PEI PMA Grade 3 common mistakes
PEI PMA Grade 3 · Understand the traps and how to avoid them.
When a scenario describes a temporary, cross-functional team delivering a new system inside an IT, engineering or HR department, the candidate labels the work 'operations' or 'a program' because the department is permanent or several teams are involved.
Candidates classify by the organisational unit or headcount named in the stem instead of testing the nature of the work itself; departmental permanence and team size feel like the deciding features, and definitions are rehearsed as isolated lists rather than as tests to apply.
What to do instead
Test the deliverable, not the department: temporary endeavour with a unique output and progressive elaboration = project; related projects under shared governance and benefits = program; ongoing repetitive output = operations. Before reading the options, write the three tests in the margin and tick them against the stem.
Given a choice between protecting a sponsor's commercial interest and disclosing a defect that endangers users, the candidate selects the confidentiality or discretion option because it looks 'professional' and loyal to the client.
Ethics stems are deliberately written to tempt loyalty and confidentiality, and candidates apply the obligation they recall most strongly without ranking obligations against each other. Reporting options feel like over-reaction to a problem the sponsor has already accepted.
What to do instead
Rank obligations explicitly each time: public safety and law first, then the profession and its code, then the employer or client, then self-interest. Any option that keeps a safety-relevant fact quiet fails at the first step, however tactfully worded. Then choose the option that discloses through the correct channel - internal escalation first, statutory regulator where a legal duty exists.
On a change-control question the candidate selects 'the project manager approves the change' or 'the sponsor approves it' without checking the delegated financial threshold and approval route stated in the scenario.
Governance items look like generic change-management items, so candidates answer from the hierarchy they memorised (project manager to sponsor to board) rather than from the authority actually granted in the charter or governance plan described in the stem.
What to do instead
Extract three facts before answering: the value or impact of the change, the approval threshold, and who owns the budget and the benefit. Approval follows the threshold, not seniority. Where no threshold is given, prefer the option that follows the documented change-control process over the option that shortcuts to a senior person.
The candidate jumps straight to a formula (CPI = EV/AC, EAC = BAC/CPI) and plugs in numbers without first writing down PV, EV, AC and BAC, or mixes up the sign convention for cost and schedule variance and which EAC variant the stem describes.
Earned value is memorised as formula strings rather than as a data model, so under time pressure the first formula that surfaces is used. The qualifier in the stem that distinguishes typical from atypical variance is skimmed past because it sits in a subordinate clause.
What to do instead
Always build the four-value table first (PV, EV, AC, BAC) from the stem, then interpret: negative variance is bad, an index below 1 is behind or over budget. Use EAC = AC + (BAC - EV) for one-off atypical variance and EAC = BAC/CPI for variance expected to continue, and state the assumption in your working so a wrong variant is visible.
The candidate recommends fast-tracking or adding resources to compress a schedule without checking whether the dependency is mandatory or discretionary, and confuses total float with free float when judging how far an activity can slip.
Schedule compression is learned as a pair (crashing costs money, fast-tracking adds risk) with no link back to network logic, and float questions are answered from the label rather than from what each float actually protects.
What to do instead
Read the network before choosing a tool: fast-tracking only works where a dependency is discretionary or where overlap does not create rework; a mandatory dependency must be crashed or the logic redesigned. Free float is the slip available before the next activity's early start; total float is the slip available before the project finish. Pick the option that matches the consequence the stem describes.
Presented with a problem that has already occurred, the candidate selects 'log it in the risk register and monitor' or a generic 'mitigate' instead of treating it as an issue requiring corrective action, and never considers opportunity strategies for positive risks.
The risk register is the most familiar artefact, so every uncertainty-shaped stem gets pushed into it, and 'monitor' sounds prudent and safe as an exam answer. Positive risk strategies are skipped because candidates rehearse threats almost exclusively.
What to do instead
Ask one question first: has it happened? If yes, it is an issue - log it, take corrective action and do root-cause analysis. If no, it is a risk, and the chosen response must name a strategy (avoid, transfer, mitigate, accept, or exploit, share, enhance, accept) with a named owner. Reject any option that only monitors when the trigger has already fired.
The candidate selects a fixed-price contract for a scope the stem says is poorly defined or likely to change, or a cost-reimbursable contract for a well-specified commodity purchase, because fixed price is remembered as the option that 'protects the buyer'.
Contract types are memorised as a risk-transfer ranking without the condition that makes transfer workable - a stable, well-defined scope and a supplier able to price the risk. The question of who is best placed to manage the risk is never asked.
What to do instead
Check scope maturity in the stem before choosing: firm, well-defined scope means fixed price (firm-fixed-price, or FPIF where some incentive is needed); evolving scope or buyer-controlled design means cost-reimbursable with fee or incentive; small, off-the-shelf, low-value buys mean purchase order or time and materials. Allocate each risk to the party best able to control it, then check the contract type matches that allocation.
When the stem describes defects recurring in a process, the candidate selects 'increase inspection' or 'add a final review' rather than changing the process, and freely substitutes quality assurance for quality control in either direction.
Inspection feels visible and immediately actionable, while assurance work such as audits and process improvement looks indirect and slow. Cost-of-quality categories are memorised without the prevention, appraisal and failure distinctions that make the answer obvious.
What to do instead
Decide what is being measured: a process producing defects points to quality assurance (audit, root cause, process change, prevention cost); a product being checked against requirements points to quality control (inspection, testing, appraisal or failure cost). Where the defect recurs, prevention beats more inspection - choose the option that changes the process and then re-verifies the output.
For a stakeholder or team problem, the candidate picks a process answer - issue a weekly status report, hold more meetings, escalate to the sponsor - instead of analysing the specific stakeholder's interest, influence and current engagement level, or addressing the root cause of the conflict.
Communication volume feels like engagement and escalation feels safe, so the analysis step is skipped. The options rarely name the analysis explicitly, and candidates default to the action they would take in their own workplace rather than the action the scenario's evidence supports.
What to do instead
Identify the stakeholder's power and interest, then the gap between current and required engagement level (unaware, resistant, neutral, supportive, leading) before choosing. Match the method to the need: push for one-way information, pull for capability building, interactive for decisions and conflict. For conflict, prefer problem-solving and collaboration, reserve forcing for safety or regulatory imperatives, and escalate only where the issue exceeds your authority.
For a hazard scenario, the candidate chooses 'provide PPE and run a toolbox talk' or 'retrain the crew' instead of a higher-order control such as elimination, substitution or engineering modification, and overlooks statutory notification and record-keeping duties triggered by the incident.
PPE and training are the cheap, familiar and visible controls, so the hierarchy of controls is recited but not applied under pressure. Regulatory duties such as reporting a serious incident, permits, consents and environmental limits are treated as administrative detail rather than mandatory obligations with deadlines.
What to do instead
Apply the hierarchy in order - eliminate, substitute, engineering control, administrative control, PPE last - and reject options that rely on behaviour change where a physical or design fix exists. Then ask what the regulator requires: notification to the enforcing authority for serious incidents, records and registers, permits and consents, and environmental monitoring, checking whether the option states the required deadline and recipient.