CRSA / RSA Exam common mistakes
CRSA / RSA Exam · Understand the traps and how to avoid them.
Treating a hierarchy-of-controls question as a vocabulary recall item: choosing 'PPE' or 'training' as the answer whenever the hazard is described vividly, instead of ranking controls from elimination down to PPE and selecting the highest feasible level.
Candidates study the five-level hierarchy as a list to memorize and never practice applying it to scenario stems, so the emotional salience of a dramatic hazard description pulls them toward the control that protects the individual worker rather than the one that removes the exposure.
What to do instead
Before reading the answer options, restate the hazard in one line and write E-S-E-A-P (elimination, substitution, engineering, administrative, PPE) in shorthand; pick the option at the highest level that is technically and economically feasible for the described operation, and only drop to PPE when the stem says higher controls are infeasible or already in place.
Confusing risk assessment with hazard identification: submitting a list of hazards found on a walkthrough as the 'risk assessment,' or scoring likelihood and severity without ever documenting the basis for the scores.
The two activities are taught together and share vocabulary, and walkthrough checklists reward finding hazards, so candidates stop at the inventory stage and never build the exposure-and-consequence reasoning that distinguishes a risk assessment.
What to do instead
Practice writing two-sentence assessments: 'Hazard X, exposed population Y, exposure frequency Z, credible outcome W, likelihood rating with rationale, severity rating with rationale, resulting risk level, control decision, re-assessment trigger.' If any of those elements is missing from an answer option, it is not a complete risk assessment.
Selecting the most stringent regulatory citation when the stem asks what the employer must do, rather than identifying the performance-oriented duty (recognize, evaluate, control) that the standard actually imposes.
Candidates memorize numeric limits, recordkeeping days, and threshold quantities because they are easy to flashcard, then over-apply them to questions about program design where the standard sets a goal and leaves the method to the employer.
What to do instead
For each major standard you study, write three columns: absolute numbers (limits, days, thresholds), employer duties stated as verbs, and employee rights. When a stem asks 'what must the employer do,' answer from the duties column and reserve the numbers for stems that explicitly ask for a value.
Answering incident investigation questions with 'who was at fault' framing: recommending discipline, retraining the injured worker, or citing carelessness as the root cause.
Everyday reasoning about accidents is blame-based, and many workplace cultures reinforce it, so the intuitive answer feels correct even though modern investigation practice targets systemic causal factors and management-system deficiencies.
What to do instead
Force every causal chain to at least three levels of 'why' and require that the terminal cause be a system condition (design, procurement, scheduling, supervision, maintenance, training adequacy) rather than a person's trait. If an option names a worker's attitude as root cause, eliminate it.
Confusing the purpose of a training evaluation level: claiming a post-test score proves the training reduced injuries, or treating attendance records as evidence of competence.
Kirkpatrick-style levels are learned as a four-item list without the causal ordering, and organizations commonly report attendance and quiz scores as training effectiveness, so candidates adopt the metrics they have seen in practice.
What to do instead
Map each evidence type to its level explicitly: reaction (survey), learning (knowledge or skills test), behavior (supervisor observation on the job), results (injury rates, near-miss frequency, audit findings). A question asking whether training 'worked' on the job requires behavior- or results-level evidence, never a quiz score.
Treating exposure monitoring as a single air sample: recommending one personal sample as proof of compliance with an occupational exposure limit without addressing sampling strategy, duration, representativeness, or the difference between TWA, STEL, and ceiling values.
Candidates memorize the numerical limit and the analytical method but not the sampling design, and single-sample results are what appear in the case examples they have seen.
What to do instead
For each substance you review, build a card with the averaging period (8-hour TWA, 15-minute STEL, ceiling), the sampling and analytical method, and the decision logic: worst-case or representative sampling first, full-shift sampling for TWA comparison, short-term sampling for excursion comparison, and re-sampling when results exceed the action level.
Misreading the stem's actor and modality: answering what a regulator or consultant would do when the stem asks what the safety professional employed by the site should do, or answering a 'should' question with what is merely legally permitted.
Stems are long and candidates skim to the technical content, missing the last clause that names the role and the obligation level, and exam writers deliberately vary actor and modality to separate knowledge from judgment.
What to do instead
Read the final sentence of every stem twice and underline the actor, the obligation word (must, should, may, is required to), and the question being asked. Predict the answer before looking at options, then reject any option that is correct in general but wrong for that actor or obligation level.
Applying a single control approach to ergonomic and industrial hygiene problems alike: prescribing lifting technique training for a high-frequency manual handling task, or respirators for a solvent that could be captured at the source.
Training and PPE are the controls candidates see most often in their own workplaces, and the ergonomics and IH domains are studied as separate fact sets rather than as applications of the same control hierarchy.
What to do instead
For ergonomics, work through the hierarchy in physical terms: eliminate the lift (mechanical assist), reduce the load and frequency, adjust the workstation geometry, then condition and train the worker. For airborne contaminants, always ask first whether the contaminant can be captured or confined at the source before considering respiratory protection.
Answering emergency preparedness questions with a written plan as the deliverable, ignoring the planning process, drills, evacuation and shelter decision criteria, accountability, and post-incident critique.
Compliance cultures reward the existence of a document, so candidates equate 'has an emergency plan' with 'is prepared,' and the exam's process-oriented options look like unnecessary elaboration.
What to do instead
Memorize preparedness as a cycle: hazard analysis, capability assessment, plan development with roles, training, exercise and drill with evaluation, corrective action, revision. When a stem asks what demonstrates preparedness, choose evidence of tested capability (drill records with identified deficiencies and closed corrective actions) over evidence of documentation.
Choosing the answer that protects the employer's legal position over the answer that reflects professional ethics: minimizing an exposure result, delaying notification, or deferring to a client's preference to omit a finding.
Candidates study ethics as an abstract preamble and do not rehearse the concrete conflicts, so under time pressure they default to organizational loyalty and to the answer that avoids conflict.
What to do instead
Prepare four or five ethics scenarios in advance (suppressed monitoring data, unqualified acceptance of a client's risk tolerance, a signature requested on work you did not perform, a conflict of interest in a vendor relationship) and for each write the obligation, the disclosure path, and the documentation step. On the exam, when an option trades public or worker safety for organizational convenience, eliminate it.