Jump to text
On this page
Copy and source
Published source confirmed at last check

Source changed 2026-11-18 21:11:39 UTC · snapshot created 2026-11-18 21:12:35 UTC · last check 2026-11-18 22:15:20 UTC

RMP.2 — Choose a Criticism-Bearing Research Design or Stop

Type: Research Method Practice practitioner method pattern Status: Stable

RMP.2:0 - Use This When

Use this method when the research question, current evidence and answer conditions are sufficiently clear to ask whether further inquiry is worthwhile. First check whether the evidence already answers the question. If it does not, determine whether the current allocation already justifies undertaking no new study. When a selection remains open, use a suitable existing strategy or develop one through §4.4. Identify what evidence could distinguish the competing claims, recover an operation that could obtain it and follow its use on available material. Compare the actual options for their fit to the question, the error each can expose, their limitations and the full effort of obtaining an answer.

Return one ResearchDesignDisposition for the question about research effort now at issue. Use EvidenceSufficientStop when current evidence answers the qualified question. Use ResearchDesignDecision when a justified feasible strategy has been selected. Remaining PostSelectionNeed items must preserve the selection throughout their jointly admissible ranges. Use ResearchDesignHold when a specific missing result prevents a selection that is still needed. Use QualifiedNoNewStudy when the current allocation calls for no new study although the answer remains limited. For a selected design, check the construct, evidence, measurement, access, protections, permissions and resources on which its criticism and feasibility depend.

The result tells a researcher or commissioner what the current evidence supports and whether a particular study is worth pursuing. A selected design describes the proposed strategy. Its protocol, permission to proceed, actual work and findings have their own grounds.

If the research question or its current basis is unclear, use RMP.1 to qualify it. Use the appropriate professional method for operational definitions, study execution, analysis or development of a reusable Method. Obtain a required ethics, safety or access decision from its holder when the work depends on it, using an existing valid permission where applicable. The researcher may hold another role under its own mandate.

RMP.2:0.1 - Working Terms

Name in this patternWhat it denotes
research designA justified way to answer a qualified question. It relates the relevant cases, material or sources to observations, intended analysis, a challenge that could defeat or narrow the claim, significant threats and a useful stopping condition.
design familyA broad kind of relation between a question, its evidence, a comparison and a challenge to the inference. Options include intervention or assignment, field observation, situated interpretation, historical source criticism, computational modeling and corpus synthesis. A design may combine contributions from several families.
criticism logicWhich material error, rival explanation, source failure, dependence, bias or unjustified transfer the design could expose, how it could expose it and what result would defeat or narrow the claim.
serious design-or-stop optionA real way to stop or a feasible or conditionally feasible strategy for the same research question. It differs from the other options in what evidence could decide, what error could be exposed, what risk or cost would be incurred or what would prevent use.
EvidenceSufficientStopCurrent evidence meets the answer condition for the intended use. Return the answer and the limitations that matter to reliance on it.
ResearchDesignDecisionA worthwhile feasible strategy has been selected, with an explanation of its distinguishing evidence, relevant alternatives including no new study, threats, required inputs, refusal conditions and useful stopping point.
PostSelectionNeedAn input still needed to instantiate, record, schedule or support the selected design. Every admissible value, including admissible combinations with other unresolved inputs, preserves its family, core contrast, criticism, feasibility and refusal conditions. The set can be empty.
ResearchDesignHoldA particular selection is still needed, but an unresolved result could change the approach, core comparison, admissible evidence, criticism, feasibility or reason for refusal. The hold identifies the result. Commissioning work to obtain it is a separate decision.
QualifiedNoNewStudyNo new study is selected for the current allocation. Return the bounded answer or still-open claim and the reason the attainable study does not justify its burden. A material uncertainty may remain a limitation or a condition for reconsidering the choice.
selection-changing uncertaintyA still-open fact or requirement whose plausible values support different strategies, ways of challenging the claim, stopping conditions or refusals.
expected information gainWhat an option is expected to clarify or distinguish about the live question. Compare that contribution with the effort, risk and limits of obtaining the evidence. Use a numerical estimate when its assumptions are supportable.

Recognize a design question when a qualified inquiry still has an open choice about further evidence. To justify a design, show how its evidence would address the same question, how it could reveal error and why the needed conditions can be met. Separately state what the current evidence already establishes and which research effort, if any, has been selected.

RMP.2:1 - Problem Frame

Research questions do not determine one universal design. The same practical use can be served by a controlled comparison, a field observation, a qualitative account of situated mechanism, historical source criticism, a computational challenge, a synthesis of an existing corpus, a mixed composition, or no new study at all. Each exposes different errors and requires different observations, sources, access, and domain judgements.

A fact about the subject or available work may change the design. If acceleration changes the failure mechanism, evidence from an accelerated seal test may no longer support the field claim. An invalid salinity sensor or inaccessible site can eliminate a field comparison. Uncertain archival provenance can prevent a historical source from discriminating two chronologies. Settle such conditions when selection depends on them or identify the resulting limit.

Other details can remain open after selection. A particular filename, serialization or schedule within an established window may make no difference to the evidence, criticism or feasibility. Check the remaining values together before treating them as implementation details.

RMP.2:2 - Problem

A familiar method can be chosen before anyone has identified the error the question requires the study to expose. Randomization, interviews, simulation or synthesis may each provide a suitable operation. A detailed implementation can still fail if the observations offer no relevant way to distinguish the rival accounts, including differences in their likelihood under the stated assumptions.

A design can fail when the construct has no suitable observable, the source cannot support the claim, the instrument lacks discrimination, safe access is unavailable or the required permission is missing. Identify such dependencies before committing effort to a protocol. They can change the strategy or make no new study the appropriate allocation.

