Making Clinical Reasoning Visible in Nursing Education

Two nursing students read the same case. A patient has sudden left-sided weakness, facial asymmetry, and difficulty attending to the left side of the environment. Both students choose the same response: initiate the emergency stroke pathway.

On a standard question, both students are correct.

But suppose the first student recognized a right-hemisphere pattern, connected the findings to an acute focal neurologic process, considered competing explanations, and understood why time-sensitive evaluation was necessary. The second student selected the same response because “weakness means stroke,” while incorrectly believing that left-sided weakness indicates injury to the left side of the brain.

The students arrived at the same answer through very different reasoning.

One response is supported by a coherent clinical model. The other contains a misconception that, by chance, did not change the final answer. If we score only the answer, that difference remains hidden.

A clinical reasoning trace helps us see this hidden difference.

The answer is visible. The reasoning is not.

Clinical judgment is often evaluated through an observable decision: the cue selected, hypothesis prioritized, intervention chosen, or outcome identified. Yet the reasoning behind that judgment includes clinical relationships that may never appear in the final response.

The National Council of State Boards of Nursing (NCSBN) developed the NCSBN Clinical Judgment Measurement Model (NCJMM), which organizes clinical judgment into six measurable cognitive operations:

1. Recognize cues
2. Analyze cues
3. Prioritize hypotheses
4. Generate solutions
5. Take action
6. Evaluate outcomes

The NCJMM was designed to support valid measurement of clinical judgment and decision-making in entry-level nursing. Research supporting its item-development method describes clinical judgment as an iterative process: nurses may repeat or revisit parts of the process as the patient’s condition or the available information changes. The model also separates clinical judgment into parts that can be assessed. NCSBN Clinical Judgment Measurement Model Betts et al., 2019 Dickison, Haerling, & Lasater, 2019

Breaking the process into parts is important. It creates the possibility of locating a reasoning breakdown more precisely than “the student got the question wrong.” Research describing the NCJMM item-development method suggests that poor judgment may be traced to a specific part of the decision process when an assessment captures enough detail. Betts et al., 2019

But an NCJMM label alone is not a reasoning trace. Saying that an item measures Analyze Cues identifies the broad cognitive operation being assessed. It does not show which cues the learner connected, which mechanism the learner believed explained those connections, which alternatives were considered, or why one concern outranked another.

The space between the patient information and the final response is where the reasoning trace lives.

What FuseRN calls a clinical reasoning trace

FuseRN uses the following working definition:

A Learner Clinical Reasoning Trace is an ordered, evidence-bearing record of a learner’s expressed clinical reasoning: which cues were selected, how those cues were interpreted and connected, which hypotheses were considered, how concerns were prioritized, which solutions and actions were chosen, and how outcomes were evaluated.

Several words in this definition matter.

Learner distinguishes the student’s pathway from an expert-authored reference pathway.

Evidence-bearing means the trace must be grounded in something the learner actually did: a selection, explanation, ranking, connection, prediction, revision, or other observable response.

Expressed sets an important boundary. A reasoning trace is not a recording of the learner’s private thoughts. It is a structured sample of reasoning made visible through carefully designed tasks.

Ordered does not mean perfectly linear. Clinical reasoning is iterative. New information can change the leading hypothesis, alter the priority, make a planned intervention unsafe, or restart the cycle after reassessment. Order simply preserves what the learner knew, believed, or selected at each point in the case.

The phrase hidden reasoning is useful, but it should not be mistaken for mind-reading. What remains hidden during ordinary answer scoring is the clinically important structure between the case and the response. A clinical reasoning trace makes parts of that structure observable.

One correct answer can conceal different reasoning states

Getting the answer right and showing sound reasoning are not the same thing.

Robust correctness

The learner selects the correct response and provides enough evidence that the relevant cues, interpretation, priority, action, and patient context are aligned.

