BCSP CSP common mistakes
BCSP CSP · Understand the traps and how to avoid them.
Computing exposure statistics with the wrong denominator or averaging method - dividing incident counts by total hours worked and calling it an incidence rate, taking a plain arithmetic mean of air samples collected for unequal durations, or interpreting overlapping confidence intervals as proof of 'no difference'.
Candidates memorize the shape of the formula but not the base or the assumptions behind it, and under time pressure they grab the familiar number rather than setting up the calculation from the data given in the stem.
What to do instead
Before calculating, write three things on scratch paper: the numerator definition, the base (incidence rates use 200,000 hours = 100 workers x 40 hours x 50 weeks), and the time basis. For air sampling, convert each sample to a time-weighted contribution (concentration x hours / total hours) rather than averaging results. For intervals, state to yourself what '95% confidence' actually means before choosing an option.
Plugging numbers into an engineering relationship without checking units or the physical assumption behind it - mixing cfm with cubic meters per minute in a dilution ventilation calculation, adding sound levels arithmetically instead of logarithmically, or applying a room-average dilution model where capture velocity at the hood is what the stem is asking about.
The stem usually supplies numbers in the units the candidate expects, which encourages pattern-matching to a memorized equation instead of deriving what is being asked for; the candidate also skips the conceptual check of whether the equation's assumptions fit the scenario.
What to do instead
Carry units through every step and cancel them explicitly. Before selecting a formula, name the control concept in the stem (capture, dilution, transport, source control) and confirm the equation matches it. For decibels, use the exchange rate stated by the standard in play rather than simple addition.
Answering with an activity instead of a system element - labeling a safety inspection as an 'audit,' calling a lagging indicator a leading indicator, or offering 'retrain the crew' when the stem describes a deficient management review, unclear policy, or missing accountability structure.
Candidates default to the visible, familiar activity they see at their own site and lose track of where that activity sits in the plan-do-check-act cycle or in the hierarchy of system elements (policy, planning, implementation, evaluation, correction, management review).
What to do instead
For every management-system item, first name the PDCA phase the stem describes, then choose the option that operates at that phase. Distinguish audit (conformance to a criterion, by a competent, ideally independent party, with findings and corrective action) from inspection (physical condition check). Classify each metric as leading (predictive, upstream) or lagging (outcome) before reading the options.
Confusing hazard with risk and picking the wrong risk-treatment option - choosing 'transfer' when the stem describes eliminating the task, choosing 'accept' when the stem shows an uncontrolled high-severity exposure, or stopping the analysis at the initial risk score without considering residual risk after controls.
Risk vocabulary sounds interchangeable in everyday speech, and candidates answer the scenario they imagine rather than the specific question of which treatment step or which stage of the assessment the stem is at.
What to do instead
Re-state the stem in one line before answering: what is the hazard, what is the consequence/likelihood, and is the question asking for identification, analysis, evaluation, or treatment? Apply the treatment options in order (avoid, reduce by changing likelihood or consequence, share/transfer, retain) and always re-score residual risk before selecting the final answer.
Mixing up occupational exposure limits and their sources - treating an ACGIH TLV as an OSHA PEL, applying a 3 dB exchange rate to an OSHA noise compliance question (or 5 dB to a NIOSH-based one), using dry-bulb temperature for a heat-stress evaluation, or answering an ergonomics stem with a medical-management action when it asks for the ergonomic risk factor.
The limits share similar numerical formats, so recognition memory fires before the candidate checks which organization or standard the stem invokes; ergonomics items are also frequently misread as clinical items.
What to do instead
Before answering any exposure-limit item, identify the issuing body named or implied in the stem and recall that limit's basis. Keep a one-page personal table of PEL, action level, TLV, and REL values with their exchange rates and averaging periods, and for ergonomics stems, sort the options into risk factors (force, posture, repetition, vibration, contact stress, duration) versus controls versus medical surveillance.
