation of the generator pair and return identical outputs across the domain.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Equality results can require differentiability, nonvanishing derivatives or fixed arity; outside assumptions the conclusion may fail.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 4 is complete only when the claimed result survives that review.
Misconception to avoid. Different code means different judgement. Equivalent representations may compile differently.
Practical evidence check
- Produce a canonical specification and cross-check two implementations.
- Return to chapter contents
CHAPTER 5 OF 10 · Apply the core method
5. Prove internality, continuity and monotonicity
Core idea. A mean should normally lie between its inputs and respond predictably when one input rises, but variable weights can complicate monotonicity.
Method. Prove properties from generator and weight conditions; test equal, ordered and endpoint vectors; retain the first counterexample to any failed claim.
Worked example. A weight increasing too sharply near the low end causes an input increase to shift relative weights enough that the aggregate moves unexpectedly.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Local monotonicity does not guarantee global monotonicity, and numerical grids do not prove continuous behaviour.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 5 is complete only when the claimed result survives that review.
Misconception to avoid. Because the result lies inside the range, it is monotone. Internality and monotonicity are different properties.
Practical evidence check
- Raise each coordinate separately across a dense boundary-focused test set.
- Return to chapter contents
Worked weight effect
| Inputs | Weight rule | Result direction |
|---|---|---|
| Low, high | Equal constant weights | Central transformed mean |
| Low, high | Higher weight near low | Moves downward |
| Low, high | Higher weight near high | Moves upward |
| Equal inputs | Any valid positive weights | Returns common value |
Exact values depend on the chosen generator pair; the example describes direction, not a universal formula.
CHAPTER 6 OF 10 · Test evidence and alternatives
6. Compare two means with comparative convexity
Core idea. Comparison asks when one generated mean is always below another; the answer depends on relations among generators and weights, often expressible through convexity-type conditions.
Method. State whether the theorem is local or global, verify derivative and weight-ratio conditions, and test crossing cases outside the proven region.
Worked example. Mean A is below Mean B for every vector in a certified interval, but the ordering reverses after a weight function crosses its allowed ratio.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. A local comparison near the diagonal cannot be advertised as global ordering.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 6 is complete only when the claimed result survives that review.
Misconception to avoid. Checking many samples proves one mean dominates. It finds evidence, not a theorem.
Practical evidence check
- Mark the exact interval and search numerically just beyond it.
- Return to chapter contents
CHAPTER 7 OF 10 · Test evidence and alternatives
7. Use Jensen and quasideviation connections carefully
Core idea. Bajraktarević and quasideviation means relate to generalized convexity, which can support bounds and comparison while changing the meaning of ordinary convexity.
Method. Define the deviation or comparative convexity notion, preserve domain conditions and translate the inequality back into the original decision scale.
Worked example. A Jensen-type bound holds in transformed coordinates, but stakeholders care about the returned original-scale difference and its asymmetry.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Convexity in one coordinate system may not survive an inadmissible relabelling of the scale.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 7 is complete only when the claimed result survives that review.
Misconception to avoid. Convex means risk-averse. Mathematical curvature needs domain interpretation.
Practical evidence check
- Show the same inequality in transformed and original coordinates.
- Return to chapter contents
CHAPTER 8 OF 10 · Handle limits and disagreement
8. Test delegativity and subgroup aggregation
Core idea. Some generalized means allow subgroup results and efforts to be aggregated without changing the final outcome, but only under associated decision and effort functions.
Method. Specify the pair of functions, grouping rule and retained effort; test all partitions rather than passing subgroup means alone.
Worked example. Two teams report subgroup decision and total effort; recombining both reproduces the all-person calculation, while averaging subgroup decisions without effort does not.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Changing group membership or using incompatible weight functions can break delegativity.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 8 is complete only when the claimed result survives that review.
Misconception to avoid. Associativity alone guarantees fair delegation. Group invariance says nothing about representation or power.
Practical evidence check
- Compare direct and two-stage aggregation for every partition of a small set.
- Return to chapter contents
Property certificate
| Claim | Evidence | Countercheck |
|---|---|---|
| Internal mean | Ratio monotonicity and positivity | Endpoint vectors |
| Monotone | Derivative/order proof | Raise one input |
| Equal operators | Generator equivalence theorem | Diverse vectors |
| Comparison | Global inequality conditions | Boundary and crossing cases |
Numerical testing is useful but cannot replace proof on a continuous domain.
