September 7, 2026 · 7 min read
TL;DR: Across this bank, answer options that establish the cause before choosing a remedy are keyed 25 times against 5 as distractors, the strongest positive pattern we have measured. It works because most wrong answers are remedies chosen before anyone knows what is wrong. It fails in three specific situations: when the stem has already given you the cause, when a documented process already covers it, and when containment outranks explanation.
Most exam advice is a list of things not to pick. Do not add people. Do not escalate straight away. Do not spend the reserve. Do not require overtime. Every one of those is supported by the data in this bank, and every one of them leaves the same question open: if all four are wrong, what does the exam actually want?
Usually the option that finds out what is going on. We measured it across the bank rather than asserting it: options phrased as determining, identifying, investigating or analysing the cause of something come back 25 keyed against 5 distractors. That is the inverse of every trap pattern we have written up, and it is the closest thing to a general rule the data supports.
Because of what the other three options in a typical scenario have in common. They are remedies, and a remedy chosen before the cause is known is a bet.
A warehouse-automation project's weekly review shows installation trending 12% behind the planned rate, still inside the approved cost baseline, with no corrective action yet taken. The options are to escalate the variance to the sponsor immediately, update the risk register to record the slip as a new risk, add resources to the installation work because 12% is significant enough to act on, or analyse the variance to determine its cause and whether it warrants a corrective action at all.
It keys the analysis. The other three are each defensible in isolation and each assume something the stem has not established: that the sponsor needs to know, that this is a new risk rather than a known one, that the fix is capacity. The keyed option is the only one that does not gamble.
A refinery shutdown project is in its hardest phase. The team has worked eight consecutive weekends, two engineers have asked to transfer, and morale is visibly gone. The options offer team-building activities, a completion bonus, closer monitoring of output, and identifying what is actually driving the low morale and addressing those causes. Three plausible morale remedies against one diagnostic, and the diagnostic is keyed, because eight consecutive weekends might mean exhaustion, or a specific unresolved grievance, or uncertainty about what happens after the shutdown. Each of those wants a different answer.
The same shape appears in quality. A project manager studying a control chart sees nine consecutive points on the same side of the mean. The keyed answer identifies the assignable cause. Redrawing the chart around a new mean accepts the shift without explaining it, and ignoring it under the rule of seven misreads what the run actually signals.
Three situations, and they account for the five distractor appearances.
| Situation | Why diagnosis fails | What the bank keys instead |
|---|---|---|
| The stem already gave you the cause | Investigating repeats work the question has done | Act on the cause you were handed |
| A plan or baseline would settle it | The document is the faster, authoritative answer | Check the plan, baseline or agreement |
| Damage is live and spreading | Explanation can wait, containment cannot | Stop the bleeding, then investigate |
The second row has a clean example. Stakeholders new to a project are confused by a report format the team reused from an earlier one. One option creates a fishbone diagram to find the root cause of the misunderstanding. It is a distractor, and it fails because the cause is not mysterious: the format was reused without agreement. The keyed answer uses monitor communications techniques, which may in turn trigger a change to the communications management plan. Reaching for a root cause tool here dignifies a known problem with an investigation.
The third row shows up in an agile scenario. Testing confirms faults in software delivered in earlier releases, and those faults must be put right immediately. One option runs a root cause analysis into why they did not surface earlier. The keyed answer stops the sprint and meets the product owner and team to rework the backlog. The root cause analysis is not wrong as an activity; it is wrong as the next action when confirmed defects are already in released software. It belongs in the retrospective, after the fire is out.
Common trap: "Root cause analysis is always the safe first move." It is the safest move when several explanations are live, and it is a distractor when the stem has already named the cause or when something is actively going wrong. This bank keys the diagnostic option 25 times and rejects it 5 times, and every one of those five rejections has one of the three shapes above. Reading which one you are in is the actual skill, and it is what a run of practice questions with each wrong answer explained rather than just marked wrong trains faster than any rule of thumb.
Worth noticing: the keyed options are rarely the ones that name a formal technique. They describe understanding the situation, and then usually pair it with a next step.
A social responsibility programme's targets are being missed. The keyed answer analyses why the targets are being missed and identifies the corrective action needed. Note the second clause. The keyed option is not "investigate and stop"; it is investigate and then act on what you find.
