Lecture 4: Data-driven fraud detection | Quizlet

What is the difference between errors and fraud in data analysis?

What is the difference between errors and fraud in data analysis?

Errors: Unintentional issues due to system failures or procedural lapses. Spread evenly in data.

Fraud: Intentional actions to deceive, involving circumvention of controls and creation of false documents, often concentrated in specific areas of data.

How do auditors detect fraud using data analysis?

Traditional methods: Statistical sampling and manual checking.

Fraud examiners: Prefer full-population analysis to thoroughly identify fraudulent patterns.

What is the reactive approach in data-driven fraud detection?

Investigation starts after receiving a tip or detecting an anomaly.
Relies on reacting to a potential issue rather than actively searching for it.

What is the proactive approach in data-driven fraud detection?

Involves actively searching for potential fraud schemes and symptoms.
Investigators do not wait for a tip-off and use data analysis to detect irregularities.

Why is understanding the business important in fraud detection?

Different businesses have unique risks and operations, requiring tailored fraud detection procedures.

What is the role of risk assessment in fraud detection?

Identifying potential types of fraud, how they occur, and their symptoms to better focus detection efforts.

What does it mean to catalog possible fraud symptoms?

Listing the indicators or red flags that may suggest fraud based on potential scenarios identified.

How do investigators use technology in fraud detection?

They extract data from databases and other sources to identify symptoms and patterns of fraud.

What happens during the analysis phase of the data-driven fraud detection process?

Examining identified anomalies to determine if they are likely indicators of fraud.

What is the final step in the proactive fraud detection process?

Investigating the most promising indicators to confirm or rule out the presence of fraud.

What is digital analysis in fraud detection?

Analyzing the digits of numerical data to identify unusual patterns, such as fictitious invoices.

How does Benford's Law help in fraud detection?

It predicts the frequency of digits in naturally occurring datasets. Deviation from expected patterns may indicate manipulation.

What are the advantages of using Benford's Law in fraud detection?

It is inexpensive, allows profiling of specific cases, and suspects are less likely to be aware of its use.

What are the limitations of Benford's Law?

It only broadly indicates possible fraud, requiring additional tests, and is most effective with large datasets.

What is outlier investigation in fraud detection?

It involves calculating a z-score to determine how far a value deviates from the norm, with values >3 warranting further investigation.

How is a z-score calculated?

Z=(Value−Mean)/StandardDeviation. It normalizes data to help identify outliers.

What is stratification in data analysis?

Splitting complex datasets into groups (e.g., by vendor) to focus on specific categories for deeper analysis.

What is summarization in data analysis?

Performing calculations (e.g., averages, totals) on groups of data to produce summary records, aiding in analysis.

What is time trend analysis in fraud detection?

Evaluates changes in data over time, using regressions to predict values and spot deviations from expected trends.

What is fuzzy matching and how is it used?

It finds approximate matches between text entries, useful for identifying similar names or addresses that could indicate dummy companies.

What is Soundex in fuzzy matching?

An algorithm that finds words that sound similar, helping to identify names that may be misspelled or altered.

What is N-grams in fuzzy matching?

Divides words into smaller parts (trigrams) for comparison, aiding in matching similar-sounding or misspelled names.

How is real-time analysis used in fraud detection?

Integrated into transaction systems to monitor data as it flows, allowing for immediate detection of anomalies.

Why is financial statement analysis important in fraud detection?

It helps identify unexplained changes or discrepancies in key accounts, suggesting potential fraudulent activities.

What is vertical analysis in financial statement analysis?

Converts financial statement numbers into percentages of a single base figure, making it easier to spot anomalies within a period.

What is horizontal analysis in financial statement analysis?

Compares figures over different periods to identify significant changes, helping to detect fraud over time.

What is the quick ratio (acid test) in liquidity analysis?

QuickRatio = (Cash+AccountsReceivable) / CurrentLiabilities. Measures a company's ability to cover short-term obligations without selling inventory.

What is the profit margin ratio?

ProfitMarginRatio = NetIncome / NetSales. Indicates how much profit is made for every dollar of sales.

Why is it important to compare account balances over time?

It helps to identify unexpected changes that may suggest fraud, like sudden increases in expenses without a corresponding rise in revenue.

How can anomalies in COGS (Cost of Goods Sold) indicate potential fraud?

If COGS increases faster than sales, it may suggest inventory theft, misstatements, or inaccurately recorded transactions.