This test may reveal unusually quick payments to certain vendors or slow payments that result in interest charges or early-payment discounts not taken.

Using the "State Vendor Payments" file, a DATE_DIFFERENCE field is created by using the @Age function in IDEA. This function calculates the differences in days between two date fields.

This equation should be used:


With over 2 million records in this file, we can stratify the date difference results into a number of bands from 0 to 120 days, as displayed in Figure 8.20.

Results of Stratification Payment and Invoice Date Differences

FIGURE 8.20 Results of Stratification Payment and Invoice Date Differences

From the stratification, we note that 826,000 payments were made on the same day as the invoice date. Testing of a sample of these transactions should be done. There were also 2,728 payments that were made before the invoice date, which is highly unusual. However, from the outset of reviewing this file, we had identified that the invoice date contains many input errors. Obvious errors should be dismissed and remaining anomalies should be reviewed.

Since we are looking at payment dates, this would be a good time to extract and review the 41 payments that were made on Sundays and the 64 payments that were made on Saturdays, as identified in our initial review of the field statistics for the date fields in Figure 8.2.

We can create a new field using the @Workday function that returns the number 1 if the payment date falls on a weekday or conversely returns a zero if the date falls on a weekend day. We can set the criteria or extraction to output only transactions falling on the weekend where the field contains a zero. In the event that some agencies or departments process payments on Saturdays, it would be better to segregate the two weekend days. We can use the @Dow or date of week function. This function will return, based on a date field, 1 for Sunday, 2 for Monday . . . and 7 for Saturday.

First we append a new numeric field to the "State Vendor Payments" file called DOW using the equation of


We then extract to a file called "Saturday and Sunday Payments," where DOW = 1 .OR. DOW = 7.

Since the file now contains a reasonable 105 records, we can clearly see in Figure 8.21 that the original INVOICE_DATE field had a number of blanks that gave us an error message relating to those blanks.

By filtering out the error or bad data, we end up with eight transactions that were paid on Sunday (DOW = 1) and 13 payments made on Saturday (DOW = 7), as displayed in Figure 8.22 .

According to the invoice date, most of these 21 payments were made by the Water Resources Board agency over a year after the invoice data. Three minor payments were made by Murray State College on one particular Sunday.

< Prev   CONTENTS   Next >