Program design CJMM in simulation

Building the NCSBN CJMM into a simulation program

Listing the six CJMM domains is easy. Getting a simulation program to actually observe and score them the same way every time is the hard part.

NCSBN CJMM Simulation Program Design Rater Consistency
01 The actual problem

The framework is short. Applying it consistently isn't.

Recognize Cues, Analyze Cues, Prioritize Hypotheses, Generate Solutions, Take Action, Evaluate Outcomes. The NCSBN's six-domain Clinical Judgment Measurement Model reads like a checklist. In a simulation lab, it behaves more like six separate judgment calls a faculty rater has to make in real time, watching a student talk to a manikin or a standardized patient, then translate what they just saw into a score.

The gap shows up between raters, not within the framework. Two faculty members watching the same encounter will usually agree on whether a student eventually took the right action. They agree far less often on whether the student actually recognized the cue that led there, or got there by luck. That's a scoring-consistency problem, and a better rubric doesn't fix it on its own. Rubrics only work if everyone applies them the same way, under the same time pressure, across every cohort a program runs.

02 Two decisions that matter

Neither one is about better manikins

Two decisions determine whether a CJMM-aligned simulation program actually works.

Decision one

Scenario design

Whether the case is built so all six domains have to surface in something a rater can actually see. A scenario where the correct action is obvious from the opening vitals doesn't test Recognize Cues or Analyze Cues at all, it tests whether the student remembers a protocol. Getting cues to hide in plausible places, the way they do with a real patient, is what makes the other five domains observable instead of assumed.

Decision two

Rater consistency

Whether two faculty members scoring the same encounter, on the same rubric, in different sections of the same course, land on the same score. Most programs manage this with periodic calibration meetings and a shared rubric document, which works until the semester gets busy and drift creeps back in.

03 Where automation fits

One half of the problem, not the other

NurseKind AI is built for the rater-consistency half of this, not the scenario-design half. Scenario design still has to come from faculty who understand the case and the cohort. What NurseKind AI does is apply the same six-domain rubric to every recorded student interaction the same way, whether it's the first cohort of the semester or the last, so a Recognize Cues score means the same thing in September and April. It doesn't replace the judgment that goes into building a good case. It replaces the judgment call about how consistently that case gets scored.

More on how the scoring works: NCSBN CJMM Assessment. For how this compares to exam-style prediction tools: Clinical Judgment Assessment Tools.

04 Further reading

For scenario design specifically

For programs building a CJMM-aligned simulation curriculum from the ground up, Laura McIlvoy's Simulation and NCSBN Clinical Judgment Measurement Model: A Nurse Educator's Guide to Innovative Program Development (Cognella, 2025) is a useful resource on the scenario-design side of this problem.

05 FAQ

Common questions about CJMM in simulation programs

Not every domain needs equal weight in every scenario, but a scenario that only tests one or two domains isn't really CJMM-aligned, it's a protocol check with CJMM language attached. Programs building toward full alignment typically design a case bank where the six domains rotate across scenarios rather than forcing all six into every case.
Most rely on periodic rater calibration sessions using recorded reference encounters, plus a written rubric with concrete behavioral anchors for each domain. That works reasonably well at small scale. It gets harder to sustain as faculty turn over and cohort size grows, which is the gap automated scoring is built to close.
No. NurseKind AI doesn't write or supply scenarios. It records and scores whatever real or simulated patient interaction a program already runs, against the same six CJMM domains, so the scoring stays consistent regardless of which scenario a faculty member built.

See consistent CJMM scoring across every cohort

The same six domains, scored the same way, whether it's week one or week fourteen.

Book a demo →

Institutional pricing available · Pilot rate for a single cohort