A bank's digital onboarding project closes without delivering one of the strategic objectives it was meant to support. The keyed answer analyses why the outcomes and the objective diverged, then discusses the findings with stakeholders to decide what happens next. Again, two clauses.
Two team members' disagreement in a design review turns personal. The keyed answer meets them privately to understand what is driving the disagreement, against distractors that issue formal warnings, call in HR, or leave them to it.
Stakeholders who cannot agree on a project's objectives or approach produce a keyed answer that breaks the situation down to identify the root causes of the disagreement, then works through roles and responsibilities.
So the pattern is not "pick the option with the word cause in it." It is that the exam rewards an option that gathers what it needs and then moves, over one that moves first. A diagnostic option offered as a terminal step, with no indication of what follows, is noticeably rarer in the keyed set.
Directly, and it is worth seeing them as one thing rather than five.
Overtime is keyed zero times in this bank. Spending a reserve is keyed zero times out of four. Handing a problem to the functional manager is keyed zero times out of seven. Doing nothing or waiting is keyed 2 against 38. Adding people to the team is keyed 1 against 7.
Every one of those is a remedy. Every one of them commits something, hours, money, somebody else's authority, or headcount, on a theory nobody has tested. The diagnostic option wins so consistently not because PMI loves analysis, but because it is usually the only option in the set that does not presume an answer.
That reframing is more useful on exam day than five separate rules, because you will meet distractors we have not measured. Ask what each option assumes. The one that assumes least about what is wrong is usually the one to pick, unless the stem has already told you.
Is "find the root cause" always the right answer on the PMP exam? No, but it is the strongest positive pattern in this bank. Options that establish the cause before choosing a remedy are keyed 25 times against 5 distractors. It fails when the stem has already given you the cause, when a documented process already covers the situation, or when the immediate priority is containing damage rather than explaining it.
How do I tell a genuine diagnose-first question from one where it is a distractor? Ask whether the stem has already told you why. If the scenario names the cause, or points at a plan or baseline that would settle the question, investigating repeats work the question already did. If several plausible explanations are live and the options each pick one, the diagnostic option is the one to pick.
Why does the exam reward diagnosis so consistently? Because most distractors are remedies chosen before anyone knows what is wrong. Adding people, spending a reserve, escalating, and requiring overtime all commit resources to a theory. The keyed option protects the decision that comes after it, which is the behaviour PMI's situational questions are written to reward.
Does a variance always need a corrective action? No. One bank question describes installation work trending 12% behind plan but still within the approved cost baseline. The keyed answer analyses the variance to determine its cause and whether it warrants corrective action before deciding on one. Acting directly on the number, escalating it, or logging it as a new risk are all distractors.
PMP Practice's 2,141 questions are re-certified against PMBOK 8 and the July 2026 ECO, with every wrong answer explained, not just marked wrong, which is the only way a pattern like this becomes a count instead of an opinion. Start the free 20-question sample, no card and no signup required to try it.
Is 'find the root cause' always the right answer on the PMP exam?
No, but it is the strongest positive pattern in this bank. Options that establish the cause before choosing a remedy are keyed 25 times against 5 distractors. It fails when the stem has already given you the cause, when a documented process already covers the situation, or when the immediate priority is containing damage rather than explaining it.
How do I tell a genuine diagnose-first question from one where it is a distractor?
Ask whether the stem has already told you why. If the scenario names the cause, or points at a plan or baseline that would settle the question, investigating repeats work the question already did. If several plausible explanations are live and the options each pick one, the diagnostic option is the one to pick.
Why does the exam reward diagnosis so consistently?
Because most distractors are remedies chosen before anyone knows what is wrong. Adding people, spending a reserve, escalating, and requiring overtime all commit resources to a theory. The keyed option protects the decision that comes after it, which is the behaviour PMI's situational questions are written to reward.
Does a variance always need a corrective action?
No. One bank question describes installation work trending 12% behind plan but still within the approved cost baseline. The keyed answer analyses the variance to determine its cause and whether it warrants corrective action before deciding on one. Acting directly on the number, escalating it, or logging it as a new risk are all distractors.