RMP.8:4 - Solution
Match the result to the assertion it bears on. Establish what should be retained, added or changed, follow its implications through the account and return the answer with the grounds, uncertainty and limits the receiving question requires.
RMP.8:4.1 - Locate the assertion and its receiving question
Identify the earlier statement or model relation, including its subject, comparison, conditions and time. Specify the supplied result, the examination that supports it and the part of the account it addresses.
Check whether the result bears on the same assertion. A factory’s decision to close and the subsequent closure are different events. A laboratory rate may refer to different conditions from a field rate. A finding about another population can extend an account without contradicting the original claim.
When developing a new account, identify the contribution and the material it would add to or qualify. If there is no earlier assertion to correct, treat the contribution as an addition.
Establish what the recipient needs to understand or do when it affects the explanation or evidence required. A scientific explanation or a reusable finding can answer the question, as can evidence for an engineering, public or personal decision.
If the source, its relation to the subject or the examination result is missing, say what conclusion this prevents. A plan can establish which test or synthesis was selected. A claim about its outcome needs evidence of what happened.
RMP.8:4.2 - State the warranted retention or change
State the contribution at the strength and scope its grounds warrant. Identify a stronger inference if a recipient could reasonably draw it from the wording or intended use.
| Warranted disposition | What to state |
|---|---|
| Retain | State why the assertion still holds under the conditions examined. Identify the basis and limits of that judgement. |
| Narrow | Restrict the population, context, time, relation or transfer to what the result supports, preserving the supported part. |
| Qualify | Keep the assertion and state the uncertainty, alternative explanation or condition of use that matters to it. |
| Reject | Remove the defeated claim or inference and explain what defeated it. Assess a proposed replacement from its own grounds. |
| Add | State the new contribution and its relation to the existing account, including whether it is supported information, an interpretation or a hypothesis for further inquiry. |
One model may need different conclusions about its premises and outputs. Give the grounds for each. To raise confidence, identify what adds support. Repetition of the assertion, a later document date or inability to inspect criticism adds none by itself.
Describe uncertainty under the method that produced the result. Give the quantity and assumptions for a numerical interval and the basis of a qualitative confidence judgement. Identify whether the open step is attribution, extrapolation, prevalence or another inference.
A result compatible with zero and nonzero effects can leave the effect unresolved. To establish equivalence, use the applicable margin and comparison rule. Limited access shows what could be examined. An inference from a missing record to the underlying event needs further grounds.
RMP.8:4.3 - Reconcile the affected account
Trace the premise to the conclusions and outputs that use it. Identify those that retain their support, those that change and any newly required relation.
In a model, distinguish the target, structure, parameter, assumption, calculation and interpretation when their grounds differ. An arithmetically correct output may rely on a failed transfer assumption for its field interpretation. The same parameter can remain useful within its calibration conditions.
If a conclusion’s warranted content or stated grounds change, revise the conclusion itself and correct any dependent outputs affected by that change. Preserve access to the earlier wording when a citation, correction or existing use requires it. An unused draft can often be edited in place with a source reference.
A replacement explanation needs grounds of its own. If the evidence rules out only the earlier explanation, retain the alternatives still possible. Use RMP.1 or RMP.2 when an unresolved question or design choice warrants further research.
Use existing records and source links when they let the intended reader examine the reasoning.
RMP.8:4.4 - Return evidence to the actual use
Give the answer to the receiving question first. For a change or retention, identify the earlier assertion it concerns. Cite the source or examination that supports the answer, with the conditions, uncertainty, dependence or access limits that affect reliance.
When a materially changed assertion was used elsewhere, provide the correction to any known affected use through its authorized route. Use FPF A.10.1 when affected uses need to be discovered or several known uses need revalidation together. A list of citations can locate possible uses. Establish which relied on the changed assertion.
The responsible editor, maintainer or practitioner can apply a correction to the account or decision they control. A correction that others rely on may also require notifying them. If communication or access is unavailable, identify the return still outstanding and what would enable it. The supported local correction can stand.
Observe applicable source rights, confidentiality, participant protection and community authority. Disclose only what the receiving question needs and permits. A qualified redacted account can convey a finding. If it lacks material needed for a particular inspection, say what cannot be checked. A reader’s own inspection requires their own access.
When a practical choice is required, the receiving practitioner weighs alternatives, values and constraints under the applicable authority. The research result supplies the evidential premise within its scope.
RMP.8:4.4.1 - Make a research Method contribution recoverable across publication forms
Use this branch when a publication presents a Method someone needs to understand, repeat, adapt or combine with other work. Identify the contribution and the intended use before choosing a publication form.
A practitioner pattern explains a Method and its use. It identifies a recurring problem, the conditions and competing demands that matter, the way of acting and its grounds, worked applications, limits, consequences and relations. Its SoTA comparison examines what the best available approaches and serious alternatives change in the Method. These elements give the operations their rationale and scope. If claiming that a particular episteme is a U.MethodDescription, establish under A.3.2 that it describes an admitted Method.
Choose a section structure, language, notation, reading order, medium and publication arrangement that give the reader access to that content. A translation or a change in file layout may preserve the contribution. Omitting a constraint that changes the choice of solution alters it. A short entry can direct readers to a developed explanation if that explanation is available.