An unresolved detail that cannot change the choice can wait. Assigning a record locator or a time inside a qualified window may be such a detail. Show why the admissible alternatives preserve the design and when a change would reopen it.

RMP.2:3 - Forces

  • Check whether the existing evidence answers the question. Residual uncertainty can be compatible with the answer condition.
  • Compare the error each design can expose. A single numerical score can conceal important differences in what the evidence would establish.
  • Compare real alternatives. If a single design is feasible, assess its contribution against the option of no new study.
  • Domain knowledge, measurement, access and permissions can determine which design is possible. Use their proper grounds, whether they are already available or supplied by someone else.
  • A design can guide later work once its consequential assumptions have been resolved or bounded.
  • A detail can wait only if its admissible values preserve the selection.
  • Weigh the information the work might supply against its whole burden. Some questions support a quantitative estimate and others a qualitative judgement.
  • In a mixed design, explain how the different evidence will be brought to bear on the question and what disagreement would mean.
  • If a reporting rule applies or supplies information the task needs, use it for that purpose. Justify a design by the relation between its evidence and the research question.
  • Identify the intended use. If a later decision is required, establish its authority separately.

RMP.2:4 - Solution

Keep the qualified question in view and test the current evidence against its answer condition. If the question remains open, determine what an obtainable study could distinguish and weigh that contribution against its full burden. Return the applicable result: sufficient existing evidence, a selected strategy, a particular choice held for a consequential input, or a completed allocation choosing no new study.

RMP.2:4.1 - Admit the Design Question

  1. Recover the question and its current basis. Identify who needs the answer, what knowledge it should change, which rivals matter, what evidence could challenge the claim and what constraints apply. Distinguish the evidence required for an answer from conditions for ending work with uncertainty. A compatible supplied brief and current account allow direct entry.
  2. Check that the question still holds. A material change to the use, question, evidence, rival accounts or useful stopping point may require a revised brief. Use RMP.1 for the missing qualification. If the revised question and its grounds are already available, use them directly.
  3. Establish the inputs and permissions this strategy needs. Identify who can substantiate a domain or measurement claim, who can supply materials or resources and who holds any required ethics, safety, access or other permission. Use applicable evidence and permissions already established. If a receiving action requires a separate decision, state the authority or the specific gap.
  4. Define the present choice. State what evidence strategy and challenge are at issue, which details may be chosen later and what result would justify stopping, refusing or reconsidering.

RMP.2:4.2 - Test the Evidence-Sufficient Stop

Compare qualified current evidence with what the ResearchQuestionBrief requires for an answer. Distinguish that requirement from a condition for ending work with an unresolved claim or a refusal.

  • Does it answer the bounded question for the named receiver and use?
  • Does the evidence cover the phenomenon, comparison, cases, population or corpus and time period to which the answer will apply?
  • Does the evidence distinguish the principal live rival or address the relevant threat to validity?
  • Are its source, evidence-use, provenance, uncertainty, and currentness limits compatible with the reliance proposed?

If the evidence meets the answer condition, return the answer with the source, scope and limitations its use requires. This is EvidenceSufficientStop. The answer may complete the assignment. Preserve any uncertainty that could change reliance on it.

If the evidence is insufficient, ask what an obtainable study could establish and whether its value warrants the full burden. Include preparation, access, measurement, execution, interpretation, work by participants, delay, displaced activity and joint resource limits where they apply. A contribution to knowledge or later inquiry can justify a study.

If the available facts justify choosing no new study for this allocation, return QualifiedNoNewStudy under §4.7 with the bounded answer or unresolved claim. Otherwise examine the inputs and options that could change the remaining selection. Budget, priority, availability and burden affect the allocation. Judge the adequacy of the evidence against the question.

RMP.2:4.3 - Expose Selection-Changing Inputs

While a choice remains open, identify the facts, requirements and permissions that can change the design, its criticism or its feasibility.

InputWhat could affect selection?Typical source or responsible party
construct or mechanism relationWould a different relation between construct and observation change what needs to be measured, manipulated, compared or modeled?Domain studies or a qualified domain investigator.
observation or source admissibilityCan the available observation or source support the claim, and under what interpretation or rights of use?Observational specialist, archive, data steward or relevant rights holder.
instrument or measurement feasibilityCould sensitivity, specificity, calibration, disturbance, resolution or construct validity eliminate a required distinction?A relevant instrument or laboratory result, or a qualified measurement specialist where needed.
case, population, material, corpus, or site accessCould access or sampling conditions make the selected contrast impossible or materially different?Site, archive, data, participant, asset, or access owner
ethics, safety, rights, confidentiality, or community authorityCould a legitimate restriction prohibit the strategy or require a different evidence relation?Applicable ethics, safety, rights, privacy, security, or community authority
intervention, exposure or comparison controlCan the relevant difference be produced or observed and attributed under the proposed conditions?Domain or operational specialist, provider or applicable authority.
resource and provider feasibilityCould capacity, timing, specialist support, compute, laboratory, or field availability eliminate the family or core criticism?Resource, provider, laboratory, platform, or programme owner
intended use and permission to use the resultCan the proposed recipient use the result for this purpose, or would that use require a different claim, different evidence or permission?Recipient or relevant decision or assurance authority, where one is required.

For an unresolved input that could affect selection, compare plausible values and their consequences. If the values support different families, core comparisons, criticism, feasibility or stopping conditions and this selection is still needed, return a hold. A remaining detail may qualify as a PostSelectionNeed if the admissible values jointly preserve the selection. Where no new study has been chosen on adequate allocation grounds, report the unknown as a limit or a reason to reconsider. Naming a dependency leaves any acquisition decision with its holder.