Fragile correctness

The learner selects the correct response but supports it with an invalid mechanism, a rule that does not apply to the case, a misconception, or too little evidence to evaluate the reasoning. The answer may have resulted from guessing, pattern recognition without understanding, or a faulty clinical model that happened to produce the keyed answer.

Preserved partial reasoning

The final response is wrong, but part of the learner’s reasoning remains clinically valid. The learner may have recognized the decisive cues and generated the correct hypotheses but ranked them incorrectly. Treating the entire attempt only as “incorrect” discards useful information about what the learner understands.

Reasoning breakdown

The response is incorrect, and the available evidence identifies an important reasoning error. Examples include selecting irrelevant cues, reversing a causal relationship, overlooking a contraindication, or comparing the observed outcome with the wrong expectation.

This separation matters educationally because each state calls for different feedback. A knowledge gap, a prioritization error, an action-selection error, and an unsupported correct answer should not all lead to the same review.

What does a reasoning trace contain?

A useful trace may contain evidence from several parts of the learner’s reasoning:

• Cue recognition: Which findings were selected, ignored, or classified as normal, abnormal, trending, or important because they were absent?
• Interpretation: What did the learner believe those findings meant in this patient?
• Clinical connection: Which cues were clustered, and what relationships did the learner identify among them?
• Mechanism: What pathophysiologic or functional explanation supported the interpretation?
• Hypothesis formation: Which possible needs, concerns, or explanations were considered?
• Prioritization: Why did one hypothesis outrank the others—urgency, likelihood, severity, risk, reversibility, or time sensitivity?
• Solution generation: What desired outcomes and feasible options were identified, and what should be avoided?
• Action: Which response was chosen, in what sequence, and under which patient conditions?
• Evaluation: What result was expected, what was observed, and what should happen next?

Not every question needs to capture every part. The depth of the trace should match the conclusion the item is meant to support. If an item is intended to assess cue recognition, the learner may only need to select the relevant findings and distinguish them from believable but unhelpful details. If an item assesses prioritization, the learner should compare the leading hypotheses and state the criterion used to rank them.

Mapping traces to the NCJMM

The NCJMM provides a useful structure for organizing reasoning-trace evidence without assuming that each operation is isolated or permanently attached to a particular clinical fact.

A smaller reasoning loop happens inside each operation

Within each NCJMM operation, the learner selects the information that matters, connects it to relevant nursing knowledge, generates possible explanations or responses, and tests those possibilities against the available evidence. Options that conflict with the patient’s findings, safety needs, or clinical constraints are ruled out. The learner then makes a provisional judgment—a best answer based on the evidence available at that point—and updates that judgment when new information appears.

FuseRN groups these actions into a reasoning loop:

Select → Connect → Generate → Test → Decide → Update

This reasoning loop is not a seventh NCJMM operation. It is FuseRN’s way of explaining the smaller thinking actions that can occur within and between the six NCJMM operations.

The same finding or clinical relationship can support different reasoning tasks. Left-sided neglect might be something a learner must recognize in one item, connect to cerebral localization in another, use to prioritize a neurologic hypothesis in a third, or reassess after an intervention in a fourth.

The NCJMM operation therefore describes how the learner is being asked to use the information in a particular context. It does not change the underlying clinical meaning of that information.

Reconstructing the trace in the stroke example

Return to the patient with sudden left-sided weakness, facial asymmetry, and left-sided inattention.

A compact reference trace might look like this:

1. The sudden focal findings are clinically meaningful cues.
2. Weakness and facial asymmetry on the left suggest dysfunction of contralateral motor pathways—the pathways controlled by the opposite side of the brain.
3. Left-sided inattention adds evidence of right-hemisphere involvement.
4. The clustered findings support an acute right-hemisphere neurologic process more strongly than a generalized cause of weakness.
5. Because acute stroke is time-sensitive and carries substantial risk, emergency stroke evaluation is prioritized.
6. The immediate action is intended to accelerate diagnostic evaluation and determine whether the patient is eligible for acute treatment.
7. Later neurologic findings, imaging, vital signs, laboratory results, and the response to treatment help confirm, refine, or reject the initial judgment.

