CFE investigative guidance recognizes online databases as valuable tools for locating public records quickly, but it also highlights practical limitations. The most significant limitation is that availability and coverage differ widely by jurisdiction. Some jurisdictions provide extensive online access to court files, property records, corporate registrations, and liens, while others provide limited access, partial indexing, or require in- person requests. This variability affects completeness and can create false negatives (records exist but are not available electronically). The guidance also cautions that online results may be abstracts, may not include supporting documentation, and can lag behind official updates-so investigators should verify key records with the originating agency when accuracy and completeness matter. Option A is not the primary limitation described; the more common problem is incomplete access rather than excessive detail. Option C is too absolute; errors can occur, but CFE methodology emphasizes validation, not the claim that records are "rarely correct." Option D is also overstated; while source credibility matters, the fundamental limitation is uneven jurisdictional availability and completeness.

The Manual defines leading questions as those that contain or suggest the answer:
"Leading questions... are framed in a way that evokes a specific reply... most commonly used to confirm facts already known... Example: 'You got promoted, right?'" Nathaniel's question, "When were you promoted to manager?" presumes the promotion occurred. Thisleads the subject by suggesting a fact that may not be true. It is not open (does not invite free narration), not complex (it asks one thing), and not closed (doesn't call for yes/no but presumes an answer).

"Credit card records... show card purchasing limits, previous payment history, account balance,when and where the subject has travelled, restaurants frequented, and assets acquired. In short, these records show the movements and habits of the person".
Therefore,Bis correct.

Within CFE investigation guidance on data analysis, predictive approaches are framed as using analytical techniques (including statistical processes and machine learning) to identify patterns and outliers that can indicate fraud risk and, importantly, help detect a fraudulent situation earlier than traditional methods.
Predictive analytics is not limited to one specific technology such as layered neural networks (A); neural networks can be one possible technique, but predictive analytics is broader than that. Option B describes robotic process automation (RPA)-software that imitates human actions to automate routine tasks-rather than predictive analytics. Option D describes artificial intelligence in a general sense (systems performing
"human tasks" like perception and decision-making), which is also broader and not specific to predictive modeling. The most accurate description is therefore C: predictive analytics uses historical data along with statistical analysis and machine learning methods to build models that reflect patterns/trends and generate projections about likely future outcomes (e.g., predicting which transactions, vendors, employees, or accounts are higher-risk).