Check the inputs together. Examine combinations of the remaining values that are compatible with the evidence and constraints. Where a dependence among values is unknown, state how the outer bounds limit the feasibility claim. A set of PostSelectionNeed items is justified if every jointly admissible combination preserves the family, core contrast, criticism, feasibility and refusal conditions. Checking each value while fixing the others favourably can miss a joint failure.

Suppose two required tasks draw on the same ten investigator-days and their efforts cannot overlap. Each requires four to six days, and the supplied relations admit the pair six and six. Six plus four and four plus six fit. Six plus six requires twelve days. The whole strategy is therefore not assured feasible across the admitted demands. If the choice is still needed, return ResearchDesignHold with the two demands, their source, the combinations that fit or exceed the limit and the information that could settle selection. If an earlier part is already complete and its result is usable, count only the work still needed.

RMP.2:4.4 - Compose Serious Design-or-Stop Options

Use a supplied strategy directly when it fits the question, evidence, criticism and obtaining conditions. A justified replication can already have specified procedures. When the required evidence relation is missing, use C.39 to develop the operation for this question.

  1. Work backward from a useful answer. Identify observations, relations among sources or interpretations that could support, narrow or defeat the claim. For each live account, ask what it permits or makes difficult to explain under the same conditions. An outcome can favor one account even when all permit it. Compare its likelihood under the accounts if suitable models are available. If they make the same predictions for the evidence being compared, that evidence cannot discriminate them. A dated record may distinguish two chronologies. A contrary case may challenge a situated interpretation.
  2. Find an operation that could obtain the distinction. Use an adequate professional method, worked study, demonstration or source. Identify its inputs, what the investigator does, what it returns and why that result bears on the question. Carry its applicable conditions into the proposed use. A controlled comparison needs a defensible assignment and basis for comparability. A source comparison needs identifiable, admissible records. The families in §4.5 can help locate operations.
  3. Connect the material to the operation. Identify the records, cases or specimens that could supply its inputs and how its output would bear on this question. Derive any needed comparison, transformation or combination from the domain premises. A lab that distinguishes mechanism signatures still needs comparable material and conditions to compare accelerated with field degradation. Identify what is already qualified, what the researcher has constructed and what remains assumed.
  4. Follow the proposed relation on concrete material. A suitable record, specimen account or earlier analysis may show what enters, what the operation returns and how it would affect the claim. Try a small application if its contribution justifies the work and the needed authority is in place. Examine a consequential rival, such as two mechanisms that produce the same observation or two archival records that concern different events. Use the inspection to assess the proposed strategy. Evidence that the proposed study took place is a separate matter.
  5. Repair a failed connection if worthwhile. Look for a more discriminating observation, a different comparison or another operation supported by the professional sources and domain premises. An analysis can recover an omitted distinction only with further observations or a justified relation that identifies it. If that relation remains unknown and selection depends on it, use §4.3 to state the gap and its consequences. If a narrower question would be useful, qualify that question explicitly through RMP.1 as needed. Section 4.7 also permits a bounded no-new-study result.

Explain how the material and operation would expose the relevant error, including any premise whose failure would defeat the comparison. One adequately developed strategy can suffice. Develop another when it could change the choice. Use available professional results directly where they fit. RMP.3 specifies the consequential operations of a selected strategy when that work is required.

For the same brief, compare strategies that differ in what the evidence could establish, their feasibility or cost. Include a current routine where it is a serious option and weigh further research against no new study when that choice is open. If an established source or constraint rules out a proposal, state that reason.

If a single research strategy remains, assess whether it can expose the relevant error, is feasible and contributes enough to justify the full burden compared with no new study. Consider another strategy when it offers a consequential difference in evidence, cost or constraints.

For each serious option, explain the following features as far as the choice requires. Use the brief and reliable inputs already available. A description is complete for this purpose when it shows how the evidence could answer or challenge the claim.

  1. what could become known and when the result would be sufficient for its use,
  2. which cases, units, materials, population or corpus the proposed comparison requires and which observations, sources or interventions would distinguish the claims,
  3. the principal error or rival the inquiry could expose and what would defeat or narrow the claim,
  4. which threats to validity it addresses, introduces or leaves unresolved,
  5. what evidence, expertise, access, protections, rights, permissions, supplies or resources it requires and which are available, assumed or missing,
  6. what useful information it could add and under which conditions it would add none,
  7. the burden, timing, risk, reversibility and conditions for stopping or refusing, and
  8. what would still have to be specified or recorded if the design were selected.

Differences in software, sample size, interview format or reporting can matter when they change the evidence, threats, cost or feasibility. Two descriptions with the same evidential and practical consequences can be treated as one strategy.