CHAPTER 9 OF 10 · Handle limits and disagreement
9. Respect admissible scale transformations
Core idea. A sophisticated generator cannot make ordinal labels cardinal or create meaningful ratios where the measurement scale lacks them.
Method. List admissible transformations, rerun the rule under each and document which properties or rankings are preserved.
Worked example. An input-dependent weight tied to raw score magnitude changes after a harmless unit conversion, revealing that the rule is not scale-admissible for that measurement.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Some applications intentionally fix a calibrated scale; invariance then depends on maintaining that calibration.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 9 is complete only when the claimed result survives that review.
Misconception to avoid. Monotone generators make the method ordinal. Weighted sums in transformed space still use more than order.
Practical evidence check
- Apply a legitimate unit change and an arbitrary monotone relabelling.
- Return to chapter contents
CHAPTER 10 OF 10 · Transfer and govern
10. Govern expressive means as contestable policy
Core idea. Generator pairs and weight functions encode compensation, influence, scale and grouping choices that affected people should be able to understand and challenge.
Method. Version functions, proofs, test vectors, elicitation, sensitivity, overrides and correction; retire the model after scale or policy change.
Worked example. A service score gives less influence to saturated indicators; governance publishes the weight curve, why it exists and cases where it reverses ranking.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Formal validity does not establish accurate inputs, lawful authority or fair burden.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 10 is complete only when the claimed result survives that review.
Misconception to avoid. The theorem transfers responsibility to mathematics. Decision owners remain accountable.
Practical evidence check
- Audit one outcome from measurement through weights, comparison and remedy.
- Return to chapter contents
Build the specialist decision record
For “The Gold Standard of Judgement | Bajraktarević Means, Weight Functions, Comparative Convexity and Scale Interpretation”, begin with one bounded decision rather than an abstract claim. Name the affected people, data or institutions; the time window; the authority for acting; and the conclusion the method is allowed to support.
Preserve source material before interpretation. Record definitions, exclusions, transformations and the person responsible for each judgement. A reviewer should be able to separate an observed fact from an inferred result, a policy choice and a presentation shortcut.
Work one ordinary case in full. Show the input, every material rule or calculation, the output and the action that follows. Then write what the same evidence would mean under the simplest credible alternative method.
Keep an explicit unresolved state. Missing, protected, conflicting or unevaluated information must not become zero, false, consent or absence merely because a form expects one value.
Finish the record with a claim ledger. For every sentence that could drive action, identify its supporting source, inferential step, confidence or legal status, intended reader and expiry condition. This keeps descriptive evidence, formal consequence, professional judgement and institutional authority from collapsing into one impressive-looking conclusion.
Run the boundary-case audit
Select a case near the method’s domain edge: a missing input, a disputed definition, an endpoint, a sparse group, a resource limit or a legal deadline. Boundary cases reveal the contract that routine examples can conceal.
Change one assumption at a time and retain every output. Mark what stays stable, what reverses and what becomes undefined. If several assumptions move together, report that dependence instead of attributing the change to one convenient cause.
Search deliberately for a counterexample to the strongest claim. A method that works on the published illustration can still fail under a permitted input, an alternative grouping, a changed unit or an inaccessible evidence pathway.
Translate the counterexample into plain language: who would receive a different decision, what evidence would be hidden, and which safeguard or narrower claim repairs the problem.
Include an adversarial but plausible case. Let a missing group, ambiguous identifier, extreme weight, conflicting rule or urgent deadline interact with normal system incentives. The aim is not theatrical failure; it is to learn whether ordinary pressure makes the method conceal uncertainty, shift burden or reward a misleading shortcut.
Test independent reproduction
Give a reviewer the bounded inputs, public method contract and permitted access, but not the desired answer. Ask them to reproduce the result and identify every place where judgement, local knowledge or unavailable evidence was required.
Compare the reviewer’s intermediate states, not only the final number or label. Matching outputs can hide different denominators, inference paths, missing-data treatments or authorities that will diverge on the next case.
Where reproduction fails, classify the cause: ambiguous definition, undocumented transformation, incompatible implementation, inaccessible source, calculation error or legitimate alternative interpretation. Each class needs a different remedy.
Retain the disagreement and resolution. A corrected record should show what changed, why it changed and which prior outputs or decisions must be revisited.
