HRM.2 - Substantiate a personnel assessment and selection
Type: Method pattern Status: Stable
HRM.2:1 - Problem frame
Use this when recruitment, promotion, allocation or another personnel choice relies on a claim about a person’s suitability for actual work. Begin with the decision and the contribution it needs, not an available test or candidate score. The first useful result is a warranted comparison and recommendation, or the exact uncertainty that prevents it. The object is the assessment and selection inference; an appointment remains a separate act.
Use a sufficient established procedure directly when its work, population, support conditions and decision remain applicable. A formative learning conversation need not become a selection exercise.
HRM.2:2 - Problem
Evidence gathered for one purpose can acquire an unjustified second meaning. Course completion becomes mastery, mastery becomes suitability for a different job, a supervisor rating becomes general potential, and an AI score becomes an appointment recommendation. The procedure can consistently rank people while measuring the wrong contribution or conditions they will never face at work.
HRM.2:3 - Forces
| Force | Consequence |
|---|---|
| Consequential choice and incomplete evidence | State what the evidence supports and obtain only the missing contribution that can change the decision. |
| Comparable treatment and attainable demonstration | Comparable opportunity may require a qualified accommodation rather than identical irrelevant conditions. |
| Actual assisted work and independent capability | Test the required configuration; add unaided performance only for a justified work condition. |
| Speed and validity | A fast, inexpensive procedure is useful only for an inference it can support. |
HRM.2:4 - Solution
HRM.2:4.1 - Define the decision and investigate the work
Name what is being decided, who may decide, the candidate population, applicable eligibility conditions and the proposed terms. Separate an entry requirement from something learnable after appointment with available support. Recover the actual work through documents, observation and knowledgeable participants. Include consequential uncommon situations, coordination and recovery where they matter; frequency alone does not establish importance. Identify which variations in work, location, tools and support the role must cover.
Translate this analysis into claims needed for the decision: for example, the ability to turn a client’s ambiguous concern into a bounded advisory question using the provided agent, explain the resulting recommendation, and recognize when a required professional conclusion is unavailable. Distinguish this from the obsolete ability to type a report unaided. If adequate work analysis is missing, obtain it before selecting the instrument. OPM’s job-analysis practice and qualified industrial/organizational assessment are useful suppliers; the reader must still specify the receiving decision.
For an internal talent or promotion opportunity, establish who can be considered, how nomination or application works, and which exclusions have an applicable basis. Exclusive, inclusive and blended arrangements make different participation promises. Consider whether a supported direct application or another access route is needed; do not treat a manager’s nomination as evidence of capability. More places do not by themselves establish that people experience inclusion or that the selection is fair.
For recruitment, make the proposed role and terms sufficiently real to offer. Choose channels capable of reaching the relevant population, communicate material requirements and conditions accurately, and provide an accessible application route. HRM.1/HRM.3 may need to change an unobtainable offer. A small response to an unattractive offer is not evidence that no capable people exist.
HRM.2:4.2 - Build the evidence argument before scoring candidates
For each consequential claim, choose evidence that could support it and explain the inferential step. A representative work sample may demonstrate a defined operation; a structured interview can investigate particular prior behavior or reasoning; a qualified test can support a specified construct or predictive claim. None is universally best. Specify the conditions, relevant comparison and decision use before knowing who benefits from the rule.
| Claim needed | Possible evidence | What must be established before use |
|---|---|---|
| Performance of a representative advisory task | Work sample with permitted tools and an explained scoring basis | The task samples consequential work, scoring concerns that work, and the support corresponds to the job |
| A contribution across a wider class of cases | Relevant history or a professionally supported assessment | The evidence can generalize to the proposed work, population and conditions |
| Eligibility or a required qualification | The applicable verified fact | The rule actually requires it and the fact concerns this person and use |
| Performance after an obtainable learning period | Appropriate evidence and a credible learning/provision arrangement | The recommendation distinguishes present capability from the conditional future contribution |
Ask a qualified assessment supplier for the validity argument for this decision, population and conditions; the consistency of measurement; relevant fairness and bias findings; limitations; permissible administration and change conditions. A vendor’s reference customer or general accuracy figure is insufficient when the inference changes. The supplier can perform technical validation; HR must obtain enough of its result to understand which use is warranted. If the result is unavailable, use an adequate alternative or leave the dependent comparison unresolved.
A work sample’s resemblance to the job alone does not establish every predictive claim. Likewise, correlation with old supervisor ratings is not automatically evidence of the new contribution if those ratings reflect obsolete tasks or unequal opportunities. Keep the intended outcome and the evidence for it explicit. Use existing HCD or OPS.22 results only within the claim each supports; a good operating result may depend on team or agent contributions not controlled by the individual.
HRM.2:4.3 - Make the assessment usable for the actual candidates
Specify administration, instructions, resources, time conditions, evidence handling and scoring. Ensure assessors can apply the criteria, and resolve material disagreements by examining the work and scoring grounds. Preserve applicable access, accommodation, privacy and representation requirements. An accommodation that removes an irrelevant barrier can improve the inference; a change that alters the assessed contribution needs its effect examined by the qualified supplier.
Tell candidates the material conditions and permitted use of assistance. If the job uses an agent, assess the relevant assisted contribution. If an outage or independently required operation matters, assess that condition for its own reason. Do not conceal a test of unaided performance inside a nominally assisted task, or interpret access to a more capable private agent as the person’s general mastery without justification.