RMP.2:4.5 - Preserve Design-Family Differences

Design familyHow evidence bears on the questionPossible errors or rival explanationsConditions that can affect selection
experimental — intervention or assignmentDeliberately vary or assign an intervention/exposure or comparison condition and estimate a bounded difference or relation.Rival causes, allocation, baseline imbalance, measurement error, model dependence, missingness, intervention fidelity, and applicability.Manipulability, admissible comparison, mechanism relation, instrument validity, safety, authority, sample/case feasibility, and effect/uncertainty use.
field or observationalCompare observed exposures, states, changes or contexts when the investigator did not assign the observed conditions.Confounding, selection, measurement error, context and seasonal effects, missing observations, reverse causation and unsupported transfer.Access to sites or usable records, valid measures, temporal and spatial coverage, available comparisons and the domain limits on causal inference.
qualitative or reflexiveDevelop a situated account of meaning, experience, practice, process or mechanism from participants, records, interactions or cases.Researcher position, rival interpretations, disconfirming cases, relevance of sampling, context, participant voice and unsupported transfer.Appropriate participation or access to records, ethical use, a suitable account of the phenomenon, interpretive competence, and protection against harm or disclosure.
historical or source-criticalUse surviving records produced for particular purposes to investigate an occurrence, interpretation, change or lineage.Provenance, chronology, source purpose, selective survival and silence, competing interpretations, anachronism and limited corroboration.Access, rights, authenticity, date, archive coverage, the relation between record and claim and any missing evidence that could distinguish accounts.
computational/modelChallenge a mathematical, algorithmic, simulation, or code-mediated claim through derivation, implementation, calibration, sensitivity, reproduction, validation, or intended-use comparison.Code/data/environment dependence, implementation defects, parameter and structural sensitivity, calibration, validation, identifiability, and intended-use fit.Model-to-phenomenon relation, data and code access, environment identity, validation target, compute feasibility, and domain acceptance criteria.
synthesisUse a defined body of evidence to answer a corpus question through eligibility, search, appraisal and suitable integration.Eligibility bias, search omissions, dependence among sources, heterogeneity, incompatible measures or meanings, appraisal limits and selective publication or reporting.A clear corpus question, traceable sources, eligibility and appraisal criteria, compatible results for any proposed pooling, access and currency.
mixedConnect two or more component designs whose different evidence addresses the question.Each component’s threats and any unsupported inference in combining the results, including disagreement, non-comparability or one component dominating the conclusion.A reason for each contribution, feasibility of the components together, a justified relation among their results and a way to interpret disagreement or partial failure.

These families illustrate different ways evidence can bear on a question. They can overlap and be combined. An observational study may estimate an effect under stated assumptions. An intervention study may examine situated experience. Enter with the available claim, corpus, model, field anomaly or proposed intervention. Stop or revise the strategy when the current evidence or a changed premise warrants it.

RMP.2:4.6 - Compare the designs and the option of no new study

Use the applicable questions below to compare real options on the same brief. State consequential differences and keep protected requirements in force. A short explanation may be sufficient.

Comparison positionRequired content
question and useWhich part of the brief could this option answer, for whom and under what conditions would the inquiry end?
evidenceWhich material, observations, intervention or exposure, sources and comparisons would the investigator use, and what would be inferred from them?
criticismWhich error or rival could it expose? What result would narrow or defeat the claim, and what material threats would remain?
validity and transferInternal, measurement, source, construct, interpretive, contextual, external, or intended-use limits that matter to this claim.
inputs and permissionsWhat domain or measurement evidence, sources, access, protections, rights, capacity or supplier work does the strategy need? Which are established, assumed or missing?
expected information gainWhat live question could the evidence resolve or distinguish? Under which conditions would the added work supply no useful information?
burden and refusalWhat work, materials, time and capacity would be needed to obtain and interpret the evidence? What risk, irreversibility, delay, displacement or joint limit would matter? Under what conditions should the work stop or be refused?
continuationWhat would need later specification or recording? What would close the inquiry or justify reopening selection?

If a bounded decision rule combines several measures, establish their meanings and the rule’s scope. Keep protected constraints, materially different losses and consequential unknowns visible. A single feasible option must still justify its burden against no new study where that allocation remains open.

RMP.2:4.7 - Return Exactly One ResearchDesignDisposition

Branch A — EvidenceSufficientStop. Use this result when qualified current evidence meets the answer condition. Give the answer, the evidence and the limitations that matter to this use. If another party will act on it, identify any uncertainty or reason that could change their reliance. The answer completes this branch.

Branch B — ResearchDesignDecision plus PostSelectionNeed. Select a feasible research strategy when its evidence could expose the relevant error and its expected contribution justifies the full burden. Compare actual alternative strategies and the option of no new study. Identify the selected strategy, its core contrast, justification, threats, required inputs, resources, refusal conditions, useful stop and any later needs. Explain why a considered alternative was set aside. Apply the joint-input test in RMP.2:4.3 to all remaining inputs. For each PostSelectionNeed, show the following.

  1. the value still needed and who can provide or establish it,
  2. the admissible range or alternatives already constrained by supplied results;
  3. why all admissible values, together with the other needs, preserve the selection, its criticism and feasibility, and
  4. the action if the value arrives outside that range.

Write an empty set as PostSelectionNeed: none. If a plausible value or joint combination could change the family, core contrast, criticism, feasibility or refusal and the choice is still needed, use Branch C. Use Branch D when the available facts justify choosing no new study for the present allocation.

Branch C — ResearchDesignHold. Use a hold when a particular choice is still needed and a missing result prevents it. Identify the question, the affected strategies, plausible results with different consequences, what work cannot start on the present basis and what would reopen selection. Name a qualified source or holder for the missing result where known, or state the gap. The hold can complete the question about what prevents selection. Deciding to obtain the input requires its own justification, feasibility and any applicable permission.

Branch D — QualifiedNoNewStudy. Use this when no new study is selected for the present allocation and the evidence still falls short of the answer condition. Return the strongest bounded answer or name the unresolved claim. Explain why the obtainable study does not justify its burden or why the discriminating evidence is unavailable. Preserve material limits and any condition that could reopen the choice. This allocation leaves the evidential and permission requirements for a stronger use in force.

Use one branch for the present choice. The result may be part of the existing explanation. Reconsider the choice when new evidence or a changed use affects it. RMP.3 requires a selected design from Branch B or a compatible design already supplied. A sufficient answer, a hold and a no-study allocation can each conclude their own work.

RMP.2:4.8 - What Changes in Practice

A researcher can develop the comparison a question requires, expose a failure on available material and repair the relation or identify the gap. The resulting account shows what a strategy could establish and whether that gain warrants the work. The present task can finish with adequate existing evidence, a justified design, a specific hold or a bounded no-new-study decision.

