One late approval in twenty-five, plus four changes the log never captured: those findings, evaluated for cause rather than counted, end reliance on a rate-change control in AC503's fourth-unit workpaper. Searches like "ac 503 unit 4 assignment example", "ac503 unit 4 sample" and "ac503 unit 4 example" land here.
What a finished AC503 Unit 4 control testing workpaper looks like
Four indexed pages with a standard header: client, year end, index, objective, preparer in brackets and a blank reviewer block. Page one quotes the control as the client's procedure manual words it, names the risk it addresses, billing at rates customers never agreed to, and lists the attributes tested: approval by the pricing manager, approval dated before activation, and agreement of the new rate to a signed contract amendment. Page two tests the population before any sampling, comparing month-end copies of the rate master file by script against the change log. Page three records twenty-five changes selected at random from the 214 logged, with one exception: an approval dated three days after the new rate had already billed eleven loads. Page four evaluates both findings and concludes.
How a AC503 Unit 4 example is structured
Completeness comes before selection, because a sample drawn from a log that misses changes cannot speak for the changes it misses. The comparison script and its output stay in the workpaper so a reviewer can rerun them. That test found four changes absent from the log, all entered through an administrator account assigned to the sales vice president, all raising accessorial rates for three customers above contract, together billing 186,400 more than the contracts allowed. The sample page follows, attribute by attribute. Evaluation then treats the two findings differently. The late approval is a deviation, investigated for cause: a manager on leave with no delegate. The four unlogged changes are not sample deviations at all but evidence that the control can be bypassed, pointing to an access failure in general controls and to possible override. The conclusion withdraws reliance and states the substantive consequence.
The procedure manual's wording
Every rate change approved by the pricing manager in the billing system before activation, quoted from the procedure manual word for word.
Is the log complete?
Month-end copies of the rate master compared by script against the change log, turning up four changes the log never recorded.
Twenty-five changes, three attributes
Approver identity, approval before activation, and the new rate agreed to a signed contract amendment, tested item by item.
One late, four missing
A late approval traced to leave with no delegate, and four administrator-account changes that raised accessorial rates 186,400 above contract.
Reliance withdrawn
No reliance on the control this year, invoices for the three customers recomputed at contract rates, and deficiencies reported in writing.
Where marks go in AC503 Unit 4
Sampling a population nobody tested for completeness is the error graders mark first in this workpaper, since the four missing changes were visible only to a test run before selection. Deviations reported as a rate, one in twenty-five, with no cause examined, earn little at this level. The four bypassed changes get misread in two directions: folded into the sample as extra deviations, which treats an access failure as a sampling result, or noted as a curiosity and left without consequence. Graders expect the vice president's access and bonus arrangement weighed as a possible override and reported to those charged with governance. Concluding reliance despite the bypass costs heavily. Workpapers without the script output, or with a reviewer's initials filled in by the preparer, lose documentation marks.
Get a AC503 Unit 4 example written to your instructions
Share the control your Unit 4 prompt describes, the population data and any testing template your section uses, plus the rubric. Completeness is tested first, then attributes item by item, and every exception is evaluated for cause, in a workpaper returned within 24-48h. The first custom sample is free, with reviewer lines left open.
AC503 Unit 4 questions, answered
Why test the change log before sampling from it?
Because a sample can only represent the population it was drawn from. If the log omits changes, a clean sample says nothing about the omitted ones. Information produced by the client and used in a control test needs its own evidence of completeness and accuracy, and the AC503 example obtains it by comparing independent copies of the rate file against the log.
Are the four unlogged changes deviations?
Not in the sampling sense. They show the control could be bypassed entirely through an administrator account, which is a failure of access controls and possibly management override, not an instance of the pricing manager forgetting to approve. The example evaluates them separately from the sample, and that separation carries much of the workpaper's grade.
How large should a sample be for a control like this?
Firms publish their own guidance, typically scaling sample size to how often a control operates and how much reliance is planned. A control performed a few hundred times a year often draws around twenty-five items. The example cites a firm table as its basis, states the planned reliance, and notes that your section may set a different figure.