AI assessment needs an identified model/procedure and evidence relevant to its use. Material changes in scoring or inputs can make prior scores incomparable. Reuse of data from other purposes, including private developmental or health material, requires its own grounds and permissions. A competent agent may perform assessment or checking; the required independence, competence and institutional authority determine the allocation.
HRM.2:4.4 - Compare candidates under the justified rule
Apply the supported procedure to the eligible population. Separate missing evidence from unfavorable evidence and record material uncertainty. A person denied a relevant opportunity should not silently receive the same interpretation as one who demonstrated failure. Seek a proportionate clarification or another valid opportunity when it can change the choice and is permitted.
Distinguish minimum conditions from compensable preferences. A required professional permission cannot be compensated by a high interview score. Among candidates who meet binding conditions, use the explained comparison relevant to the work. Do not construct weights retrospectively to select a favored person. When candidates are not distinguishable on the evidence, use the applicable tie or further-evidence procedure and state the limit; artificial numerical precision supplies no missing information.
Construct the recommendation with the proposed work and terms, evidence and uncertainty, applicable conditions, and any learning or support on which it depends. The authorized decision maker may accept it, request a defined missing contribution or make another supported choice within the actual rules. Communicate the decision and available correction or challenge route through HRM.4. Use ADM and the relevant organizational process to make an appointment or assignment effective; selection itself does not provide access, remuneration or authority.
HRM.2:4.5 - Revisit the inference when its grounds change
Follow the receiving work and appropriate assessment evidence. A failure can concern the selection inference, missing promised support, a changed task or an operating arrangement. Investigate before treating it as proof that the person was unsuitable. Reopen the relevant validation or procedure when the work, population, tool configuration, scoring, criterion or applicable rule changes materially. Preserve results that still support their narrower uses.
HRM.2:5 - Archetypal Grounding
A reporting service needs two advisers. Its old test ranks unaided spreadsheet speed. Work analysis now shows that agents prepare the tables; advisers must frame client questions, identify unsupported conclusions and negotiate an actionable response. The owner constructs two representative cases with the provided agent and criteria covering those operations. A qualified assessor examines the work coverage, scoring and limits. One case contains missing client evidence; a warranted refusal to recommend is a successful response, not an incomplete answer.
Candidate A produces a polished recommendation but misses that absence. Candidate B identifies it, obtains the permitted clarification and supplies a usable bounded answer. Under the established critical condition, fluency does not compensate for the unsupported recommendation. The comparison favors B for this role, subject to the remaining eligibility and employment conditions. It does not establish B’s general intelligence or unaided spreadsheet mastery.
Now the proposed service must continue during agent outages. The owner may obtain fallback provision or require an additional human operation. Only after that choice should the assessment change. A blanket ban on tools would have tested a different job; a blanket promise that tools remove all human requirements would have missed the changed dependency.
HRM.2:6 - Bias-Annotation
The incumbent population, historical ratings and recruitment channels can encode unequal opportunities. A subgroup difference is a reason for the appropriate professional and institutional inquiry, not by itself proof of a particular measurement mechanism. Candidate acceptability, psychometric validity and legal permissibility remain distinct. Use qualified grounds for each relevant claim.
HRM.2:7 - Conformance Checklist
- Is the personnel decision connected to actual work and a specified population and support configuration?
- Does each material inference have appropriate evidence, scope and limits?
- Are administration, access and scoring sufficient for the intended comparison?
- Are missing evidence, unmet requirements and preference differences distinguished?
- Are recommendation, authorized act, actual provision and further challenge kept separate?
- Can a change reopen the affected inference without rewriting unrelated findings?
Recognizing a plausible work sample does not assure its validity. That conclusion requires the evidence appropriate to the actual claim.
HRM.2:8 - Common Anti-Patterns and How to Avoid Them
Completion equals competence. Recover what was actually demonstrated. Potential as an unexplained score. Name the contribution and the inference. Vendor accuracy equals local validity. Obtain the relevant population, work and use grounds. The highest total overrides every requirement. Keep binding conditions separate. AI implies mandatory human rescoring. Select a checker for actual competence, independence or institutional necessity.
HRM.2:9 - Consequences
The comparison becomes usable for its defined decision, and unsuitable instruments can be rejected before affecting candidates. Some decisions remain uncertain or require professional validation. A supported rejection or no-selection result can be preferable to a confident appointment based on irrelevant evidence. Administrative and participation work still remain after selection.
HRM.2:10 - Architectural Rationale
The Method owns the personnel inference joining work evidence to a consequential choice. It consumes job analysis, learning and operating results without equating them. The distinction allows one observation to retain its valid use while a proposed promotion or appointment remains unsupported.
HRM.2:11 - SoTA-Echoing
SIOP’s 2018 Principles and OPM’s job-analysis/work-sample practice ground work relevance and appropriate use. SIOP’s January 2023 AI-assessment guidance adds attention to changed scoring, data sources and the applicability of validation; it concerns AI used to assess people and expressly does not cover all employee–AI collaboration. This Method’s treatment of actual assisted work is a construction from work relevance and current obtaining alternatives, not a claim that the 2023 document evaluates 2026 agents. A sufficient qualified assessment service is a valid alternative to local procedure development. Its narrower technical result still needs the actual personnel decision and effective act.
HRM.2:12 - Relations
HRM.1 supplies the connected work and terms; HRM.3 supplies reward terms; HRM.4 supplies voice and challenge; HRM.5 constructs the resulting transition. OPS.22 supplies a warranted operating finding, HCD capability evidence, and qualified assessment suppliers the necessary validation. ADM and OCE establish the effective personnel relation and usable work.