RMP.2:5 - Archetypal Grounding

The following examples stipulate the records, access, professional results and other conditions to show how the four dispositions differ.

RMP.2:5.1 - Accelerated Seal-Test: Different Supplied Values

The supplied ResearchQuestionBrief asks whether high temperature accelerates the same failure mechanism as the field conditions, so that accelerated degradation can inform the Pump-P4 seal-life claim. Choosing a maintenance interval is a separate operational decision.

RMP.2:5.1.1 - Derive and repair the mechanism comparison

Start with the question and the domain evidence. The higher temperature might accelerate the field mechanism or introduce a different dominant mechanism. Engineering supplies the material and operating envelope, the relevance of both accounts and the limits of a field-representative condition. In the laboratory method, a qualified practitioner prepares an identified specimen, observes its surface and chemistry and interprets the joint signature using mechanism references valid for the material and damage range. These results and the available laboratory competence and equipment are inputs to the research comparison.

Examine the familiar operation. A routine test returns an overall wear index. Suppose the proposed strategy compares that index under accelerated and field-representative conditions and takes matching values as evidence of the same mechanism. Examine the two reference records below.

Supplied referenceIndependently established mechanismOverall wear indexQualified surface/chemical signature
Reference A, within the named material and damage rangeThe field mechanism6A
Reference B, within the same qualified rangeThe rival heat-induced mechanism6B

Both constructed references have wear index 6, although their mechanisms differ. Greater precision in that index alone would still permit both explanations of a matching value. The number may answer a wear question. It fails here as a discriminator of mechanism.

Construct a comparison that can distinguish the mechanisms. Within the supplied material and damage range, the laboratory operation can return a mechanism-sensitive signature. Compare the signatures in accelerated and field-representative conditions. Use comparable specimens, preserve batch identity and assign conditions so that batch differences can be distinguished from temperature effects. Compare at appropriate damage levels, since equal elapsed times can correspond to different degradation. Once the strategy is justified, RMP.3 can specify the allocation, observation schedule and handling procedures.

The proposed strategy obtains mechanism-sensitive observations under both conditions, preserves material and damage comparability and interprets the signatures against the qualified references. A discrepancy can challenge preservation within the stated envelope. Agreement can support the claim to the extent allowed by sensitivity, sampling and interpretation. A different field condition or a maintenance interval requires its own grounds.

Test the proposed inference against the references. A and B are distinguishable on the failure-analysis records despite their equal wear index. Their identities and qualified conditions allow their use as mechanism references. In the proposed study, a qualified rival signature would challenge preservation, a compatible signature could support it within the stated limits, and inadequate discrimination would leave the mechanism unresolved. Use the professional method’s relevant classification and uncertainty criteria when deciding whether the strategy is justified.

Check the material identity and qualified damage range of a proposed reference. A signature label on an unidentified specimen may leave the interpretation open. If the instrument cannot distinguish A from B under the study conditions, use an adequate alternative or field comparison when one is available and worthwhile. Otherwise state the missing measurement and how it affects the choice, or conclude that no new study is justified for the present allocation. The wear index still has its original limitation.

Assess the proposed study. Suppose matched specimens and a justified assignment are available. The laboratory supplies the signature relation, uncertainty criterion and qualified measurement range. Access, protections, qualified personnel and resources cover preparation, measurement and interpretation. The attainable mechanism answer warrants that burden. The available field records do not connect material to exposure as required. The available model admits both mechanisms and cannot independently distinguish them here. These facts favor the experimental comparison for this case. Its results remain to be observed.

The scenarios below vary the supplied facts for the same question. A suitable prior result can answer it directly. A missing discriminator can hold a choice that is still needed. If the comparison cannot be obtained at a justifiable cost, the present allocation may end without a new study and with the mechanism question open.

RMP.2:5.1.2 - Select the disposition under the supplied facts

Compare the following relations when their supplied conditions make them serious:

OptionEvidence and criticism relationPrincipal loss or dependency
stop on current evidenceUse a result that already answers whether acceleration preserves the relevant mechanism within the specified conditions.The conclusion is limited to that use. A stronger life or maintenance claim needs suitable further grounds, which may already exist elsewhere.
experimental mechanism comparisonCompare degradation under the candidate accelerated and field-representative conditions using the mechanism-sensitive relation derived in §5.1.1; challenge acceleration-induced mechanism change and alternative material effects.Requires qualified indicator relation, instrument feasibility, specimen/material comparability, safety, and authority.
field or observational comparisonRelate field exposures, failures, material batches, installation histories and operating contexts to assess whether the laboratory relation transfers.Confounding, selection, few failures, measurement limitations or lack of access may limit the distinction.
computational/model challengeTest the proposed mechanism and acceleration relation through a domain-qualified model, sensitivity, calibration, and validation target.Model-to-phenomenon relation, parameter identifiability, implementation, data, and validation can dominate.

EvidenceSufficientStop scenario. Suppose an applicable result for compatible materials and the operating envelope shows that acceleration changes the dominant degradation mechanism. Return that finding and its scope. It answers whether evidence from this test can support the proposed mechanism-preserving inference. A stronger use would need a different basis.

ResearchDesignDecision scenario. Under the conditions in §5.1.1, the signature comparison is discriminating, obtainable and worth its burden compared with the field, model and no-new-study options. Select that strategy. The remaining PostSelectionNeed items concern the location and serialization of the required specimen, condition, observation and deviation records. Every admissible choice preserves the comparison and permits those records to be retrieved. If a required field cannot be recorded or retrieved, reconsider selection.

ResearchDesignHold scenario. The measurement result does not yet establish whether the indicator distinguishes the two degradation mechanisms. If it can, the experimental comparison may work. If it cannot, that comparison fails and a field or model strategy may be preferable. If the selection still matters, return this dependency and its consequences. The measurement specialist may already have suitable evidence. A decision to obtain a new result is a separate commitment.

