CFI (Flooring Installation Certifications) common mistakes
CFI · Understand the traps and how to avoid them.
Recording an origin-and-cause determination as a single narrative paragraph that blends observed fire effects with the investigator's conclusions, so the file cannot show which statements are data and which are opinions.
Candidates write the report the way they would explain the case verbally, mixing observation and inference in one flow; under time pressure they skip separating the two because it feels repetitive.
What to do instead
Use a fixed two-column structure in every file: documented observations (with photo/evidence numbers and measurement units) on one side, hypotheses and the basis for each on the other. Before submitting, read the narrative and highlight every sentence containing 'because', 'likely', 'consistent with' or 'probable' - each must sit in the conclusion section with an evidence citation.
Treating the scene as a walk-through before establishing control: no entry log, no sketch before overhaul, and evidence moved before photographs are taken.
Candidates are eager to reach the area of origin and start digging; they assume the scene will stay stable and that memory will substitute for documentation.
What to do instead
Adopt a mandatory sequence: perimeter and entry log first, wide-to-close photography and a dimensioned sketch before any disturbance, then excavation. Keep a running evidence log with item number, description, location, collector and time. If a scene must be altered for safety, record the alteration and the reason in the log at the moment it happens.
Misidentifying the seat of the fire by reading the largest area of damage or the lowest burn as the origin, rather than reconstructing fire spread and ventilation effects.
Char patterns, floor-level burning and 'V' patterns are taught as shortcuts, and candidates apply them without accounting for ventilation-limited burning, flashover, or post-flashover damage that erases lower patterns.
What to do instead
Before fixing an origin, write out the fire dynamics narrative: ignition sequence, ventilation openings, fuel package, and whether flashover occurred. Confirm the origin hypothesis against at least two independent indicators (e.g., pattern plus witness timeline plus debris analysis) and explicitly state alternative hypotheses and why they were rejected.
Describing electrical causation from damage alone - declaring a receptacle, conductor or appliance 'the cause' because it is melted, arced or beaded, without distinguishing fire-caused damage from electrically caused damage.
Arc beads, melted insulation and severed conductors are visually dramatic and appear in almost every structure fire; candidates who lack electrical fundamentals read them as proof of an electrical fault.
What to do instead
Apply the standard test: electrical damage must be shown to be a cause, not an effect. Look for bead morphology and metallurgy, whether arcing is at a point of contact versus a severed end, and whether the circuit was energized at the time. Where you cannot distinguish, state 'electrical activity of undetermined significance' and send samples for laboratory examination instead of assigning cause.
Failing to secure and document the electrical system before it is altered - service disconnected, breakers repositioned, or wiring cut during suppression and overhaul - losing the ability to establish circuit status at ignition.
Suppression and safety priorities override preservation, and candidates do not know what data the electrical system holds or how quickly it is destroyed.
What to do instead
Photograph the panel and meter before any switching, note breaker positions and meter readings, and request the utility's records for the service. Record who disconnected the supply, when, and why. Treat the service entrance, panel and any suspect circuit as evidence to be mapped, not debris to be cleared.
Presenting a conclusion as certain when the physical evidence supports only a probable or possible determination, or the reverse - defaulting to 'undetermined' to avoid challenge.
Candidates conflate the confidence classification with personal confidence, and fear cross-examination on a definitive finding.
What to do instead
Use the four-level classification deliberately: cause determined, probable, possible, undetermined. For each, write the test that separates it from the next level (e.g., what evidence would move 'possible' to 'probable'). An 'undetermined' file must still list what was ruled out and what evidence was sought and not found.
Citing a standard, code section or scientific principle in the report without the edition, section number, or an accurate statement of what it actually says.
Candidates rely on remembered summaries and training slides; they cite the general idea of a standard rather than the operative text, and editions change.
What to do instead
Keep a citation log built from the current edition of each referenced document, quoting the section you rely on. Every technical claim in the report should map to a citation or to your own documented testing. Verify the edition in force at the time of the incident, not the edition you trained on.
Confusing the classes of fire and the associated agent or extinguishing method - for example treating a cooking-oil fire or a metal fire as an ordinary flammable-liquid or Class A problem.
Class A/B/C are learned first and over-applied; Class K and Class D situations are less common and their special agents are not intuitive.
What to do instead
Build a one-page matrix of class, fuel example, agent, and the failure mode of using the wrong agent (e.g., water on cooking oil, water on burning metal). Test yourself by fuel type rather than by letter, and state the class explicitly in every report where suppression method matters.
Giving testimony that goes beyond the scope of retained expertise, or offering opinions on ultimate legal issues without laying the foundation of methodology and data.
Candidates want to be helpful and answer the question asked; they do not rehearse the boundary between technical opinion and legal conclusion, or the basis requirement.
What to do instead
Prepare a written opinion outline with three layers for each opinion: the data relied on, the method applied, and the limits of the conclusion. In deposition and trial, answer only the question asked, and when a question exceeds your expertise say so and identify what discipline would address it. Rehearse challenges to your method and your data before testimony.
Overlooking insurance subrogation and liability implications when deciding what to collect and preserve - failing to identify all potentially responsible parties and their insurers early.
Candidates focus on the physical cause and treat insurance as an administrative afterthought; they do not know how spoliation exposure and subrogation recovery shape the evidence plan.
What to do instead
At case intake, list every potentially liable party (product manufacturer, installer, contractor, landlord, utility) and the insurer for each. Set the evidence retention plan accordingly, send preservation letters early, and document chain of custody so the file can survive a spoliation challenge.
Applying passenger-vehicle fire and collision investigation habits to marine or heavy-vehicle scenes - misreading fuel systems, battery banks, or hull and compartment ventilation as if they were automotive.
Most training and field exposure is automotive; marine compartments, diesel systems and 12/24-volt battery installations behave differently in ignition and fire spread.
What to do instead
Before scene work, identify the vessel or vehicle type, its fuel and electrical architecture, and its ventilation and compartment layout, and adjust the origin hypothesis accordingly. For marine cases, account for bilge accumulation, fuel-line and tank grounding, and shore-power connections; for heavy vehicles, account for auxiliary systems and aftermarket electrical additions.