Run the reproduction on a changed but structurally similar case after the first discrepancy is repaired. A method is not transferable merely because its authors can explain the original example. The new case tests whether the public contract carries the reasoning or whether hidden expertise is still doing essential work.
Design correction, challenge and remedy
Assign a visible route for correcting inputs, challenging assumptions and reviewing the application of the method. Technical accuracy without a usable remedy leaves the affected person carrying an error they cannot contest.
Separate urgent live-case repair from later system learning. A current decision may need pause, reassignment, fresh evidence or a human review before the organisation finishes its longer investigation.
Protect confidential and privileged material during review, but do not use confidentiality as a reason for unrecorded authority. Layer public explanation, protected oversight and individual reasons according to lawful access.
Verify that correction reaches cached results, derived records, dashboards and future decisions. Closing a complaint is not enough if the old conclusion continues to travel.
Measure remedy from the affected person’s position. A revised database row is useful, but it is incomplete when a hearing, service allocation, published claim or downstream model still relies on the old result. Record notification, restored opportunity, compensation or procedural repair where the governing framework provides it.
Set review and retirement triggers
Review after a change in definitions, population, law, source access, measurement scale, software, vocabulary or incentives. A stable formula can become invalid when the world or the meaning of its inputs changes.
Monitor leading signals as well as final outcomes: coverage gaps, timeouts, overrides, complaints, disagreement, missingness and boundary frequency. These often reveal failure before a headline metric moves.
Predefine thresholds for recalibration, suspension and retirement. Do not force a method to answer after its assumptions fail merely because replacing it is inconvenient.
At retirement, preserve the versioned contract, evidence, known limitations and migration decision. State which historic results remain interpretable and which require warning, restatement or withdrawal.
Schedule one prospective review before confidence fades or controversy forces it. Assign named owners for evidence, implementation and rights impact, and invite at least one reviewer outside the design team. The review should be able to narrow the claim, change the method or stop use without requiring proof of catastrophic failure.
Questions readers ask
What is a Bajraktarević mean?
- A generated mean formed from a monotone function ratio and positive weighted sums.
How does it extend quasi-arithmetic means?
- A constant denominator generator and fixed weights recover the quasi-arithmetic case.
Why use input-dependent weights?
- To model influence such as effort or reliability that changes with the input, when justified.
Can two different formulas define the same mean?
- Yes, under generator-equivalence conditions.
What is comparative convexity doing here?
- It helps characterize ordering and inequality relations between generated means.
Does a valid mean justify a decision?
- No; scale meaning, authority, impact and remedy remain external obligations.
Practise, verify and continue
Choose one consequential case. Reproduce the complete method, work one ordinary and one boundary case, test the strongest rival explanation, and state the evidence that would reverse or suspend the result. Build an evidence pack containing original inputs, definitions, calculations or rule steps, alternatives rejected, uncertainty, reviewer comments and the final action. Give the pack to a reader who did not design the method. Ask that reader to reproduce the result, identify one missing assumption and construct a case that would make the method return unresolved. Revise the public explanation only after the evidence pack and the operational process agree.
Move the method into an education or public-service setting. Keep the reasoning structure while replacing domain authority, population, privacy, evidence and correction routes. Start with a low-stakes historical example, then compare the specialist method with a transparent baseline. Check whether the same people, records and decisions remain inside scope; whether the scale or vocabulary still means the same thing; and whether a person affected by error can understand and challenge it. Pilot under human review, record reversals and burden, and stop the transfer if the formal resemblance hides a different ethical or legal problem.
Related eduKate reading
Quasi-arithmetic means, generator elicitation and scale admissibility
Uninorms, neutral elements and associativity limits
Sugeno integrals and non-compensatory choice
Bipolar Sugeno integrals and qualitative gains and losses
Choquet integrals and non-additive expectations
Authoritative sources and scope
Grünwald and Páles, Local and global comparison of generalized Bajraktarević means
Páles and Pasteczka, Decision making via generalized Bajraktarević means
Grünwald and Páles, Jensen convexity of quasideviation and Bajraktarević means
Research on generalized Bajraktarević means and comparison inequalities
This article is mathematical and decision-design education. Verify definitions, regularity assumptions, arity and theorem domains in the cited research. Examples are original teaching cases. A valid generalized mean does not establish scale admissibility, accurate inputs, legitimate weights, fairness or legal authority. Preserve proofs, counterexamples, sensitivity, reasons and correction.