QualifiedNoNewStudy scenario. The evidence leaves the Pump-P4 relation unresolved. The matched specimens have been lost and the available field and model material cannot distinguish the competing mechanisms for this case. Select no new study for the present allocation. Future specimens would belong to a different case. Newly released field records could make a relevant comparison attainable and justify reconsideration. A maintenance interval remains a separate decision.

RMP.2:5.2 - Citizen Stream Salinity: Dispositions without an Affiliation Gate

A resident team asks what explains the salinity increase in a specified reach and period. Possible accounts include road-salt transport, groundwater, a changed outfall, seasonal hydrology and sensor drift. Residents can frame and pursue this research in their own capacity. Available resources, measurement competence, safe access and applicable permissions still affect what they can do.

  • Return EvidenceSufficientStop if validated monitoring and source results already distinguish the relevant explanation for this reach, period and use. State the scope and any remaining limits.
  • Return a field or observational ResearchDesignDecision when the measurements, comparison reaches or periods, safe access and applicable permissions support a spatial or temporal contrast that can challenge the live rivals, and the contribution warrants its burden. A record locator may be chosen later if every admissible arrangement preserves the observations. Unknown calibration, access or comparison matters if it could change this selection.
  • Return ResearchDesignHold if the field choice is still needed and sensor validity, site or record access, or a seasonal comparison could make the strategy work or fail. State which result could distinguish the options and who can establish it when known. The community may itself be able to answer.

The environmental finding can support a bounded explanation and identify evidence limits. If remediation is proposed, the applicable authority decides on that action.

RMP.2:5.3 - Historical Source-Critical Case

An archive group asks whether a maintenance-practice change followed Directive-D or began earlier. An authenticated dated record may already answer the chronology question. A source-critical study could compare the dates, provenance and purposes of workshop records, consider archival silences and examine rival interpretations. If the question concerns a contested corpus already traced, synthesis is another possible strategy.

If the discriminating records have suitable provenance and access, a historical design may be justified with only invariant transcription and locator conventions left to settle. If an open provenance or access question could make the comparison viable or impossible and a selection is still needed, return ResearchDesignHold. Use source criticism appropriate to the chronology rather than ranking the evidence solely by experimental criteria.

For the narrow chronology question, suppose one authenticated workshop series is the only accessible material that could distinguish practice before and after the directive. An index and another scan format expose the same records. Their content does not provide independent corroboration. If the source-critical examination fits the two available investigator-days and evidence of an earlier or later practice would alter the claim, this one strategy can warrant ResearchDesignDecision. If the work would instead take twenty days, only two are available and no feasible reduction preserves its discrimination, retain the open chronology and return QualifiedNoNewStudy for this allocation. A newly accessible independent ledger could change the comparison.

RMP.2:5.4 - PostSelectionNeed Counterfactual

Consider two questions about a proposed field design.

Remaining valueCounterfactual testDisposition
directory and identifier syntax for required observations and deviationsEvery admissible choice permits the required records to be retrieved and preserves the field comparison and its criticism. A syntax that loses a record falls outside this range.PostSelectionNeed
whether the available sensor can discriminate salinity from temperature-dependent drift at the required scaleIf it can, the field comparison remains possible. If it cannot, another instrument, a different evidence relation or no new study may be appropriate.ResearchDesignHold when that selection remains necessary.

Classify a remaining input by the effect its admissible values could have on the selection, including their effects together. The date on which the input is expected to be settled is a separate scheduling matter.

RMP.2:6 - Bias-Annotation

Recurring biasLikely driftRepair
favorite-method biasThe practitioner’s familiar design defines the question and rivals.Hold the brief fixed and compare actual answer-changing relations with the relevant no-new-study outcome.
hierarchy biasOne design family is presumed strongest regardless of claim and threat.Compare the specific evidence relation and criticism each option enables.
novelty-of-tool biasA new sensor, model, AI workflow, or platform is treated as an informative design.State the answer-changing contrast, threats, and validation relation independently of the tool.
false-stop biasLimited time, funds or willingness is reported as sufficient evidence.Compare the evidence with the answer condition. An allowed return with uncertainty may end the work while the question remains open.
provisional-design biasThe proposal is called selected while a consequential input remains open.Compare plausible results for that input. Return ResearchDesignHold if the needed selection depends on which occurs, or QualifiedNoNewStudy if no study has been selected on adequate grounds for this allocation.
hold-everything biasA harmless record name or schedule choice prevents selection.Check that the admissible values, including their combinations with other inputs, preserve the comparison, feasibility and refusal conditions.
reporting-as-design biasA reporting guideline or protocol template is treated as design evidence.Use it later for transparency where applicable; select the question-to-criticism relation here.
mixed-method accumulationMore methods are assumed to strengthen an inference automatically.Explain each component’s evidence contribution and how to handle agreement, disagreement and partial failure.
unsupported information-gain estimateA numerical gain is assigned without a defensible model of outcomes or use.Identify the live uncertainty and what the possible answers could change. Use a qualitative comparison when the evidence supports that degree of precision.
authority inferred from designSelection is treated as permission for access, safety, ethics, spending or an action by the recipient.Identify the decision and the grounds for any applicable permission. Use an established decision or state the limit when the needed authority is absent.

RMP.2:7 - Conformance Checklist

Apply the following questions to the disposition being returned. For sufficient evidence, give the answer and limits. For no new study, explain the present allocation. For a hold, state the blocking dependency. The remaining design details apply to a selected strategy.