This trace contains more information than the final action alone. It also creates several possible points of divergence:

• The learner may fail to recognize neglect as important.
• The learner may identify the cues but localize the lesion to the wrong hemisphere.
• The learner may form the correct hypothesis but fail to appreciate its urgency.
• The learner may prioritize stroke but choose an unsafe or mistimed action.
• The learner may take an appropriate action but identify the wrong evidence for evaluating its effect.

Each error is different and calls for a different teaching response.

How can educators capture a trace without turning every question into an essay?

A clinical reasoning trace does not require an exhaustive think-aloud response after every item. Researchers have used think-aloud methods to study nursing decision-making, including during simulation, but everyday learning systems need shorter ways to capture meaningful reasoning evidence. Abdulmohdi & McVicar, 2023

A useful trace can often be captured with five short prompts:

1. Which findings most influenced your answer?
2. What is your leading hypothesis or client concern?
3. Why does it outrank the strongest alternative?
4. Why should your selected action address the problem?
5. What finding would show that the action worked—or that your judgment must change?

Other efficient trace tasks include:

• Selecting the most and least important cues
• Connecting two findings that belong in the same clinical cluster
• Ranking two or three hypotheses
• Identifying the strongest contradictory finding
• Rejecting the most attractive distractor
• Predicting the next change in the patient’s status
• Choosing what must be reassessed and when
• Revising an answer after one new finding is introduced
• Reporting confidence before and after additional evidence

The goal is not to collect more text. The goal is to collect the smallest amount of evidence needed to distinguish clinically important reasoning states.

How should the trace be evaluated?

Trace evaluation should preserve uncertainty. The absence of an explanation is not proof that the learner used an incorrect explanation. It may simply mean the task did not capture enough evidence.

A useful evaluation vocabulary includes:

• Valid: The expressed reasoning is supported by the case, clinical knowledge, and patient context.
• Valid alternative: The reasoning differs from the reference pathway but remains clinically defensible.
• Invalid: The expressed reasoning conflicts with the case, clinical evidence, or an applicable constraint.
• Missing: A required component was not supplied.
• Insufficient evidence: The response does not show enough to judge this part of the reasoning.
• Not applicable: The component was not required in this task or case state.

This prevents two common errors: treating every difference from the expected answer as wrong and treating every unexpressed thought as a misconception.

Multiple valid reasoning traces may support the same judgment. Clinical reasoning is constrained, but it is not always limited to one path. Learners may use different supporting cues, reach the same priority through different defensible comparisons, or choose different safe solutions that meet the same patient need.

Why ordinary rationales are not enough

Many nursing questions provide a rationale after the learner answers. That rationale explains why the keyed answer is correct, but it does not show how the learner reached it.

Even asking the learner to provide a rationale after answering has limits. The explanation may be incomplete, reconstructed after the decision, or shaped by the available answer choices. A stronger approach gathers small pieces of evidence throughout the reasoning process:

• Cue selection before the answer
• Hypothesis ranking before the action
• Confidence before feedback
• Expected outcome before reassessment data appear
• Revision after an important new finding

This creates a time-ordered record of how the learner’s expressed clinical model changes as the patient situation unfolds.

A practical self-tracing method for nursing students

Students can use the same idea while practicing NCLEX and course questions. Before reading the rationale, record:

1. Task: What is the question asking me to do—recognize, analyze, prioritize, generate, act, or evaluate?
2. Cues: Which two or three findings actually control the answer?
3. Meaning: What do those findings indicate in this patient?
4. Alternatives: What is the strongest competing explanation or response?
5. Priority: Why does my choice come first?
6. Mechanism: Why should the selected action produce the desired effect?
7. Evaluation: What result would confirm that the choice worked?

