Page 82 - Accounting Information Systems

CHAPT E R 2 Introduction to Transaction Processing 53

The Digital Audit Trail
Let’s walk through the system represented in Figure 2-11 to illustrate how computer files provide an audit
trail. We begin with the capture of the economic event. In this example, sales are recorded manually on
source documents, just as in the manual system. The next step in this process is to convert the source
documents to digital form. This is done in the data-input stage, when the transactions are edited and a
transaction file of sales orders is produced. Some computer systems do not use physical source docu-
ments. Instead, transactions are captured directly on digital media.
The next step is to update the various master file subsidiary and control accounts that the transaction
affects. During the update procedure, additional editing of transactions takes place. Some transactions
may prove to be in error or invalid for such reasons as incorrect account numbers, insufficient quantities
on hand, or customer credit problems. In this example, the system determines the available credit for each
customer from the credit file before processing the sale. Any records that are rejected for credit problems
are transferred to the error file. The remaining good records are used to update the master files. Only these
transactions are added to the archive file that serves as the sales journal. By copying the valid transactions
to the journal, the original transaction file is not needed for audit trail purposes. This file can now be
erased (scratched) in preparation for the next batch of sales orders.
Like the paper trail, this digital audit trail allows transaction tracing. Again, an auditor attempting to
evaluate the accuracy of the AR figure published in the balance sheet could do so via the following steps,
which are identified in Figure 2-11.
1. Compare the accounts receivable balance in the balance sheet with the master file AR control account
balance.
2. Reconcile the AR control figure with the AR subsidiary account total.
3. Select a sample of update entries made to accounts in the AR subsidiary ledger and trace these to
transactions in the sales journal (archive file).
4. From these journal entries, identify specific source documents that can be pulled from their files and
verified. If necessary, the auditor can confirm the accuracy and propriety of these source documents
by contacting the customers in question.

Documentation Techniques

The old saying that a picture is worth a thousand words is very applicable when it comes to documenting
systems. A written description of a system can be wordy and difficult to follow. Experience has shown
that a visual image can convey vital system information more effectively and efficiently than words.
Accountants use system documentation routinely, as both systems designers and auditors, The ability to
document systems in graphic form is thus an important skill for accountants to master. Five basic docu-
mentation techniques are introduced in this section: data flow diagrams, entity relationship diagrams, sys-
tem flowcharts, program flowcharts, and record layout diagrams.

DATA FLOW DIAGRAMS AND ENTITY RELATIONSHIP DIAGRAMS
Two commonly used systems design and documentation techniques are the entity relationship diagram
and the data flow diagram. This section introduces the principal features of these techniques, illustrates
their use, and shows how they are related.
Data Flow Diagrams
The data flow diagram (DFD) uses symbols to represent the entities, processes, data flows, and data
stores that pertain to a system. Figure 2-12 presents the symbol set most commonly used. DFDs are used
to represent systems at different levels of detail from very general to highly detailed. In Chapter 14, we
will study the construction of multilevel DFDs. At this point, a single-level DFD is sufficient to demon-
strate its use as a documentation tool. We see an example of this in Figure 2-13.
Entities in a DFD are external objects at the boundary of the system being modeled. They represent
sources of and destinations for data. Entities may be other interacting systems or functions, or they may