The two matrix formats
The Next Generation NCLEX uses two grid formats. Matrix multiple choice presents rows of findings or interventions and columns that are mutually exclusive, so you select exactly one option per row — for example, effective, ineffective or unrelated. Matrix multiple response allows more than one selection per row, so a single finding might be marked as indicating two different conditions.
Recognising which of the two you are facing takes three seconds and prevents a whole category of error. If the directions say select one option in each row, resist the urge to hedge by ticking two. If they permit multiple selections per row, do not force yourself to choose only one.
Both formats are scored with partial credit. In matrix items, each row generally earns its own point when answered correctly, which means the grid is not one question with one score but a set of small questions bundled into one screen.
- •Matrix multiple choice: exactly one column per row
- •Matrix multiple response: one or more columns per row
- •Read the directions line before touching the grid
- •Scoring is per row, with partial credit
The row-by-row method
The instinct on seeing a large grid is to scan columns and look for patterns. Resist it. Column scanning invites you to distribute your answers evenly for aesthetic reasons — three in each column feels right — and the exam does not distribute correct answers evenly. It is perfectly normal for six of eight rows to fall in the same column.
Instead, cover the rest of the grid mentally and take the first row alone. Ask the question the column headers pose, about that finding only, and commit. Then move to the second row and repeat. This converts an intimidating twenty-four-cell decision into eight independent decisions, each of which you are perfectly capable of making.
Where a row genuinely stumps you, apply a default rather than freezing. In evaluation matrices, ask whether the intervention could plausibly influence this value within the timeframe: if not, unrelated is often correct. In expected-versus-unexpected matrices, ask whether a textbook description of this diagnosis would list this finding.
- •Never distribute answers evenly for the sake of balance
- •Isolate each row and answer it as a standalone question
- •Commit and move — do not revisit rows repeatedly
- •Use a default rule for genuinely uncertain rows rather than stalling
Reading the column headers properly
Column headers change the entire task and are the most commonly skimmed part of the item. A grid asking whether findings are expected or unexpected for a diagnosis is a pathophysiology question. A grid asking whether findings indicate improvement, no change or worsening is an evaluation question requiring comparison with earlier data. A grid asking whether actions are indicated, contraindicated or non-essential is a scope and safety question.
The word indicated deserves particular attention. Indicated means the action is appropriate and should be performed for this client now. Non-essential means the action is not harmful but does not address the current problem. Contraindicated means performing it would harm this client. Candidates routinely mark a safe but irrelevant action as indicated, which is scored as wrong.
When headers reference time — such as findings that require immediate follow-up — the timeframe is part of the criterion. A finding can be abnormal, worth documenting, and still not require immediate follow-up, and the grid is testing precisely that discrimination.
- •Expected vs unexpected → pathophysiology of the diagnosis
- •Improved vs unchanged vs worsened → compare with earlier chart values
- •Indicated vs contraindicated vs non-essential → scope, safety and relevance
- •Immediate follow-up → urgency, not merely abnormality
Traps built into matrix items
The first trap is the unrelated-but-real finding. In an evaluation grid, item writers include values that genuinely changed but could not have been influenced by the interventions given. Marking them as evidence of effectiveness shows content knowledge without causal reasoning, which is exactly what the format is designed to detect.
The second trap is the correct-but-slow parameter. A haemoglobin A1c or a wound culture result cannot change within a shift. If the grid asks about the response to interventions performed an hour ago, values on that timescale belong in the unrelated or unchanged column.
The third trap is scope. Actions that require a prescription — starting a medication, ordering a test, changing a dose — are contraindicated or not indicated for the nurse unless the chart contains an order or protocol authorising them. Always check the orders tab in case-study grids before marking a clinical action as indicated.
- •Unrelated findings: real changes the interventions could not have caused
- •Slow parameters: values that cannot move within the stated timeframe
- •Scope violations: actions requiring an order that the chart does not contain
- •Expected findings presented as if they were complications
Timing and partial credit strategy
Budget roughly ninety seconds to two and a half minutes for a matrix item, depending on the number of rows. Because rows score independently, the optimal strategy is breadth first: give every row a reasoned answer, then return to the two or three you found hardest if time allows.
Never leave a row blank. There is no penalty for an incorrect row, and an empty row is a guaranteed zero. Even a coin-flip between two columns has positive expected value, and in practice your clinical instinct is better than a coin flip.
Do not let one difficult row consume the item. The candidate who spends four minutes agonising over row three and then rushes rows four through eight typically scores worse than the candidate who answers all eight thoughtfully in two minutes. This is arithmetic, not attitude.
- •90 seconds to 2.5 minutes per matrix item
- •Answer every row before refining any row
- •No penalty for wrong rows — blanks are strictly worse than guesses
- •Cap the time any single row may consume
How to practise grids
Grid confidence comes from repetition of the method rather than from grid-specific content. Any practice question can be converted into grid training: take a scenario, list six findings from it, and classify each as expected, unexpected or unrelated before checking the rationale. You are rehearsing the exact discrimination the format scores.
When you review a missed grid, identify the row you got wrong and name the reason in one phrase — wrong timescale, wrong scope, expected finding, or knowledge gap. Row-level review is far more useful than noting that the item as a whole was missed, because the grid contains several independent lessons.
Then practise grids inside mixed sets and full mock exams so that switching between item types stops costing you attention. Use the content-area question sets on this site to build the underlying clinical knowledge, and the full-length mock exams to rehearse format switching under time.
- •Convert ordinary questions into self-made classification grids
- •Review missed grids row by row and name the failure reason
- •Practise format switching inside mixed timed sets