EN
Five different object prototypes arranged in a circle and connected to a common center by blue lines.

Assessing a skill is not asking whether someone has it

A skill is a useful construct for organizing capabilities, but it never appears as a complete object to the person assessing it. What we see are partial manifestations in specific tasks, decisions, and contexts.

PRYSMAP4 min read

Asking whether someone “has” communication, analysis, or leadership invites a response based on an overall impression. The person seems persuasive, has held a particular position, or has a strong reputation. These signals may contain information, but they may also reflect halo effects, selective memory, and different expectations among assessors.

The problem begins before we choose the scale. It begins with the object.

A broad label allows too many interpretations

Two people may rate analytical thinking while assessing different things. One looks at whether the person structures a problem; another, whether they use quantitative methods; a third, whether they explain limitations and assumptions. Their answers are not comparable merely because they share a word.

The OPM policy on competencies proposes translating underlying knowledge, skills, and abilities into observable, measurable indicators of performance. This translation does not reduce a capability to a single behavior. It defines which manifestations are relevant to the work.

A well-constructed skill must clarify at least:

  • the kind of performance it organizes;
  • the elements that comprise it;
  • how it progresses across levels;
  • the contexts in which it should be applied;
  • the evidence relevant to its interpretation.

Decomposing is not the same as fragmenting

Decomposition is useful when it reveals internal relationships. In facilitation, for example, important elements might include preparing the purpose, guiding participation, managing tension, and producing agreements or next steps. This is not a universal list; the elements depend on the role and intended use.

Each element should connect to observable manifestations. We can observe how a person frames a question, redirects a conversation, explains a decision, or adjusts their intervention when new information emerges. We do not observe the skill as a substance separate from those actions.

The opposite risk is atomization. A checklist of twenty microbehaviors may be easy to complete while missing the integration that makes performance competent. A person may execute every step under simple conditions without knowing when to combine them or when to depart from the procedure.

Levels should change the demands, not just the adjective

Defining levels as basic, intermediate, and advanced shifts the problem to the assessor. What, exactly, changes?

The distinction may lie in depth, autonomy, complexity, scope, consistency, or the ability to guide others. The SFIA levels of responsibility show how attributes such as autonomy, influence, and complexity can be combined with professional skills. The relevant idea is not to copy its seven levels, but to make explicit which variable progresses.

The anchors need examples of performance distinct enough to separate one level from another. The OPM structured interview guide recommends developing example behaviors that help distinguish degrees of competency. Without observable boundaries, adding levels merely multiplies disagreement.

A key with a blue tag inserted in the lock of a partly open white door.

Context changes the meaning of evidence

A prepared presentation, a negotiation among opposing interests, and a technical explanation for a nonspecialist audience can all involve communication. They do not impose the same demands.

That is why the evidence must match the intended use. A self-report may reveal how a person understands their practice; a work sample may show application; a manager can provide observations over time; peers may see day-to-day coordination. No source is inherently superior. Each answers different questions and has different limits.

Combining sources without clarifying their functions does not automatically strengthen the assessment. It may collapse perceptions, demonstrated performance, and unequal access to relevant situations into a single number.

The conclusion must preserve the path

After an assessment, we should be able to explain which manifestations were observed, in what context, against which expectation, and under which interpretive rules. We should also be able to state what could not be observed.

That traceability changes the language. Instead of claiming that someone “has level 3 leadership,” a defensible conclusion might describe consistent evidence of autonomous coordination within a team alongside insufficient evidence of leadership across teams. The second formulation takes more space, but it supports a better decision.

It becomes assessable when the organization defines what counts as a relevant manifestation, how the demands change, and which evidence supports an interpretation without extending beyond it.