| Course | ACC 427 Investigating with Computers |
|---|---|
| Module | Module 3 |
| Paper type | undergraduate data analysis assignment applying forensic tests to transaction data |
| Length | About 1,010 words, 6 pages |
| Format | APA 7 student paper |
| School | Southern New Hampshire University |
| Program | BS Accounting |
| Updated | October 2026 |
Free sample paper for ACC 427 Module 3
Voids, Timing and One User ID: Five Forensic Tests on Landfill Scale-House Data
[Student Name]
Southern New Hampshire University
ACC 427: Investigating with Computers
Module Three Assignment
[Instructor Name]
[Date]
The organization, setting and figures below are a composite written as a model document. No real employer, client, colleague or patient is described.
Voids, Timing and One User ID: Five Forensic Tests on Landfill Scale-House Data
Introduction
The validated scale data from the Pennsylvania landfill contain 412,600 tickets over 22 months, 118,300 of them paid in cash, and 3,214 voids. Module Two showed that cash recorded net of voids matches bank deposits, so any skimming would most likely appear as voids of cash tickets after the money was collected. This assignment runs five tests on the void data, each designed to separate legitimate corrections from voids that remove completed sales, and interprets the combined results (Nigrini, 2020).
Test 1: Voids by User, Against Opportunity
Of the 3,214 voids, 1,604 involved cash tickets. Comparing each operator's share of cash voids with his share of cash-ticket shifts adjusts for the fact that some operators simply work more.
Table 1. Cash Voids Compared With Shifts Worked
| Operator | Share of cash-ticket shifts | Cash voids | Share of cash voids |
|---|---|---|---|
| Operator 3 | 31% | 1,148 | 72% |
| Operator 1 | 27% | 162 | 10% |
| Operator 2 | 24% | 171 | 11% |
| Operator 4 | 18% | 123 | 8% |
| Total | 100% | 1,604 | 100% |
Operator 3 voided cash tickets at more than five times the rate of his colleagues per shift worked. His voids of check and card tickets, by contrast, were in line with theirs, which matters because voiding a check or card payment does not produce money anyone can keep.
Test 2: Time From Sale to Void
A legitimate correction, such as a mistyped weight, is usually fixed immediately. A skim requires the customer to pay and leave before the ticket disappears. The median time from ticket creation to void was one minute for Operators 1, 2 and 4 and 14 minutes for Operator 3; 86 percent of his cash voids came more than five minutes after the ticket, compared with 9 percent for the others.
Test 3: Voids by Hour
Eighty-one percent of Operator 3's cash voids occurred after 3 p.m., when the scale-house supervisor leaves for the day and the late operator works alone until closing at 5 p.m. The other operators' voids were spread evenly across the day.
Test 4: Reason Codes
Operator 3 entered reason code CUST, customer dispute, on 1,091 of his 1,148 cash voids. Customer disputes are supposed to generate a dispute form in the customer service system. A join between the void table and the dispute log found matching dispute records for 12 of them. The other operators used CUST rarely and had dispute records for most of those uses.
Test 5: Customers on Voided Tickets
Voided cash tickets clustered among 46 small contractors and residents who paid cash regularly. Several appeared dozens of times. Those repeat customers are the most practical source of corroboration, because they can be identified from their truck plates and could confirm that they paid.
The Combined Pattern
Each test alone has innocent explanations. A new or careless operator might make more corrections; a late shift might see more disputes. Together, the pattern is specific: one user, cash only, long after the sale, late in the day when unsupervised, with a reason code unsupported by dispute records. The voided fees on his cash tickets total $212,400 over 22 months, an average of $185 per ticket, close to the typical charge for a small contractor's load. Debreceny and Gray (2010) argue that the value of data mining in fraud work lies in combining weak signals into a pattern no single test would reveal, which is what the five tests do here.
How the Tests Were Run
The tests were run in a data analysis tool on the joined table described in Module Two, with every step saved as a script. Test 1 summarized cash voids by user ID and divided each user's count by his count of cash-ticket shifts from the schedule. Test 2 subtracted the ticket creation time from the void time for every void and summarized the results by user. Test 3 grouped voids by hour of the day. Test 4 joined the void table to the customer service dispute log on ticket number and counted matches. Test 5 grouped voided cash tickets by truck license plate. Each script was run twice, once by the lead analyst and once by a second analyst working independently, and the results matched. Keeping the scripts means that if counsel or an opposing expert questions a figure, it can be regenerated from the preserved data in minutes, and anyone reviewing the work can see exactly what was counted.
Estimating the Amount
The $212,400 is the sum of fees on Operator 3's voided cash tickets. It is an upper estimate of what may have been taken through this method, not a proven loss, because some of those voids may be genuine corrections. A more conservative estimate excludes voids within two minutes of the sale, the window in which the other operators made almost all of their corrections. That leaves 987 voids and $183,100. Both figures will be presented, so that the company and counsel understand the range before any recovery or insurance claim is made.
What the Data Do Not Show
The data show that the operator's user ID voided the tickets. They do not show who physically did it, whether the trucks actually unloaded or whether the customers paid. Shared logins, a known weakness at many scale houses, could mean another person used his ID. Appelbaum et al. (2017) note that analytics in assurance produce leads whose evidential weight depends on corroboration. Three steps follow: review video for a sample of voided tickets still within the camera's retention period, check whether Operator 3's ID was ever logged in while he was off shift, and, through counsel, ask a sample of repeat customers whether they paid.