  • RMP2-01: The question and its current basis are sufficiently qualified. Where a change leaves that qualification open, identify or obtain the contribution needed from RMP.1.
  • RMP2-02: The intended use, answer condition and work-stop conditions are clear. Where a later action requires a decision, identify the applicable authority or the specific gap.
  • RMP2-03: Before selecting work to obtain new evidence, determine whether the current evidence already answers the question.
  • RMP2-04: A claim of evidence sufficiency accounts for the scope, relevant rivals, provenance, currency, uncertainty and reliance required in this case.
  • RMP2-05: Constraint, cost, preference, or low priority is not mislabeled as evidence sufficiency.
  • RMP2-06: The actual serious options address the same brief; a sole real research relation is judged against no new study without a quota of invented alternatives.
  • RMP2-07: An applicable current practice or available routine remains in the comparison if it offers a serious option.
  • RMP2-08: For each serious strategy, the account shows how the available material and operation could address the question, what would challenge the result and what obtaining it would require. If the connection had to be constructed, follow it on concrete material and repair or report a material failure.
  • RMP2-09: Use the evidence and criticism appropriate to each family considered. When combining designs, explain how their contributions relate.
  • RMP2-10: The selected option can expose the principal error or rival named in the brief.
  • RMP2-11: For an alternative that was considered and rejected, state the reason for that decision.
  • RMP2-12: Identify the construct, evidence, measurement, access, rights, capacity and other conditions that could change this strategy or its use.
  • RMP2-13: Identify the grounds for inputs already established. Distinguish assumptions and missing results.
  • RMP2-14: For an unresolved input that could still affect selection, explain how its plausible values or joint combinations could change the design or the decision to stop.
  • RMP2-15: A value that prevents a still-needed selection by changing family, core contrast, criticism, feasibility or refusal produces ResearchDesignHold; an unknown does not create a pending investigation when the present allocation is already settled as QualifiedNoNewStudy.
  • RMP2-16: A hold identifies the unresolved result, a qualified source or holder if known, the options affected, their plausible consequences and what would allow the choice to be completed. State which work the present basis cannot support.
  • RMP2-17: In a hold, distinguish conditional strategies from a strategy already selected.
  • RMP2-18: Across every admissible remaining value and their joint combinations, a PostSelectionNeed preserves the family, core contrast, criticism, feasibility and refusal conditions.
  • RMP2-19: A PostSelectionNeed value outside the admitted range reopens selection.
  • RMP2-20: An empty PostSelectionNeed set is written explicitly as none.
  • RMP2-21: One of EvidenceSufficientStop, ResearchDesignDecision plus genuine PostSelectionNeed, ResearchDesignHold, or QualifiedNoNewStudy answers the present selection boundary. QualifiedNoNewStudy preserves insufficient evidence or a bounded answer and the reason no new study is selected.
  • RMP2-22: Entry to RMP.3 requires a selected strategy or a compatible supplied equivalent. Its specification, execution, findings and permissions have their own grounds.
  • RMP2-23: Expected information gain is tied to the uncertainty the strategy could reduce and the conditions under which the answer would be useful.
  • RMP2-24: Where reporting guidance is used, identify what it supplies. Establish design selection and actual execution from evidence for the particular claim.
  • RMP2-25: Domain and measurement conclusions have their professional grounds. Decisions about rights, access, safety, ethics, resources, assurance or further action stay with the applicable authority.

RMP.2:8 - Common Anti-Patterns and How to Avoid Them

Anti-patternWhy it failsRepair
“Use the gold-standard design”A design suited to one question or threat can miss another.Compare what each serious option could expose under the available conditions.
“List the ingredients and the design will follow”A list of sources and measurements can leave unexplained how they could distinguish the claims.Use §4.4 to connect the needed distinction to an obtainable operation and follow it on actual material. Repair a failure that matters or report the resulting limit.
“The protocol is the design”A detailed procedure may leave its evidential rationale open.Look for the basis of the question-to-evidence relation, relevant alternatives, threats and selection. A protocol that already explains them can supply this information.
“Evidence is incomplete, therefore study”An unresolved claim does not select a useful obtainable acquisition.Compare the whole burden and contribution; return QualifiedNoNewStudy when the current allocation closes without a new study.
“No budget, therefore evidence sufficient”Resource refusal is not an epistemic conclusion.Preserve the bounded answer or unresolved question in QualifiedNoNewStudy when no current study is selected; resource and receiving decisions remain with their owners.
“Select provisionally and validate later”A missing mechanism, measure, access, protection or permission could overturn the selected strategy.If the selection is still needed, return a hold and describe the missing result and its consequences. If the current allocation already justifies no study, return that decision and the evidence limit.
“Everything must be fixed before selection”Some implementation details cannot change the justified strategy.Check that jointly admissible values preserve the family, comparison, criticism, feasibility and refusal conditions. Record a remaining need as PostSelectionNeed when it passes that test.
“Two software packages are two research designs”Different implementations may support the same evidential comparison.Determine what differs. An eight-hour run and a two-hour run may realize the same research design at different costs. Compare practical alternatives where that difference matters, and distinguish a change of design when evidence, criticism or meaningful constraints change.
“Mixed means stronger”Components may share a bias, disagree or lack a justified relationship between their results.Identify what each contributes, how the results could be combined and what can still be concluded if one fails or the components disagree.
“A rerun is a replication”Code/data/environment reproduction, new-data replication, validation, and robustness test different claims.Name the live threat and select the credibility move that bears on it.
“The selected design authorizes the study”Its justification establishes a strategy. Acting may require permission for access, safety, ethics, rights, resources or another use.Identify any applicable authorization already in force or the decision still needed from its holder.

RMP.2:9 - Consequences

A researcher can use an adequate answer, justify a feasible design, hold a particular choice or close the present allocation with no new study. The account shows how the evidence bears on the question and what conditions protect the inference. A detail can be left to later work when its admissible values and combinations cannot change the selection.