Selecting an analysis technique or cause level that does not match the question - proposing a fault tree for a single already-occurred event, using a JSA to explain a multi-causal incident, or writing the immediate (symptom) cause as the root cause and pairing it with a person-focused corrective action.
Technique names are learned as a list rather than as inductive versus deductive tools with different inputs, so candidates match on keywords; and investigating near-misses mentally is easier than tracing causal chains to management-system factors.
What to do instead
Classify the technique before choosing: deductive top-down from an undesired top event (fault tree), inductive bottom-up from component failure (FMEA), task-based prospective (JSA/JHA), or simple causal chain (5-why, cause-and-effect). Then sort candidate causes into immediate, contributing, and root/systemic, and verify that the corrective action removes a cause you actually identified rather than the person present.
Choosing a lower-order control as the 'best' answer - selecting PPE, additional training, or a new warning sign when elimination, substitution, or an engineering control is technically and economically feasible in the scenario.
Administrative controls and PPE are what most candidates see implemented in practice, so they feel realistic; options that require redesign feel impractical even when the stem signals feasibility.
What to do instead
Rank every option against the hierarchy (elimination, substitution, engineering, administrative including training, PPE) and select the highest-ranked feasible option, not the most familiar one. Watch for qualifiers like 'most effective,' 'first step,' or 'initially' - 'first' often points to hazard anticipation or elimination of the task, not to the last line of defense.
Mismatching fire chemistry and suppression - treating a flammable liquid's flash point, fire point, and autoignition temperature as interchangeable, confusing the lower and upper flammable limits, or selecting water, CO2, or an ABC dry chemical for a combustible metal fire or an energized electrical fire.
Fire terminology is learned as a vocabulary list without the physical mechanism attached, and suppression agents are remembered by brand familiarity rather than by the class of fire and the agent's mechanism (cooling, smothering, chemical interruption, oxygen displacement).
What to do instead
Anchor each term to its mechanism: flash point is the temperature at which a vapor ignites momentarily, the fire point sustains combustion, and autoignition needs no ignition source. Before choosing an agent, classify the fuel (A, B, C, D, K) and check the agent against that class and against energized-equipment status; remember metal fires require a class D dry powder applied by a compatible technique.
Blurring emergency plan requirements - treating the emergency action plan and the fire prevention plan as the same document, assuming a written plan is always required, choosing evacuation when the scenario calls for shelter-in-place or defend-in-place, or omitting the requirement to account for all employees after evacuation.
Both plans are usually one binder at the candidate's own site, so the regulatory distinction is invisible in practice; and candidates answer the most dramatic response rather than the risk-appropriate one.
What to do instead
Separate the two plans by purpose: emergency action covers reporting, evacuation routes and procedures, accounting for all employees, rescue and medical duties, and contacts; fire prevention covers hazard control, housekeeping, training, and control of ignition sources. Check the small-employer exemption before selecting 'written plan required,' and choose the protective action (evacuate versus shelter) based on the hazard, time available, and building characteristics given in the stem.
Mixing legal and ethical obligations - treating first-aid-only cases as recordable, citing the General Duty Clause where a specific standard applies (or vice versa), missing recordkeeping reporting timelines, or signing off on a plan or certification outside one's competence and failing to protect confidential or privileged information.
The legal items reward precise recall of thresholds and exceptions, which candidates often blur, and ethics items feel like common sense, so candidates answer with the outcome that benefits the employer or the client rather than the obligation stated in the code of ethics and the standard of care.
What to do instead
Memorize the recordkeeping decision path (injury or illness, work-related, new case, meets a recording criterion; first aid is not medical treatment) and the reporting timelines for fatalities and severe injuries. When two standards could apply, cite the specific standard and reserve the General Duty Clause for recognized hazards with no standard. On ethics items, identify the duty owed, to whom, and whether competence, confidentiality, or public safety is engaged before reading the options.