After reviewing the answer, do not record only whether it was correct. Identify the earliest point at which your reasoning diverged from the supported pathway.

Did you miss a cue? Misinterpret it? Use the wrong mechanism? Rank the hypotheses incorrectly? Apply a familiar rule outside its proper context? Choose the wrong action despite sound analysis? Compare the result with the wrong expected outcome?

That earliest divergence is usually more useful for learning than the final score.

From answer explanations to reasoning-specific feedback

A clinical reasoning trace is useful not because it produces more data, but because it can support more specific feedback.

Instead of:

Incorrect. The correct answer is to initiate the stroke pathway.

Feedback could say:

You recognized the focal neurologic findings and selected the appropriate immediate response. However, your localization rationale reversed the relationship between the affected hemisphere and the side of the motor deficit. Preserve the action choice, but repair the mechanism connecting the lesion to the observed weakness.

That feedback does three things:

1. Preserves what the learner understood correctly.
2. Identifies the earliest meaningful error.
3. Targets review to the missing or incorrect relationship.

This is more useful than repeating the complete rationale because it responds to the learner’s actual evidence.

The larger opportunity

Nursing education already has strong models of clinical judgment. The NCJMM gives educators a shared way to describe and measure it. Tanner’s model of clinical judgment focuses on noticing, interpreting, responding, and reflecting. Research on information processing, decision-making, think-aloud methods, and clinical reasoning education also helps explain how learners organize and express clinical information. Tanner, 2006 Muntean, 2012 Dickison et al., 2016

The opportunity is to turn these ideas into a structured learner record that educators can review.

A clinical reasoning trace does not replace clinical judgment models. It records the evidence a learner shows while doing the work described by those models.

That record can support:

• More intentional item construction
• Partial credit for parts of the reasoning that were correct
• Identification of fragile correctness
• Feedback aimed at the earliest meaningful divergence
• Separation of missing evidence from invalid reasoning
• Comparison of multiple valid reasoning pathways
• Tracking recurring misconceptions over time
• Practice that targets the failed reasoning operation rather than the diagnosis label alone

The goal is not to require nurses to explain every private thought or to reduce clinical judgment to a rigid sequence. The goal is simpler:

When a learner makes a clinical decision, capture enough evidence to understand what supported it, where the pathway became unreliable, and what should be repaired next.

The final answer still matters. In practice, clinical decisions have consequences. But if nursing education wants to develop clinical judgment instead of only scoring outcomes, it must get better at examining what happened between the cue and the answer.

References

• Abdulmohdi, N., & McVicar, A. (2023). Investigating the clinical decision-making of nursing students using high-fidelity simulation, observation and think aloud: A mixed methods research study. Journal of Advanced Nursing, 79(2), 811–824.
• Betts, J., Muntean, W., Kim, D., Jorion, N., & Dickison, P. (2019). Building a method for writing clinical judgment items for entry-level nursing exams. Journal of Applied Testing Technology, 20(S2), 21–36.
• Dickison, P., Haerling, K. A., & Lasater, K. (2019). Integrating the National Council of State Boards of Nursing Clinical Judgment Model into nursing educational frameworks. Journal of Nursing Education, 58(2), 72–78.
• Dickison, P., Luo, X., Kim, D., Woo, A., Muntean, W., & Bergstrom, B. (2016). Assessing higher-order cognitive constructs by using an information-processing framework. Journal of Applied Testing Technology, 17(1), 1–19.
• Muntean, W. J. (2012). Nursing clinical decision-making: A literature review [Unpublished manuscript]. Department of Psychology, University of Oklahoma.
• National Council of State Boards of Nursing. (n.d.). Clinical Judgment Measurement Model.
• Tanner, C. A. (2006). Thinking like a nurse: A research-based model of clinical judgment in nursing. Journal of Nursing Education, 45(6), 204–211.