An article may present several connected claims and Methods with their construction, evidence and limits. Its contributions may warrant one pattern, several related patterns, a change to an existing language or a different knowledge product. A pattern language connects recurring problems and ways to resolve them. A DPF serves a field of practice with a defined scope and maintenance boundary. Use E.4.DPF when deciding which framework contribution the material warrants.
Explain how to recognize another instance of the problem, the conditions that matter, which actions recur, what can vary, and the preparation, support and result required. Another qualified application can reproduce the intended effect while yielding a different object or number. Recomputing a reported number establishes that computation. Applying the Method to a new case tests a different claim. Correct recognition that the Method does not apply can answer a boundary question. Claiming its positive effect requires a case where that effect occurred.
For a reader to use the published Method, establish the following connections as the task requires.
- Explain how the Method works. Identify the recurring problem, situation, objects and relations, the competing demands and how the way of acting addresses them. A construction may combine, adapt or reinterpret existing Methods. Show the sequence, alternatives, feedback and returns where they matter. For work performed together, show how its constituent actions realize the encompassing Method. E.4.CM helps establish compatible contributions and construct the connections a list of source methods leaves open.
- Specify the intended repetition. Distinguish reconstructing a derivation, rerunning a computation on the same inputs, making new measurements and applying a Method under changed conditions. State what remains comparable, what changes, the result to compare and the criterion for agreement. Equality can be appropriate for a derivation or deterministic calculation. New measurements usually need a justified uncertainty or agreement criterion. RMP.6 helps choose an examination or interpret its observed result.
- Provide the means the examination needs. Supply or locate the relevant definitions, assumptions, input preparation, reasoning, operations, material, comparison and interpretation. Give versions, environment, calibration, timing, randomization or required capability when a difference could change the result. For a reused Method, identify the usable instructions and their conditions. RMP.4 helps establish what surviving records show about conduct. State what they leave unknown.
- Connect the publication forms. An empirical article often separates the question, methods, results and interpretation. A theoretical or methods article may develop a construction and its argument. A protocol or Stage 1 Registered Report specifies proposed work. A results report gives what was done and found. Apply the reporting requirements of the relevant venue and study type. A pattern explains the recurring problem, competing demands, solution, grounds, cases, boundaries and comparison with other approaches. Preserve the substance when changing section names or order.
- Explain what readers can examine. Link each presentation to the source contribution and the material needed for the stated examination. Give access conditions and consequential omissions. A protocol supports a claim about intended work. An available artifact can permit a rerun. For a claim about an actual rerun, provide its result. A repeatable calculation needs an additional applicability basis for a different setting. When a claim on which others relied changes, correct affected presentations and uses through §§4.3–4.4.
Choose an explanation and access arrangement sufficient for the receiving use. A theoretical derivation may be reproducible from its premises without an empirical dataset. Respect restrictions on protected participant material. If they prevent the requested public inspection, identify the limit and any qualified finding available to that reader.
RMP.8:4.5 - Finish the answer and select further inquiry only when warranted
Finish when the account and its required recipients have the conclusion the evidence warrants, or when the specific missing input or delivery limit has been identified. Distinguish a supported local correction from an outstanding promised delivery. A broader open question can be reported separately.
If further inquiry could improve an explanation, reusable finding or receiving decision, compare its attainable contribution with the work it requires. Include preparation, access, execution, interpretation, protection, shared resources and displaced work. Keep the present conclusion within the evidence already obtained.
When a research design still needs choosing, RMP.2 helps compare designs. For a question about credibility, use an adequate existing result or use RMP.6 to assess the examination. Distinguish a supported decision to commission no further study from a choice that remains open because a needed input is missing. Revisit the choice if a useful investigation becomes feasible.
Repeated failures in selecting or performing research Methods can identify a question for RMP.9 or FPF Method development. State what needs improvement and which practice could address it, if known. The present research result can be returned while that work remains open.
Continue an inquiry from a result that changes the question. Formulate each live hypothesis as a claim about its object and conditions. Derive an observation, argument or construction on which the plausible accounts differ. State what each possible result would justify retaining, revising or testing next. A judgement that a branch is promising can justify the cost of such a test. Support for the claim comes from what the examination can establish. If no feasible examination can distinguish the alternatives, report the unresolved distinction and the missing means.
Carry shared material forward at its warranted scope. Preserve the source, conditions and reasoning that a new branch uses. Mark which conclusions depend on the same measurements, calibration or assumptions. Several analysts can expose sensitivity to analytic choices on one data set. An error in their common source can still affect all of them. In Silberzahn et al.’s many-analyst study, teams working on a shared question and data obtained a range of estimates. Reanalysis is a suitable way to examine that dependence. A claim about reproducibility with independently obtained observations calls for a different design.
When the question changes, keep the supported answer to the old one. Carry forward only the premises that serve the new inference. RMP.1 and RMP.2 determine whether the new question needs a study and whether a feasible design could answer it. If the branches compete for a scarce facility or specialist, the research contribution identifies which tests share inputs and what each test could establish. STR.16 supports the separate decision about commitments.