To compare unlike strategies, establish what each could contribute and what it might miss. A change in access, measurement, the proposed mechanism, protections or authority can alter which strategy is feasible or worthwhile. Reconsider those parts of the comparison on which the use depends.

A later reader can distinguish a sufficient answer, a selected strategy, a choice still held open and a completed decision to undertake no new study on the available basis. Evidence of execution, findings, a commission or a subsequent decision must establish that particular occurrence.

RMP.2:10 - Rationale

A design identifies how relevant material could be obtained and what its examination would establish. Following the proposed operation on concrete material can expose a gap that a family name or option list conceals. Both reference seals have wear score 6, so matching scores cannot establish matching mechanisms. The qualified signatures and comparability conditions support a different proposed test. When selection is still needed, weigh that contribution and its limitations against the alternatives and full burden.

A remaining detail belongs to PostSelectionNeed when all admissible values and joint combinations preserve the family, core comparison, criticism, feasibility and refusal. A selection still needed belongs on hold if plausible values support materially different outcomes. Where the present allocation has already selected no study, the same unknown can remain a limitation or a reason to reconsider.

Use EvidenceSufficientStop to return an already adequate answer. Use QualifiedNoNewStudy when a decision to undertake no new study is justified and the evidence is still insufficient. Substantiate domain facts, instrument capability and permissions on their proper grounds when the chosen strategy relies on them.

RMP.2:11 - SoTA-Echoing

The methodological question is how to develop and choose a strategy that bears on the research question, or to finish without a new study, while making consequential missing information visible.

The MRC/NIHR complex-intervention framework uses programme theory, context and key uncertainties to guide inquiry, revision and stopping. Its phases concern complex health interventions. The transferable contribution here is the connection between explanatory premises and research choice. In Cochrane Handbook chapter 2, §2.5, conceptual models identify assumptions, relate effects to required conditions and shape a review’s comparisons. Apply those operations to the question they serve. Review eligibility and synthesis retain their specific methods.

C.39 guides the construction in §4.4. Work backward from the required answer to an operation, follow its use on available material and repair a failure where worthwhile. A specialist may already have supplied an adequate strategy. When the evidence relation is missing, a list of candidate fields leaves the gap open. In the seal example the wear index admits both mechanisms. Professional premises allow a different comparison using the signatures. The example illustrates the reasoning and its additional design effort. An empirical claim about the method’s effectiveness would need evidence about actual use.

Different professional sources guide different inferences. The Standards for Reporting Qualitative Research supports transparent reporting while allowing different qualitative paradigms and methods. The American Historical Association’s Statement on Standards of Professional Conduct addresses source use, assumptions, bias and alternative interpretations. The Cochrane Handbook relates review eligibility, search, collection and synthesis to the question. The National Academies’ Reproducibility and Replicability in Science distinguishes computational reproduction from new-data replication. Use the contribution that bears on the current threat.

A domain ranking or a familiar sequence of studies can guide a recurring question when the applicable conditions justify it. Applied across all questions, a fixed ranking can conceal the error that historical source criticism, qualitative reflexivity, observational comparison, model validation or corpus synthesis could expose. Compare the actual evidence and losses. Existing evidence or a justified no-study allocation can also finish the work.

For domain claims, instrument validity, rights, access and protections, use the appropriate evidence and authority. If a professional source or local rule changes the feasible options or the error a design can expose, revisit the affected comparison. If the joint-input test repeatedly misclassifies a design, examine both the criterion and its application and repair the source of the error.

RMP.2:12 - Relations

Use RMP.10 when ordinary A.3.1.MR recovery leaves a consequential question about a candidate account of how work was done. Use RMP.11 for a source-critical explanation of how an identified Method or its relationship to another Method changed. Either may supply an evidential distinction for design selection. Establish access, suitable sampling and a selected strategy separately when they are needed.

  • RMP.1 — Decide Whether a Question Needs Research and When to Stop. Use its qualified brief and current account to enter this method. Return there if a change in question, use, evidence or stopping conditions requires qualification that is still missing.
  • RMP.1:4.3. Use F.1, F.0.1 and A.10 to compare available meanings and evidence where needed. A source comparison may already answer the question, reveal a different question or identify a material uncertainty requiring research.
  • C.39 — Find and Develop a Way to Obtain a Result. Use §4.4 to recover a relevant operation, reason backward from the needed distinction, follow the proposed connection on material and repair it where needed. Use professional competence and domain evidence for the underlying relations.
  • Core Method development and A.3.1.MI. Use Method Engineering for the identity, repertoire, architecture, qualification, fit, worth or introduction of a reusable Method. RMP.2 selects a research strategy for one qualified question.
  • Operationalization and trace. Enter RMP.3 with a selected design and any genuine PostSelectionNeed items, or their compatible equivalents. A stop, hold or no-new-study decision can complete the present work. To report that the research took place, use a trace or other suitable evidence of the actual work.
  • Professional evidence and authority. Use domain research, measurement and instrument evidence, archives and other suppliers for the facts the strategy needs. For participation, rights, access, safety, confidentiality, resources or consequential use, rely on the applicable permissions and decisions. Identify a missing permission when it affects the choice.
  • MRC/NIHR complex-intervention framework: supports perspective pluralism, uncertainty-led design, non-linear entry, reconsideration, and stopping for its declared intervention-research scope.
  • SRQR, AHA professional standards, Cochrane Handbook and the National Academies account. Use the family-specific contributions to reporting, source criticism, synthesis, reproduction or replication described in §11.

RMP.2:End

Referenced in the corpus

82 literal mentions in other sections. Read their context to establish the relation.