Conclusion
Five tests on the full population of voids isolate a pattern concentrated in one operator's user ID: 72 percent of cash voids on 31 percent of shifts, a median of 14 minutes after the sale, mostly after the supervisor leaves and with an unsupported reason code. The voided fees total $212,400. The pattern is strong enough to direct the investigation but must be corroborated before any conclusion is drawn about who took the money.
References
Appelbaum, D., Kogan, A., & Vasarhelyi, M. A. (2017). Big data and analytics in the modern audit engagement: Research needs. Auditing: A Journal of Practice & Theory, 36(4), 1-27. https://doi.org/10.2308/ajpt-51684
Debreceny, R. S., & Gray, G. L. (2010). Data mining journal entries for fraud detection: An exploratory study. International Journal of Accounting Information Systems, 11(3), 157-181. https://doi.org/10.1016/j.accinf.2010.08.001
Nigrini, M. J. (2020). Forensic analytics: Methods and techniques for forensic accounting investigations (2nd ed.). Wiley.
What the ACC 427 Module 3 instructions ask for
The Module Three assignment in ACC 427 usually asks you to apply analytical tests to a data set to identify possible fraud. Expect tests such as stratification and summarization by user or vendor, duplicate and gap tests, time-based analysis, comparison of a group with the rest of the population and joins with other tables. For each test, state what it is designed to find and why, describe how it was run, report results against a baseline and interpret the findings, including innocent explanations. Many versions ask you to rank results and recommend follow-up. A table of tests, results and interpretations, followed by prose that ties them together, usually fits the assignment well.
How this ACC 427 Module 3 data analysis assignment example is built
The sample tests voids in a landfill's scale data. By user, one operator made 72 percent of cash voids though he worked 31 percent of cash shifts; other operators' voids are spread across all payment types. By timing, other operators void a median of one minute after the ticket, consistent with correcting a weight entry, while his median is 14 minutes, long enough for a truck to unload and leave. By hour, 81 percent of his voids come after 3 p.m., when the supervisor has left. His voids use the reason code for customer dispute, but no dispute records exist. Repeat customers appear on many voided tickets. The voided fees total $212,400. The paper notes that video review is needed to confirm.
Where the ACC 427 Module 3 rubric puts the points
Rubrics for the ACC 427 data analysis assignment typically score test design and logic, execution, presentation of results with baselines, interpretation and recommended follow-up. Top papers explain what each test would show if fraud were present and if it were not, compare a suspect group with a meaningful baseline, report both flagged counts and their share of the population and acknowledge innocent explanations. Graders reward tests that build on each other toward a coherent pattern and a short list of next steps that would confirm or rule it out. Common deductions include reporting raw counts without a baseline, treating correlation with one employee as proof and failing to recommend corroborating evidence outside the data.
ACC 427 Module 3 help: the mistakes that cost points
Data analysis papers lose points most often by reporting that one person had the most voids without asking whether he also worked the most shifts, and by presenting each test in isolation instead of showing how they combine into a pattern. Another common gap is forgetting innocent explanations, such as a newer employee who makes more entry errors. If your data involve purchases, payroll or expense reports, the same discipline applies: define the test, set a baseline, report flags and explain them. Ask of every result what a fair-minded skeptic would say about it, and answer that in the paper before your grader does; the answer is usually a second test or a piece of outside evidence.
Get ACC 427 Module 3 written to your instructions
Send the ACC 427 Module 3 data description and instructions. The paper will design each test with its logic, run it on the population, report flags and expected baselines, and explain what the results show and do not show. Your first one costs nothing; allow about two days. The paper above is an original model document written by our desk, not a submitted student paper and not an official Southern New Hampshire University document.
More ACC 427 papers and related BS Accounting samples
- ACC 427 Module 1 Discussion: Where the Digital Evidence Lives
- ACC 427 Module 2 Data Acquisition Assignment: Getting a Complete Copy of the Scale Data
- ACC 427 Module 4 Project One: Fuel Cards Matched to Truck GPS
- FIN 320 Module 1 Financial Manager Discussion
- ACC 317 Module 6 Discussion: What Would Change Under IFRS
- ACC 345 Module 5 Forecasting Assignment: Pro Forma Statements for Next Year
- ACC 201 Module 3: sample paper, in real form
ACC 427 Module 3 questions, answered
Where can I find a free ACC 427 Module 3 data analysis sample?
This page holds a full ACC 427 Module 3 assignment running five forensic tests on scale-house void data at a landfill.
Why compare a suspect's activity with a baseline?
Raw counts can mislead. Comparing a person's share of an activity with his share of opportunities, such as shifts worked, shows whether the activity is unusual.
What does time-to-void analysis reveal?
Legitimate corrections usually happen within seconds or a minute. Voids long after a sale suggest the transaction was completed and then removed.
Does a strong data pattern prove fraud?
No. It identifies where to look. Corroborating evidence, such as video, documents or interviews, is needed to establish what happened.
Which tests are common in forensic data analysis?
Summaries by user or vendor, duplicate and gap tests, Benford and number tests, time-based tests and joins between tables or systems.