CFI (FAA) common mistakes
CFI · Understand the traps and how to avoid them.
Answering 'NFPA 921 requires...' when the item is only a recommendation of a guide, or mixing up NFPA 921 with NFPA 1033 - e.g., treating the 921 chapters as mandatory 'shall' requirements or naming 1033 as the source of the scientific method rather than the professional-qualifications standard.
Candidates memorize buzzwords and chapter titles without internalizing that 921 is a guide (recommended practice) while 1033 is a standard listing job performance requirements; exam items exploit the assumed equivalence.
What to do instead
Build a one-page comparison: NFPA 921 = guide, methodology and technical content (scientific method, hypothesis testing, origin/cause, explosion, electricity, appliances/vehicles); NFPA 1033 = standard, the 16 JPRs and knowledge/skill levels. Before answering any 'which document' item, ask whether the stem describes a method (921) or a qualification (1033).
Selecting a cause by elimination - answering 'electrical' or 'arson' because no other ignition source was found, rather than because positive evidence supports that hypothesis.
Elimination feels rigorous and mirrors how many field investigators narrate conclusions; the exam tests whether the candidate recognizes negative corpus as an invalid methodology.
What to do instead
For any cause item, require a stated piece of affirmative evidence for the surviving hypothesis (arc mapping plus a competent ignition source plus a first-fuel connection, or an accelerant pour pattern plus ignitable liquid residue). If no hypothesis can be positively supported, the correct answer is undetermined.
Using flashover, ventilation-limited burning, or 'alligatoring' as a point-of-origin or cause indicator - e.g., choosing the area of greatest char as origin, or selecting 'flashover' as the explanation for a specific burn pattern.
Damage patterns are visually vivid and intuitive, so candidates substitute pattern-reading for dynamics reasoning; the item usually tests whether the candidate knows flashover and ventilation obscure rather than reveal origin.
What to do instead
For pattern items, first classify the fire regime (fuel-limited vs. ventilation-limited) and the ventilation profile, then state that flashover and post-flashover burning consume early patterns. Treat char depth, alligatoring, and 'V' patterns as supporting evidence only, corroborated by witness, alarm, and suppression data.
Concluding an electrical cause from a single bead, a 'short', or a breaker in the tripped position, without distinguishing fire-caused (fire-induced) arcing from arc-induced fire, and without confirming the circuit was energized at the time.
Electrical terms are familiar and a tripped breaker feels like proof; candidates conflate evidence of arcing with evidence that arcing started the fire.
What to do instead
Apply a three-part test in every electrical item: (1) was the conductor energized and was the overcurrent device intact at the time; (2) is the damage bead/melting consistent with high-temperature arcing versus fire-melting; (3) does arc mapping place the arc in the origin area rather than downstream. If any part fails, the answer is not 'electrical failure.'
Treating a sprinkler or detection system as having operated (or failed) based on a single indicator - e.g., assuming the sprinkler controlled the fire because water was flowing, or that the alarm 'failed' because no one heard it - without evidence of activation time, coverage, obstruction, or water supply.
Fire protection systems are taught as code facts rather than as physical evidence, so candidates answer from the design intent instead of the observed performance data.
What to do instead
For any system item, separate design (coverage, spacing, hazard classification) from performance (activation, waterflow/pressure, valve position, obstruction, impairment). Answer only from the physical evidence described; if the stem gives no performance evidence, the answer must be that system performance cannot be determined.
Mixing up explosion classifications - calling a gas deflagration a detonation, attributing backdraft or smoke explosion to a fuel leak, or failing to separate the pre-explosion fire from the post-explosion fire in describing damage.
The vocabulary is overlapping and dramatic, and candidates reason from damage severity rather than from confinement, fuel/air mixing, and ignition timing.
What to do instead
Use a decision sequence: fuel type and state, confinement and venting path, whether a flammable atmosphere existed before ignition, and whether injury/damage patterns show a pressure front (detonation) or a slower pressure rise (deflagration). Explicitly label pre- and post-explosion fire in the answer.
Answering evidence-collection items with methods that compromise the sample - ignitable liquid debris in plastic bags, samples taken from the wrong depth or without a comparison sample, items collected before being photographed in place, or a chain of custody that is described but not continuous.
Candidates focus on the analytical result they expect rather than on the admissibility and reliability of the sample, and they skip documentation because it feels administrative.
What to do instead
Memorize the sequence: photograph with scale and orientation, document location and condition, collect in a clean airtight container appropriate to the analysis, take a comparison/background sample, seal and label, and log every transfer. Any answer that skips a step in that chain is wrong regardless of the chemistry.
Interviewing errors framed as answers - endorsing group interviews of multiple witnesses, feeding details of the fire to the witness, or asking leading/closed questions that confirm the investigator's hypothesis.
Candidates treat the interview as a formality for gathering facts rather than as a memory-contamination risk, and they underestimate how much the investigator's own hypothesis shapes the statement.
What to do instead
Use the sequence: separate witnesses before interviewing, free recall first, open-ended prompts, then specific non-leading questions, and record the interview in the witness's own words. Any item where the investigator supplies information the witness did not volunteer should be rejected.
Giving opinion testimony in the report or on the stand that exceeds the underlying data - opining on cause without stating the hypotheses tested, or phrasing conclusions as 'consistent with arson' without ruling out accidental causes.
Candidates blur the roles of fact witness, origin-and-cause investigator, and expert, and are trained to write conclusory reports rather than transparent ones.
What to do instead
Separate the report into facts observed, methodology applied, hypotheses considered, and opinion with its basis and confidence level. State the reliability foundation (training, data, testing, peer review) and be prepared to give the basis for every opinion; an opinion without a stated basis is the wrong answer.
Treating scene safety and role/ethics questions as soft items - entering a post-suppression scene without atmospheric monitoring or structural assessment, or accepting an assignment that combines investigative and adjusting/financial roles without disclosure.
Candidates prioritize the investigation objective and answer from what would be most efficient, forgetting that safety and independence are non-negotiable and are tested with concrete facts.
What to do instead
Apply a fixed pre-entry checklist (structural stability, atmospheric monitoring, utilities isolated, PPE, scene security) and a conflicts screen (prior involvement, financial interest, employment relationship) before any technical answer. If an item describes an unsafe or conflicted scene, the controlling answer is to stop, secure, or disclose.